Evidence ledger · Sổ dẫn chứng
Every number printed in the report, with who measured it, its status after checking, and a link to the source.
Mọi con số in trong báo cáo, kèm người đo, trạng thái sau khi đối chiếu và link tới nguồn gốc.
← Back to the report (EN) · ← Quay lại báo cáo (VI)
| Claim Nhận định | Verified value Con số đã kiểm | Source Nguồn | Status Trạng thái | Link Link |
|---|---|---|---|---|
| Adobe Analytics reported that U.S. consumers spent a record $257.8 billion online during the Nov. 1–Dec. 31, 2025 holiday season, up 6.8% year over year. Confirmed via Adobe's own January 2026 press release, verbatim figure and wording. This is the FINAL, full two-month season total, not a forecast and not a BFCM/Cyber Week figure — the A0 conflict note correctly flags that x#116 mislabels this as 'BFCM 2025'. | $257.8 billion online, up 6.8% YoY (US, Nov 1 – Dec 31, 2025, full holiday season) | Adobe (Adobe Digital Insights / Adobe Analytics) 2026-01-07 | Confirmed Đã xác nhận | ↗ |
| Adobe's October 2025 forecast projected the 2025 U.S. online holiday season (Nov. 1–Dec. 31) would reach $253.4 billion, up 5.3% year over year; the actual result beat this forecast at $257.8 billion (+6.8%). Confirmed as a pre-season FORECAST (not actual) — clearly distinct from A-S01's final figure. 'Actual beat it' framing in the claim is accurate. | $253.4 billion forecast, +5.3% YoY (Adobe's pre-season Oct 6, 2025 forecast for the full Nov 1–Dec 31 season); actual season came in higher at $257.8B/+6.8% | Adobe (Adobe Digital Insights) 2025-10-06 | Confirmed Đã xác nhận | ↗ |
| Adobe Analytics reported that Cyber Week (the five days from Thanksgiving through Cyber Monday) 2025 drove $44.2 billion in online spend, up 7.7% year over year. This is the FINAL actual figure, from Adobe's January 2026 season wrap-up. Adobe's pre-season Oct 2025 forecast for Cyber Week was different and lower ($43.7B, +6.3% YoY) — the two should not be conflated. | $44.2 billion online, up 7.7% YoY (US Cyber Week / Cyber 5, Thanksgiving–Cyber Monday, actual, final) | Adobe (Adobe Digital Insights) 2026-01-07 | Confirmed Đã xác nhận | ↗ |
| Adobe Analytics reported U.S. shoppers spent a record $11.8 billion online on Black Friday 2025, up 9.1% year over year from Black Friday 2024's $10.8 billion. Confirmed both in Adobe's day-of Nov 29/Dec 2, 2025 release and its final January 2026 season report — the 9.1% figure is unchanged/unrevised, so the A0 conflict note's suspicion that x#278's '+7.1%' is a mixed-up reference to Cyber Monday's growth rate (also | $11.8 billion online, up 9.1% YoY (US Black Friday 2025, actual, final) | Adobe (Adobe Digital Insights) 2026-01-07 | Confirmed Đã xác nhận | ↗ |
| Adobe Analytics reported Cyber Monday 2025 was the season's biggest online shopping day at $14.25 billion, up 7.1% year over year — beating Adobe's own pre-day forecast of $14.2 billion (+6.3% YoY). Both the forecast and the actual figures are genuine Adobe numbers from two different releases (pre-day forecast ~Dec 1, actual recap after Cyber Monday); they are not competing/contradictory, just forecast vs. final. | Actual: $14.25 billion, +7.1% YoY. Pre-day forecast: $14.2 billion, +6.3% YoY. | Adobe (Adobe Digital Insights) 2025-12-01 | Confirmed Đã xác nhận | ↗ |
| Adobe Analytics reported U.S. online spend of $10.8 billion on Black Friday 2024 and a then-record $13.3 billion on Cyber Monday 2024. Both figures are genuine 2024 Adobe totals, confirmed. The A0 conflict note is correct that x#259 (Nov 2025) misattributes these to 'new Omnisend' stats — that's a sourcing/attribution error in the original post, not an error in the numbers themselves. | Black Friday 2024: $10.8 billion (+10.2% YoY); Cyber Monday 2024: $13.3 billion | Adobe (Adobe Digital Insights) 2024-12-03 | Confirmed Đã xác nhận | ↗ |
| Adobe Analytics reported that 25 days saw U.S. consumers spend more than $4 billion online in a single day during the 2025 holiday season, up from 18 days in 2024. Exact match to Adobe's own wording. | 25 days with $4B+ online spend in the 2025 season vs. 18 days in 2024 | Adobe (Adobe Digital Insights) 2026-01-07 | Confirmed Đã xác nhận | ↗ |
| Adobe Analytics reported that mobile devices accounted for 56.4% of U.S. online holiday transactions in the 2025 season overall, rising to 66.5% on Christmas Day. Exact match. (Adobe's pre-season forecast had projected 56.1% mobile share for context — close to, but not identical to, the actual 56.4%.) | 56.4% of online transactions via mobile for the 2025 season overall; 66.5% on Christmas Day | Adobe (Adobe Digital Insights) 2026-01-07 | Confirmed Đã xác nhận | ↗ |
| Adobe Analytics reported that traffic to U.S. retail sites referred by generative AI sources grew 693% year over year during the 2025 holiday season. Confirmed growth rate. Adobe does not publish the absolute share of total retail traffic that AI referrals represent alongside this figure — only that 'the base of users remains modest.' The unsourced '4,700%' (x#18) and '6,900% agentic traffic' (li#45) figure | +693.4% YoY growth in generative-AI-referred traffic to US retail sites, 2025 holiday season (Nov 1–Dec 31) | Adobe (Adobe Digital Insights) 2026-01-07 | Confirmed Đã xác nhận | ↗ |
| Adobe Analytics reported that traffic to U.S. retail sites from AI sources increased 805% year over year on Black Friday 2025 and 670% year over year on Cyber Monday 2025. Both figures confirmed as ACTUAL results (Black Friday figure via Digital Commerce 360's report of Adobe Analytics data; Cyber Monday figure directly from Adobe's own Jan 2026 release), not pre-season forecasts. The A0 conflict note's claim that Adobe/Salesfor | AI-referred traffic to US retail sites: Black Friday +805% YoY; Cyber Monday +670% YoY (both actual) | Adobe (Adobe Digital Insights) 2025-12-01 | Confirmed Đã xác nhận | ↗ |
| Salesforce data showed U.S. online order volume fell about 1% year over year on Black Friday 2025 while the average selling price rose 7% and units per transaction fell 2%, per Salesforce director of consumer insights Caila Schwartz. Could not locate a standalone Salesforce web press release carrying these exact Black-Friday-only figures. Salesforce's own Dec 5, 2025 Cyber Week press release (fetched: https://www.salesforce.com/news/press-releases/2025/12/05/cyber-week-ai-agents-sales/) gi | US Black Friday 2025: online order volume -1% YoY; average selling price (ASP) +7% YoY; units per transaction -2% YoY | Salesforce (via Retail Dive) 2025-11-29 | Confirmed Đã xác nhận | ↗ |
| Mastercard SpendingPulse reported U.S. Black Friday 2025 retail sales (excluding autos) up 4.1% year over year, with e-commerce sales up 10.4% and in-store sales up 1.7%; apparel rose 5.7% overall (6.1% online, 5.4% in-store), jewelry rose 2.75%, and restaurants rose 4.5%. Mastercard's own press release (mastercard.com/us/en/news-and-trends/press/2025/november/...) returned HTTP 403 on WebFetch and businesswire.com also 403'd; confirmed instead via a reputable trade outlet (Mass Market Retailers) that reproduces the full set of | US Black Friday 2025 YoY (Mastercard SpendingPulse, ex-autos): overall retail +4.1%; e-commerce +10.4%; in-store +1.7%; apparel +5.7% (online +6.1%, in-store +5.4%); jewelry +2.75%; restaurants +4.5% | Mastercard SpendingPulse (via Mass Market Retailers / Mastercard press release) 2025-11-29 | Confirmed Đã xác nhận | ↗ |
| Visa Consulting & Analytics' Retail Spend Monitor found U.S. holiday retail spending (a 7-week period starting Nov 1, 2025, all payment types, not adjusted for inflation) up 4.2% year over year, while Mastercard SpendingPulde measured U.S. spending between Nov 1 and Dec 21, 2025 up 3.9% year over year. Visa figure and methodology (4.2%, not inflation-adjusted, 7-week period from Nov 1, in-store 73%/e-commerce 27% of volume, e-commerce +7.8%) confirmed directly from Visa's own newsroom press release. Mastercard's 3.9% figure and Nov 1-Dec 21 date range confir | US holiday 2025 (all channels) YoY: Visa +4.2% (not inflation-adjusted, 7-week period from Nov 1); Mastercard SpendingPulse +3.9% for Nov 1-Dec 21 | Visa Consulting & Analytics; Mastercard SpendingPulse (via Retail Dive) 2025-12-23 | Confirmed Đã xác nhận | ↗ |
| Shopify merchants generated a record $14.6 billion in global sales over the BFCM 2025 weekend, up 27% year over year (24% on a constant-currency basis). Confirmed verbatim from Shopify's own newsroom recap page. | $14.6B global Shopify merchant BFCM weekend 2025 sales, +27% YoY nominal, +24% YoY constant currency | Shopify 2025-12-02 | Confirmed Đã xác nhận | ↗ |
| Shopify President Harley Finkelstein's year-by-year recap and Shopify's own filings put global BFCM-weekend merchant sales at $6.3B (2021), $7.5B (2022), $9.3B (2023), $11.5B (2024, up 24% from 2023 per Shopify's SEC 6-K), and $14.6B (2025, up 27% per Shopify's investor press release). No single fetched document lists all five years together, so this was built from three fetched primary sources this session: (1) a Harley Finkelstein LinkedIn post (fetched) verbatim reading 'Total global Black Friday Cyber Monday sales by Shopify merchants th | Shopify merchant global BFCM weekend sales by year: 2021 $6.3B; 2022 $7.5B; 2023 $9.3B; 2024 $11.5B; 2025 $14.6B | Shopify / Harley Finkelstein (Shopify President) 2025-12-02 | Confirmed Đã xác nhận | ↗ |
| Shopify merchants generated $6.2 billion in global sales on Black Friday 2025 alone, up 25% year over year (22% on a constant-currency basis), with an average cart price of $117.93, according to Shopify President Harley Finkelstein. This is the Black-Friday-single-day figure, distinct from the BFCM-weekend figures in A-S20/A-S21 ($14.6B) and from the weekend average cart price of $114.70 in A-S23 -- the periods differ (single day vs 4-day weekend), which explains why the two average-cart | Shopify global Black Friday (single day) 2025: $6.2B, +25% YoY (+22% constant currency); average cart $117.93 | Shopify / Harley Finkelstein (Shopify President), via Retail Technology Innovation Hub 2025-11-29 | Confirmed Đã xác nhận | ↗ |
| Over BFCM weekend 2025, Shopify reported 81+ million shoppers worldwide, 15,800+ merchants making their first-ever sale, 94,900+ merchants having their best-selling day ever, an average cart of $114.70 ($112.29 constant currency), a sales peak of $5.1 million per minute at 12:01pm EST on Black Friday, cross-border orders at 16% of the global total, Shop Pay sales up 39% YoY and used for 32% of orders, and 136M+ packages tracked via the Shop app. Every figure in the claim matched verbatim across Shopify's own newsroom recap page and its investor press release (both fetched). These are weekend-wide (not Black-Friday-only) figures; the live-tracker point-in-time readings noted in the conflicts field ($3. | Shopify BFCM 2025 weekend (global): 81M+ shoppers; 15,800+ first sales; 94,900+ merchants best day ever; avg cart $114.70 ($112.29 cc); peak $5.1M sales/min at 12:01 PM EST on Black Friday; cross-border 16% of orders; Shop Pay sales +39% YoY, 32% of orders; 136M+ packages tracked | Shopify 2025-12-02 | Confirmed Đã xác nhận | ↗ |
| Klaviyo reported that brands on its platform sent 22.7 billion messages over BFCM 2025 (up 25% YoY), generating more than $3.8 billion in Klaviyo Attributed Value (up 27% YoY), with a 71% YoY surge in revenue from AI product recommendation-powered messages, as nearly 20,000 customers hit their best sales day ever on a platform serving over 183,000 brands. Every number in the claim matches Klaviyo's own BFCM 2025 press release (also mirrored at klaviyo.com/newsroom/bfcm-holiday-shopping and Businesswire/Nasdaq). The +71% figure is specifically 'revenue lift from AI recommendation-powered messages YoY', not a lif | 22.7B messages (up 25% YoY); $3.8B Klaviyo Attributed Value (up 27% YoY); +71% YoY revenue from AI product recommendation-powered messages; nearly 20,000 customers hit their best sales day ever; platform serves over 183,000 brands | Klaviyo (investor press release) 2025-12-02 | Confirmed Đã xác nhận | ↗ |
| Stripe reported that businesses on its platform processed more than $40 billion in payment volume across more than 578 million transactions from Black Friday through Cyber Monday 2025, including more than 152,000 peak transactions per minute, a record-breaking Cyber Monday with more than $10 billion in volume, cross-border volume growing 37% YoY to more than $4.4 billion, and more than 150,000 businesses (including Shopify, Wix, Lovable and beehiiv) having their best day ever. All six figures in the claim match Stripe's own newsroom post verbatim. This is Stripe's own processed-volume figure (payments run through Stripe), not total US/global ecommerce sales — worth keeping that distinction in the report text. | >$40B business volume; >578M transactions; >152,000 peak transactions/minute; >$10B Cyber Monday payment volume (record); cross-border volume grew from $3.2B to >$4.4B (+37% YoY); >150,000 users had their best day ever | Stripe (newsroom) 2025-12-01 | Confirmed Đã xác nhận | ↗ |
| Sensor Tower's analysis of more than 100,000 real Amazon shopping sessions found that on Black Friday 2025, Rufus-assisted sessions made up about 40% of total sessions but drove about 66% of purchases, converting at 3.5 times the rate of non-Rufus sessions — a gap that held steady across October, November and December — with nearly all of Amazon's incremental Black Friday growth attributable to Rufus-assisted shopping. Confirmed via Sensor Tower's own blog (which itself states the 40%-of-sessions figure) corroborated word-for-word by The Drum's report on the same Sensor Tower analysis, which supplies the full 100,000+/66%/3.5x/'nearly all incremental growth' figures not spel | Sensor Tower analyzed 100,000+ Amazon shopping sessions; on Black Friday 2025, Rufus-assisted sessions were ~40% of total sessions but drove ~66% of purchases; Rufus sessions converted at 3.5x the rate of non-Rufus sessions (steady Oct-Dec); nearly all incremental Black Friday growth on Amazon came from AI-assisted (Rufus) shopping | Sensor Tower 2025-12 | Confirmed Đã xác nhận | ↗ |
| Adobe Analytics found that shoppers referred by generative AI tools converted 31% more than visitors from other traffic sources over the 2025 holiday season (versus 38% on Black Friday specifically), and 40% better on Prime Day June 2026; based on this trend, Practical Ecommerce's Armando Roggio forecasts AI-referred shoppers will convert at least 25% better than non-AI traffic during the 2026 holiday season. All four numbers (31%, 38%, 40%, >=25% forecast) confirmed. The 31%/38% figures were independently confirmed via Digital Commerce 360's direct reporting of Adobe's data (published 2026-01-13, quoting Adobe verbatim: 'they were converting 31% more... On Black F | Adobe Analytics: generative-AI-referred shoppers converted 31% better than other traffic over the full 2025 holiday season, 38% better on Black Friday 2025, and (per Adobe's Prime Day report) 40% better on Prime Day June 2026; Practical Ecommerce's Armando Roggio forecasts the premium will be at least 25% for holiday 2026 | Adobe Analytics, via Practical Ecommerce (Armando Roggio) and Digital Commerce 360 2026-08-16 | Confirmed Đã xác nhận | ↗ |
| Bain & Company's 2026 Holiday Shopping Outlook forecasts US holiday sales will exceed $1 trillion for the first time (reaching $1,016 billion, up 4.5% YoY), with headline inflation of 3.4% accounting for more than half of that growth, in-store sales growing 2.5% and online sales growing 9% — with online contributing roughly 60% of total dollar growth despite being about a third of total sales. Confirmed directly on Bain's own infographic page: $1,016B/+4.5%, in-store +2.5% and nonstore +9%, headline inflation 3.4%. The 'online = 60% of total growth' figure is not stated as a single number by Bain but is arithmetically consistent with their channel s | US Nov-Dec 2026 holiday retail sales forecast: $1,016B (>$1 trillion for the first time), +4.5% YoY; headline inflation 3.4% (core 2.5%), accounting for more than half of the projected growth; in-store +2.5% (~$695B, ~68% of total); online/nonstore +9% (~$321B, ~32% of total, contributing roughly 60% of total dollar growth) | Bain & Company 2026 | Confirmed Đã xác nhận | ↗ |
| Bain & Company found that 24% of online shoppers expect to begin their 2026 holiday shopping journey on AI platforms such as ChatGPT, Google Gemini and Claude (up from 17% in 2025), with an additional 13% of online shoppers planning to use a retailer's own AI shopping agent, and that more than 90% of shoppers plan to shop on Black Friday, Cyber Monday, or another major sales event. All three figures confirmed verbatim on Bain's own page, including Bain's own 'an additional 13%' framing — which resolves conflict li#57 in the claim's favor (13% genuinely is presented by Bain as incremental/additional, not a rephrasing error) and shows li#3 | Bain 2026 Holiday Shopping Outlook: 24% of online shoppers plan to start their holiday shopping journey on generative-AI platforms (ChatGPT, Gemini, Claude), up from 17% in 2025; an ADDITIONAL 13% of online shoppers say they plan to use a retailer's own AI shopping agent (e.g. Walmart Sparky, Amazon Alexa); more than 90% of shoppers plan to shop Black Friday, Cyber Monday, or another major sales event | Bain & Company 2026 | Confirmed Đã xác nhận | ↗ |
| "The 2026 holiday season is expected to see similar growth rates to 2025, approximately 2.6% for overall US retail and 6.6% for ecommerce, according to EMARKETER's February forecast." Confirmed via direct WebFetch. The 6.6% figure and the framing of the 2026 season as growing at 'similar' rates to 2025 (i.e. steady, not accelerating) both check out against this EMARKETER article. The corpus's own 'steady, not accelerating' phrase is a fair | ~6.6% | EMARKETER, "It's never too early: What's ahead for the 2026 holiday season" 2026-02-23 | Confirmed Đã xác nhận | ↗ |
| PwC's 12th annual Holiday Outlook (US survey, n=4,093, June 15-26, 2026) found holiday gift spend intentions down about 2% to $708 per person, Millennials cutting gift budgets 10% and travel budgets 37% as they juggle mortgages/childcare, 29% of consumers (up from 22% a year earlier) planning to use AI to research products/compare prices/stay on budget, 64% of all consumers (78% of Gen Z) prioritizing screen-free gifts, and 60% of consumers planning to buy holiday gifts in physical stores. PwC's own page returned HTTP 403 on direct fetch each time; figures corroborated via secondary reporting that quotes the PwC report directly (search snippets attributing exact PwC wording and numbers, cross-checked against multiple outlets). Confirmed: -2% gif | PwC Holiday Outlook 2026 (US consumers, n=4,093, fielded June 15-26, 2026): average planned gift spend down ~2% to $708 (from $721); Millennials plan -10% on gifts and -37% on travel YoY; 29% plan to use AI to research/compare/budget, up from 22% last year (+7 points); 64% of all consumers (78% of Gen Z) prioritizing screen-free gifts; 60% plan to buy in-store. | PwC Holiday Outlook 2026 2026 (survey fielded June 15-26, 2026) | Confirmed Đã xác nhận | ↗ |
| Basis Technologies' 2026 Winter Holiday Shopping Trends study (Basis ImpactIQ + GWI survey, nationally representative, n=2,006 US consumers age 16+, fielded May 2026) found 54% of shoppers say Black Friday and Cyber Monday are not critical to shop because deals are available all season (though 56% still plan to participate), and 71% of consumers used at least one digital tool — search, social, or AI — to find gift ideas in 2025. Both figures confirmed directly from basis.com pages. Sample size (2,006) and methodology match the claim exactly. | 54% of US consumers say Black Friday/Cyber Monday are no longer critical to their shopping journey (because deals run all season) even though 56% still plan to participate; 71% used at least one digital tool (search, social, or AI) to find gift ideas in 2025. | Basis Technologies (Basis ImpactIQ + GWI survey) 2026 (survey fielded May 2026) | Confirmed Đã xác nhận | ↗ |
| Narvar's 2026 Holiday Shopping Report (1,348 US consumers, 100 senior retail decision-makers) found 65% of shoppers plan to use AI for at least one part of their holiday shopping — 43% for gift discovery, 33% to compare products or summarize reviews, and 29% for budgeting — while only 8% of retailers call themselves 'very confident' in using AI to improve the shopping experience. All five figures (65/43/33/29/8) and both sample sizes confirmed via a combination of the Narvar landing page itself and CPA Practice Advisor's detailed write-up, which quotes Narvar directly for the 43/33/29 breakdown that Narvar's own page omitted. | Narvar 2026 Holiday Shopping Report (1,348 US consumers + 100 senior retail decision-makers): 65% of shoppers plan to use AI for at least one part of holiday shopping; of AI users, 43% for gift discovery, 33% for comparing products/summarizing reviews, 29% for budgeting/planning spend; only 8% of retailers describe themselves as very confident using AI to improve the shopping experience. | Narvar 2026-08-24 | Confirmed Đã xác nhận | ↗ |
| "20% of all 2026 holiday ecommerce traffic will originate from AI chat agents" (consumer-facing bots, autonomous backend agents, and competitor price-matching systems); "1 in 3 ecommerce sites will have a personal, site-specific shopper agent live by Cyber Week 2026"; "59% higher holiday sales growth in 2025: +6.2% for brands with shopper agents versus +3.9% for brands without" Found the actual primary Salesforce source this time (first pass only found secondhand citations via Thyris.AI/LinkedIn). All three sub-claims in the corpus match this post closely, including the specific 59%/+6.2%/+3.9% figures. Correct the attribution in the | 20% of 2026 holiday ecommerce traffic from AI chat agents; 1 in 3 ecommerce sites with own shopper agent by Cyber Week 2026; brands with shopper agents saw +6.2% holiday sales growth in 2025 vs +3.9% without (59% higher) | Salesforce Blog, "2026 Holiday Predictions: A Unified Yet Fractured Journey" (Caila Schwartz, Director of Consumer Strategy & Insights) 2026-07-20 | Confirmed Đã xác nhận | ↗ |
| PYMNTS Intelligence's Global Digital Shopping Index: Merchant Edition (commissioned by Visa Acceptance Solutions, published Sept 2, 2026) found 23% of merchants can clearly identify both AI-driven traffic and purchases, 21% can recognize agentic traffic but not connect it to completed sales, and 61% agree AI-generated results will influence purchases more than traditional search — with large merchants (27%) still ahead of SMBs (19%) on attribution capability. All three headline figures (23%/21%/61%) confirmed exactly, plus the large-vs-SMB breakdown as a bonus. Report name and exact commissioning sponsor identified: this is not a plain PYMNTS in-house survey but a Visa Acceptance Solutions-commissioned index that P | 23% of merchants can clearly identify both AI-driven traffic and the purchases it generates; 21% can recognize agentic/AI traffic but cannot connect it to completed purchases; 61% agree AI-generated results will influence purchases more than traditional search. (Large merchants: 27% can identify agent-driven purchases vs. 19% of SMBs.) | PYMNTS Intelligence (Global Digital Shopping Index: Merchant Edition, commissioned by Visa Acceptance Solutions) 2026-09-02 | Confirmed Đã xác nhận | ↗ |
| Radial's annual Peak Season Consumer Survey (with Dynata, fielded August 2026, ~1,000 US adults 18+) found 70% of consumers would forgo a 5% discount for a guaranteed delivery date (with 54% giving up a 10% discount and 37% a 20% discount for the same guarantee), up from 66% in Radial's 2025 survey of the same question. 70% and the 'up from 2025' framing both confirmed; the 2025 comparison point (66%) confirmed via Radial's own 2025 press release title ('Radial Uncovers 66% of Holiday Shoppers Would Give Up Discounts for Guaranteed Delivery', businesswire, 2025-09-08), so 'up | 70% of holiday shoppers would give up a discount (specifically, forgo a 5% discount) for a guaranteed delivery date, up from 66% in Radial's 2025 survey. | Radial (with Dynata) 2026 | Confirmed Đã xác nhận | ↗ |
| AdImpact's June 2026 Political Projections Report update forecasts $11.6 billion in total 2026 election-cycle ad spending (comfortably above the loosely-stated '>$10B'), and there are exactly 24 days between Election Day (Nov 3, 2026) and Black Friday (Nov 27, 2026), confirmed by direct calendar calculation. The '>$10B' framing understates AdImpact's actual, more precise updated figure of $11.6B — still directionally correct and not misleading, so counted as confirmed rather than corrected. The 24-day gap was verified independently via a date calculation (Nov 1, 2 | AdImpact's updated Political Projections Report (June 11, 2026) forecasts $11.6B in total political ad spending for the 2026 election cycle — comfortably above the '>$10B' figure cited. Independently verified by calendar: Election Day 2026 is Tuesday, Nov 3; Black Friday 2026 is Friday, Nov 27 — exactly 24 days apart. | AdImpact (Political Projections Report); calendar calculation 2026-06-11 | Confirmed Đã xác nhận | ↗ |
| NRF: 'retail sales in 2026 ... will grow 4.4% year over year to $5.6 trillion' (published March 19, 2026). Deloitte's 2026 Retail Industry Global Outlook (330 retail executives, 86% at companies with $1B+ revenue) found 'nine in 10 expect AI to be increasingly used over search engines by 2026' and 'nearly seven in 10 retail executives surveyed agree that behaviors such as trading down, shopping value channels, or swapping convenience for savings represent a structural change, not a temporary response to inflation.' McKinsey & Company, cited alongside ICSC, projects 'agentic commerce in the U.S. business-to-consumer retail market may reach $1 trillion in revenue by 2030.' All four sub-claims confirmed against primary or near-primary sources. Deloitte's exact wording is 'AI to be increasingly used over search engines,' which the compiled claim renders as AI tools 'overtaking' search — a slight strengthening but consistent in mea | NRF forecasts 2026 US retail sales growth of 4.4% YoY to $5.6 trillion (published March 19, 2026). Deloitte's 2026 Retail Industry Global Outlook (330 executives surveyed, 86% at $1B+ revenue retailers): nine in 10 retail executives expect AI to be increasingly used over search engines by 2026; nearly seven in 10 say value-seeking behaviors (trading down, value channels, swapping convenience for savings) are a structural change, not a temporary inflation response. McKinsey & Company (with ICSC): agentic commerce could reach $1 trillion in US B2C retail revenue by 2030. | NRF; Deloitte (2026 Retail Industry Global Outlook); McKinsey & Company (with ICSC) 2026-03-19 (NRF); 2026 (Deloitte outlook); 2026-05-05 (McKinsey/ICSC figure reported) | Confirmed Đã xác nhận | ↗ |
| McKinsey's ConsumerWise survey of more than 4,000 US consumers, fielded July 29-August 5, 2026, found 45% of respondents planned to have started holiday shopping by the end of October, with November-before-Black-Friday (24%) the single most common starting window and Black Friday weekend itself picked by only 12%. Confirmed the full breakdown via a full-text mirror of the McKinsey article: already-started 10%, Aug 5%, Sep 11%, Oct 19% (=45% cumulative by end of October), Nov-before-BF 24%, BF weekend 12%. Could not independently confirm the precise n=4,013 cited in the | 45% of US consumers plan to start holiday shopping by end of October 2026 (10% already started + 5% August + 11% September + 19% October); 24% plan to start in November before Black Friday; 12% plan to start on Black Friday weekend. | McKinsey & Company, ConsumerWise Global Sentiment Survey 2026-08-28 (survey fielded 2026-07-29 to 2026-08-05) | Confirmed Đã xác nhận | ↗ |
| Cordial's Holiday 2026 Consumer Research report, based on a survey of 1,000 US consumers, found that Black Friday weekend accounts for 19% of shoppers' planned start dates, while 53% say promotions starting in October or early November make them more likely to engage with that brand. Both figures (19%, 53%) confirmed directly from Cordial's own report page fetched this session. Note the 53% wording is slightly more specific than the corpus paraphrase ('engage more with brands that spread promotions over a longer window'): Cordial's own fra | 19% of 2026 holiday shoppers say they'll start shopping on Black Friday weekend; 53% say holiday promotions starting in October or early November make them more likely to engage with a brand's messages. | Cordial, Holiday 2026 Consumer Research Data 2026 (exact publish/fielding date not stated on the page) | Confirmed Đã xác nhận | ↗ |
| Basis, in partnership with GWI, surveyed 2,006 US consumers aged 16+ (nationally representative, fielded May 2026) for its 2026 Holiday Shopping and Advertising Trends Report and found that '56% plan to participate in Black Friday and Cyber Monday — but 54% say it is not critical to shop those events because deals are available throughout the season.' Confirmed exact wording, sample size, age range, geography and fielding date from the GlobeNewswire press release (primary distribution of Basis's own release) fetched this session, cross-checked against basis.com's own report landing page which independently | 56% of US consumers plan to participate in Black Friday and Cyber Monday, but 54% say it isn't critical to shop those events specifically because deals are available throughout the season. | Basis Technologies (Basis ImpactIQ) and GWI, 2026 Holiday Shopping and Advertising Trends Report 2026-07-22 (survey fielded May 2026) | Confirmed Đã xác nhận | ↗ |
| Adobe Analytics/Adobe Digital Insights reported that US consumers spent $257.8 billion online during the 2025 holiday season (Nov 1-Dec 31), up 6.8% year-over-year, the first time a consecutive two-month online spending period exceeded a quarter-trillion dollars, and above Adobe's own pre-season forecast of $253.4 billion. Confirmed directly from Adobe's own January 7, 2026 press release, fetched this session. The $253.4B forecast comparison was corroborated by Forbes/Talk Business coverage of the same release. | US online holiday spending Nov 1-Dec 31, 2025 reached $257.8 billion, up 6.8% YoY, beating Adobe's own October forecast of $253.4 billion (+5.3%). | Adobe Digital Insights (Adobe Analytics, based on 1 trillion+ US retail site visits) 2026-01-07 | Confirmed Đã xác nhận | ↗ |
| EMARKETER's February 2026 holiday forecast projects 2026 US holiday ecommerce growth of 6.6%, against roughly 2.6% growth for overall US retail — broadly similar to 2025's growth rates rather than an acceleration. Confirmed via EMARKETER's own article. This is a forecast/model estimate, not a survey — treat 'growth' as a projection subject to revision, and note EMARKETER has also published a separate, higher Black Friday 2025 actual ecommerce growth figure (10.4%) for a | EMARKETER projects US holiday ecommerce/online sales growth of 6.6% for 2026 (vs ~2.6% for overall US retail). | EMARKETER 2026-02-23 | Confirmed Đã xác nhận | ↗ |
| Adobe Analytics reported Black Friday 2025 US online spending of $11.8 billion (up 9.1% YoY), Cyber Monday 2025 spending of $14.25 billion (up 7.1% YoY, a new single-day record), and total Cyber Week (5-day, Thanksgiving through Cyber Monday) online spending of $44.2 billion (up 7.7% YoY). Confirmed all three figures directly from Adobe's own Cyber Monday 2025 recap press release, fetched this session. Corpus's own claim value states Cyber 5 as '>$44B ($44.2B)' which matches exactly. The corpus conflict note about x#20 quoting a pre-event CM pro | Black Friday 2025 US online spending: $11.8 billion (+9.1% YoY). Cyber Monday 2025: $14.25 billion (+7.1% YoY). Cyber Week/Cyber 5 (Thanksgiving-Cyber Monday) total: $44.2 billion (+7.7% YoY). | Adobe Digital Insights (Adobe Analytics) 2025-12-01/02 (Cyber Monday recap release) | Confirmed Đã xác nhận | ↗ |
| Adobe Analytics reported Black Friday 2024 US online spending of $10.8 billion (up 10.2% YoY) and Cyber Monday 2024 spending of $13.3 billion (up 7.3% YoY); Cyber Week 2024 (5-day) totaled $41.1 billion (up 8.2% YoY). Confirmed directly from Adobe's own December 2024 press release, fetched this session — this stat is no longer 'unattributed,' it is Adobe's own reporting (correction to the corpus's attribution field only, the figures themselves were already accurate). Implie | Black Friday 2024 US online spending: $10.8 billion (+10.2% YoY). Cyber Monday 2024: $13.3 billion (+7.3% YoY). | Adobe Digital Insights (Adobe Analytics) 2024-12-03 | Confirmed Đã xác nhận | ↗ |
| Adobe Digital Insights reported that traffic to US retail sites from generative AI sources rose 693.4% for the full Nov 1-Dec 31, 2025 holiday season compared with the year prior, with single-day spikes of +805% YoY on Black Friday and +670% YoY on Cyber Monday. All three figures confirmed via Adobe's own press materials. Note the corpus's own conflict entry is correct to flag that these are three different measurement windows from the same underlying Adobe dataset (full season vs Black Friday day vs Cyber Monday day) | Generative-AI-referred traffic to US retail sites: +693.4% for the full 2025 holiday season (Nov 1-Dec 31); +670% on Cyber Monday 2025; +805% on Black Friday 2025 (all YoY). | Adobe Digital Insights (Adobe Analytics) 2026-01-07 (season total); Black Friday and Cyber Monday figures from Adobe's respective Nov 29 and Dec 1-2, 2025 recap releases | Confirmed Đã xác nhận | ↗ |
| Per Salesforce data reported by Retail Dive, US Black Friday 2025 online order volume fell 1% year-over-year even as average selling price rose 7% and units per transaction fell 2%; separately, Salesforce said AI and agents drove $14.2 billion in global online sales and $3 billion in US online sales on Black Friday. Confirmed via Retail Dive's direct, quoted attribution to Salesforce (including a named Salesforce spokesperson, Caila Schwartz). Could NOT find a Salesforce-owned page (press release or newsroom story) that itself prints this exact Black-Friday-only breakdown | Black Friday 2025 (US): online order volume -1% YoY, average selling price +7% YoY, units per transaction -2% YoY; AI & agents drove $14.2B in global online sales and $3B in US online sales on Black Friday. | Salesforce (Salesforce Shopping Index / consumer insights data, via Retail Dive and Salesforce's own Cyber Week newsroom coverage) 2025-11-29/12-01 | Confirmed Đã xác nhận | ↗ |
| Adobe Digital Insights reports that for the full 2025 US holiday season (Nov 1-Dec 31, 2025), mobile devices drove 56.4% of online sales, consumers spent $20.0 billion via Buy Now, Pay Later (up 9.8% YoY), and Cyber Monday alone saw $1.03 billion in BNPL spend, up 4.2% YoY. Value matches claim closely (corpus said '+~10%' vs actual +9.8%, and 56.4% vs sometimes-quoted mid-season 52.8%/56.1% interim figures -- 56.4% is the final full-season number). Confirmed via business.adobe.com/resources/holiday-shopping-report.html, corrobora | Mobile 56.4% of online sales; BNPL $20.0B (+9.8% YoY); Cyber Monday BNPL $1.03B (+4.2% YoY, all-time high) | Adobe Digital Insights (Adobe Analytics) 2026-01 (full-season final report; Cyber Monday figures released 2025-12-01 via news.adobe.com/news/2025/12/adobe-cyber-monday-hits-record) | Confirmed Đã xác nhận | ↗ |
| Reprice's 'The Black Friday Freeze Report' (published 2026-08-14), a survey of 180 European e-commerce pricing leaders whose companies run Black Friday campaigns, found 30% changed no prices at all during Black Week, a further 37% changed prices once or twice (67% combined at most twice), and 74% said they hit a real barrier when trying to react to competitors mid-week. All three figures confirmed exactly by fetching the primary report page. Vendor sells repricing software, so has an incentive to frame manual/slow pricing as risky -- flag as vendor-sponsored survey when quoting, per corpus's own conflict note. | 30% changed no prices; 67% changed prices at most twice (30% zero + 37% once-or-twice); 74% reported a real barrier to reacting mid-week | Reprice ('The Black Friday Freeze Report 2026') 2026-08-14 | Confirmed Đã xác nhận | ↗ |
| Shopify's official BFCM 2025 press release reports merchants achieved a record $14.6 billion in global sales over the Black Friday-Cyber Monday weekend, up 27% year-over-year (24% on a constant-currency basis). Corpus's own conflict note is correct and worth keeping: this is the full BFCM WEEKEND total, not 'Black Friday' alone (Black Friday alone was $6.2B per the same release cycle, per Retail Technology Innovation Hub's report of Shopify data). Caption accordingly | $14.6 billion in global merchant sales over BFCM 2025, +27% YoY (+24% on a constant-currency basis) | Shopify (official investor press release) 2025-12-01 | Confirmed Đã xác nhận | ↗ |
| The CNBC/NRF Retail Monitor, using actual credit/debit card transaction data compiled by Affinity Solutions and released by the National Retail Federation on 2026-01-12, confirmed that US holiday retail sales from Nov 1 to Dec 31, 2025 grew 4.1% year-over-year and exceeded $1 trillion for the first time. NRF's pre-season forecast (Oct 2025) had projected holiday sales of $1.01-1.02 trillion (+3.7% to +4.2% YoY); the actual measured result released Jan 2026 confirmed the season did cross $1T, at +4.1% YoY -- squarely inside the forecast range. So this is a real | >$1 trillion in US holiday retail sales (Nov 1-Dec 31, 2025), +4.1% YoY -- measured, not just forecast | National Retail Federation (CNBC/NRF Retail Monitor) 2026-01-12 | Confirmed Đã xác nhận | ↗ |
| Adobe Digital Insights reports UK shoppers spent £26.9 billion online during the 2025 holiday season (1 November-31 December), up 4.1% year-over-year and the highest online spending Christmas period on record. Adobe's £26.9bn/+4.1% figure was originally published as a pre-season FORECAST, but a post-season recap (Retail Technology Innovation Hub, citing Adobe Digital Insights, dated 2026-01-14) confirms the actual final result landed at the same £26.9bn/+4.1% -- i.e | £26.9 billion UK online spend, Nov 1-Dec 31 2025, +4.1% YoY | Adobe Digital Insights 2026-01 (final result; forecast of the same figure was published pre-season) | Confirmed Đã xác nhận | ↗ |
| Adobe Analytics' final 2025 holiday season data shows overall e-commerce returns down 1.2% year-over-year, with returns during the Dec 26-31 post-Christmas window up 4.7% YoY, and roughly one in seven of the season's returns occurring in that six-day window, versus about one in eight in 2024. The -1.2% overall figure is the FINAL full-season number; an earlier, widely-reported mid-season Adobe figure (Dec 17, 2025) said returns were down 2.5% for just Nov 1-Dec 12 -- that number describes a narrower window and should not be conflated with the final | Overall holiday-season returns -1.2% YoY; returns during Dec 26-31 +4.7% YoY; ~1 in 7 of the season's returns fell in the Dec 26-31 window (vs ~1 in 8 in 2024) | Adobe Analytics / Adobe Digital Insights 2026-01-07 | Confirmed Đã xác nhận | ↗ |
| Ryder System's 12th annual e-commerce consumer study ('Beyond Price: What's Driving Consumer Shopping Behavior'), a survey of 1,160 US online shoppers conducted in March 2026, found 79% of shoppers who had a negative returns experience stopped shopping with that brand, and 64% of shoppers now use AI-assisted shopping tools. Both figures confirmed exactly via the primary Ryder newsroom press release. Corpus's conflict note about AI-adoption figures varying by survey (Cordial 42%, unattributed 71%) is valid context -- different surveys, different populations/question wording, not a | 79% of shoppers with a bad returns experience stopped shopping with that brand; 64% of shoppers now use AI-assisted shopping tools | Ryder System, Inc. (2026 E-commerce Consumer Study) 2026 (survey fielded March 2026; press release dated 2026) | Confirmed Đã xác nhận | ↗ |
| TikTok Shop's US Fulfillment Policy, effective September 3, 2026, requires sellers to keep Valid Tracking Rate ≥95%, On-Time Delivery Rate ≥80%, Late Dispatch Rate at or below a 4% recommended level (with enforcement kicking in above 10%), and Seller-Fault Cancellation Rate ≤2.5%, with regular orders dispatched within 2 business days, auto-cancelled if untracked after 5 business days, and delivered within 6 business days. All five thresholds and the three SLA windows match the claim exactly, fetched directly from TikTok's own policy page (effective date 09/03/2026 stated on the page itself). The corpus's own 'conflicts' note is worth keeping in the report: TikTok's separate Cam | VTR ≥95%; OTDR ≥80%; LDR ≤4% recommended (enforcement triggers >10%); SFCR ≤2.5%; regular-order SLA: dispatch within 2 business days of Awaiting Shipment, auto-cancel if no tracking within 5 business days, deliver-by within 6 business days. Policy effective 09/03/2026. | TikTok Shop Seller Center — US Fulfillment Policy 2026-09-03 | Confirmed Đã xác nhận | ↗ |
| Amazon's own Seller Central announcement states that, starting September 30, 2026, US seller-fulfilled offers to Amazon Business customers must maintain a 90%-or-higher Business Hour Delivery Rate over a rolling 14-day period; sellers under the threshold on September 30 get a notification, and offers may be deactivated for Business customers if the rate is still below 90% by October 30, 2026 — FBA and retail-offer eligibility are not affected. Fetched Amazon's own Seller Central announcement directly and it matches the claim's every detail (90% threshold, rolling 14-day window, Sep 30 notice date, Oct 30 enforcement deadline, FBA exempt). Secondary trade coverage (myamazonguy.com) independently corr | Amazon Business Hour Delivery Rate (BHDR) ≥90%, measured on a rolling 14-day window, for US seller-fulfilled shipments to Amazon Business customers, effective September 30, 2026; sellers below threshold get a notification on Sep 30 and may have FBM offers deactivated for Business customers if still non-compliant by October 30, 2026; FBA and retail-offer eligibility are unaffected. | Amazon Seller Central — official forum announcement ('New business hour delivery requirement for seller-fulfilled shipments') 2026 | Confirmed Đã xác nhận | ↗ |
| In both 2024 and 2025, Amazon's Black Friday Week / Cyber Monday event ran for approximately 12 days ending on Cyber Monday, consistent with the panel's 'starts 12 days before Cyber Monday' pattern; Amazon has not yet published official 2026 event dates. The underlying 12-day pattern checks out for the two most recent years the corpus can point to, but I could not find any independent source using the term 'Turkey 12' — that label appears to be the panel's/Porter's own shorthand, not an industry-standard term, | Amazon's combined Black Friday Week + Cyber Monday promotion has run ~12 days in each of the last two years: Nov 21–Dec 2, 2024 and Nov 20–Dec 1, 2025. Official 2026 dates have not yet been announced (as of this search). | Trade press coverage of Amazon's own announced event dates (PhoneArena, GSMArena) for 2024 and 2025 2025-11 | Confirmed Đã xác nhận | ↗ |
| By calendar computation, Black Friday 2026 falls on November 27 and Cyber Monday on November 30 — the first time Cyber Monday has landed in November since 2023, since it fell in December in both 2024 and 2025 — and Christmas Eve 2026 falls on a Thursday. Verified by direct computation (not just search): Black Friday dates were 2022-11-25, 2023-11-24, 2024-11-29, 2025-11-28, 2026-11-27; Cyber Monday dates were 2022-11-28, 2023-11-27, 2024-12-02, 2025-12-01, 2026-11-30. All facts in the claim check out exactly, | Black Friday 2026 = Friday, November 27, 2026. Cyber Monday 2026 = Monday, November 30, 2026 (falls in November). Christmas Eve 2026 = Thursday, December 24, 2026. Cyber Monday last fell in November in 2023 (Nov 27); it fell in December in both 2024 (Dec 2) and 2025 (Dec 1) — so 2026 is the first November Cyber Monday since 2023, i.e. a 3-year gap. | Direct calendar computation (Gregorian calendar, US Thanksgiving = 4th Thursday of November) 2026-09-14 | Confirmed Đã xác nhận | ↗ |
| Recomputing the worked example (€100 revenue, €60 contribution margin, €40 cost, discount taken directly off margin) confirms the claimed erosion exactly: a 15% discount cuts contribution margin by 25%, a 20% discount cuts it by 33%, and a 25% discount cuts it by roughly 42%. Recomputed in Python: 15%→CM=45 (-25.0%), 20%→CM=40 (-33.33%), 25%→CM=35 (-41.67%, rounds to ~-42%). All three figures match the claim exactly. | On €100 revenue with €60 contribution margin (€40 variable cost), a straight price discount taken dollar-for-dollar off margin produces: 15% off → CM falls to €45 (-25%); 20% off → CM falls to €40 (-33.3%); 25% off → CM falls to €35 (-41.7%, ≈-42%). | Direct recomputation (arithmetic, no external source needed per the task) 2026-09-14 | Confirmed Đã xác nhận | — |
| Recomputing the stock-order illustration confirms it: a flat 8-week run-rate forecast at 100 units/week totals 800 units, while weighting the 4 peak weeks at 3x volume plus the 4 normal weeks and then applying 20% growth totals 1,920 units (≈1,900) — exactly the (400+1,200)×1.2 arithmetic the source describes. Recomputed exactly: flat = 800; peak-weighted = (4×100×3 + 4×100) × 1.2 = (1200+400)×1.2 = 1920. Matches the claimed '800 vs ~1,900 units' and the stated arithmetic check. | Flat 8-week forecast at 100 units/week = 800 units. Weighting 4 peak weeks at 3x (400 normal-week units become 1,200) plus 4 normal weeks (400 units) = 1,600 units, then +20% growth = 1,920 units (~1,900). | Direct recomputation (arithmetic, illustrative example) 2026-09-14 | Confirmed Đã xác nhận | — |
| Sierra, the conversational-AI vendor Minted uses for customer support, states that Minted saw a 387% month-over-month increase in conversation volume handled by its AI agent during the Black Friday-Cyber Monday 2025 period, covering questions about customization, order modifications, shipping timelines, and holiday delivery deadlines. Fetched Sierra's own LinkedIn post, which states the 387% MoM figure verbatim and in the exact BFCM 2025 context claimed. Confirmed this is vendor-sponsored/self-reported case-study data with no independent audit — flag as such in the report. The claim's own c | Minted saw a 387% month-over-month increase in AI-agent-handled conversation volume during Black Friday–Cyber Monday 2025 (per Sierra, the AI-agent vendor). | Sierra (vendor) — LinkedIn post on Minted's BFCM 2025 performance 2025-12 | Confirmed Đã xác nhận | ↗ |
| Klaviyo delivered more than 22.7 billion messages during BFCM 2025 (up 25% year over year), generating more than $3.8 billion in Klaviyo Attributed Value for its customers (up 27% YoY), and nearly 20,000 customers had their best sales day ever. Matches the claim closely. li#82's 183K brands is a different, unrelated denominator (Klaviyo's total customer count, not the BFCM best-sales-day cohort) and is not a real conflict, as the task itself notes. | 22.7B messages (+25% YoY); $3.8B+ Klaviyo Attributed Value (+27% YoY); ~20,000 brands hit their best sales day ever | Klaviyo, "Klaviyo Breaks Records with First AI-powered BFCM" (press release) 2025-12-02 | Confirmed Đã xác nhận | ↗ |
| Klaviyo reports that during BFCM 2025, shoppers who received messages across multiple channels (e.g., email and SMS/RCS) placed 11% more orders, added 34% more items to cart, and viewed 71% more products than single-channel recipients. Confirmed as stated. As the task notes, this is a correlational comparison (multichannel-opted-in shoppers vs. single-channel), not a randomized baseline — likely reflects engagement/intent selection as well as channel effect. Should be reported as an observed | Multichannel (email + SMS/RCS) shoppers viewed 71% more products, added 34% more items to cart, and placed 11% more orders than single-channel shoppers | Klaviyo BFCM 2025 Recap / "How Cross-Channel Orchestration & Text Messages Defined BFCM 2025" 2025-12-02 | Confirmed Đã xác nhận | ↗ |
| Klaviyo's 2025 Global AI Shopping Index found that 56% of consumers surveyed plan to use AI shopping assistants during Black Friday/Cyber Monday 2025 (part of a broader finding that 81% had used AI for shopping/product research in the prior 3 months). This is a distinct survey (different question, sample, and vendor) from Omnisend's finding that 'about 1 in 4' shoppers use AI to find deals/gifts (C-S13's AI figure). The two are not in conflict — 'plan to use an AI assistant at all during BFCM' (56%, broad d | 56% of consumers plan to use AI shopping assistants during BFCM 2025 | Klaviyo 2025 Global AI Shopping Index 2025-08 | Confirmed Đã xác nhận | ↗ |
| Omnisend's BFCM 2024 data shows automated email flows made up only about 3% of all email sends in November but generated roughly 30% of all email-driven orders, with automations averaging a 45.43% open rate and 26.45% click-to-conversion rate. Confirmed via the exact figures across multiple independent quotations of the same Omnisend report. This is a share-of-orders-vs-share-of-sends comparison, not a conversion rate — it is not directly comparable to the agency dashboards' 'flows as % of attribute | Automated messages = ~3% of email sends but ~30% of all email orders in November 2024 (BFCM period); automated email open rate 45.43%, click-to-conversion rate 26.45% | Omnisend BFCM/holiday email, SMS & push statistics report 2024-2025 (BFCM 2024 period: Nov 24-Dec 2, 2024) | Confirmed Đã xác nhận | ↗ |
| "On Black Friday, email-driven orders peaked between 7 AM and 12 PM, with the 8 AM hour generating one-third more orders than the next closest hour." "For SMS, the noon hour stole the spotlight, though most purchases happened earlier in the morning." "Email shopping reached its peak during lunchtime hours (12 PM-3 PM)" [Cyber Monday]. "SMS orders dominated the late morning window (10 AM-12 PM)" [Cyber Monday]. Confirmed via direct WebFetch -- all four sub-claims match this article closely, including near-verbatim wording. Confirms the first pass's own flag: NO timezone is specified anywhere in the article for any of these peak-hour windows -- keep that caveat in the | BF email orders peak 7 AM-12 PM with 8 AM generating one-third more orders than the next closest hour; BF SMS clicks peak at noon but most purchases happen earlier in the morning; CM email orders peak 12 PM-3 PM; CM SMS orders peak 10 AM-12 PM | Omnisend Blog, "Early BFCM Sales: A New Norm in Ecommerce" 2024-12-09 | Confirmed Đã xác nhận | ↗ |
| Loop Returns reports that over Cyber Week 2025, its merchants processed 15.9 million orders totaling $1.94 billion in GMV from 12.8 million unique shoppers, representing 16% of everyone who bought from a Shopify store that weekend. Core figures confirmed exactly. The task's own conflict note is correct and independently confirmed: the returning-vs-new-buyer growth comparison in the same corpus (returning shoppers GMV +19.5% YoY vs. new +13.4% YoY) is explicitly 2024 Cyber Week data — Loo | Cyber Week 2025 (Loop merchants): 15.9M orders; $1.94B GMV; 12.8M unique shoppers; 16% of all Shopify BFCM shoppers that weekend | Loop Returns, "Black Friday/Cyber Monday by the numbers: Exclusive data insights from Loop" 2025 Cyber Week data (blog page last updated 2026-08-20) | Confirmed Đã xác nhận | ↗ |
| Attentive's 2026 BFCM Marketer Pulse surveyed 240 marketers who use the Attentive platform, fielded June 2-18, 2026, spanning company sizes from under $1M to over $100M in revenue across 12+ verticals — so '240' is the precise figure to quote, and the '200+' reference elsewhere in the panel is a loose rounding of the same underlying survey, not a conflicting one. Confirmed and internally consistent, but flag the sampling frame: because respondents are all Attentive's own customers, findings (e.g., '79% of improvers retained BFCM-acquired customers') reflect a self-selected, platform-using population and shouldn't be ge | 240 marketers surveyed, June 2-18, 2026 — confirmed on Attentive's own report page; '200+' (li#215) is a compatible, rounded-down reference to the same 240 figure, not a separate survey. | Attentive (self-reported vendor research, methodology disclosed on its own site) 2026-06 (fielded); published as part of Attentive's BFCM 2026 guide | Confirmed Đã xác nhận | ↗ |
| AppLovin's own published case study for DTC sleepwear brand Tumble states Black Friday spend scaled 6.5x versus the pre-BFCM daily average (baseline window Nov 11-18, 2025), new-customer day-zero ROAS rose 15% on Black Friday versus the pre-BFCM average, and 93.7% of day-zero Black Friday checkouts were from new customers. Figures match the claim exactly. Caveat inherent to the source: this is AppLovin's own case study of its own platform, 'new customer' is AppLovin's definition, and ROAS is platform-attributed, not independently verified via holdout. Fine to cite as AppLovin's | 6.5x Black Friday spend scaling vs pre-BFCM daily average; +15% new-customer day-zero ROAS; 93.7% of day-zero Black Friday checkouts from new customers | AppLovin (official case study) 2025/2026 (Black Friday 2025 data) | Confirmed Đã xác nhận | ↗ |
| AdImpact's 2025-2026 Political Projections Report projects 2026 midterm political ad spending at $10.8 billion (subsequently updated to $11.6 billion in a later AdImpact release), supporting the claim's 'more than $10 billion' framing. The 24-day gap between Election Day and Black Friday is a plain calendar fact for 2026 and checks out exactly. The $10B+ figure traces to AdImpact, not to Tatari itself -- Tatari (as a CTV buying platform) is almost certainly relaying AdImpact's number rather than publishing its own independent projection, so attribute the dollar figure to AdImpact when citing it, with | $10.8 billion projected 2026 midterm political ad spend (AdImpact, later revised up to $11.6 billion); 24 calendar days between Election Day (Tue Nov 3, 2026) and Black Friday (Fri Nov 27, 2026) | AdImpact (political ad spend projections); calendar arithmetic for the day-count 2026 | Confirmed Đã xác nhận | ↗ |
| Tatari's advertiser survey, reported by Modern Retail (sponsored) and Tatari's own TVREV-hosted recap, states verbatim: 'Nearly six in 10 advertisers told Tatari they will increase their Black Friday and Cyber Monday (BFCM) TV spend in 2026 compared to 2025,' with named respondents including Bylt, Tecovas, Manscaped, and Bearbottom Clothing. Wording matches exactly. Sample size is still not disclosed anywhere found, matching the corpus's own conflict note -- do not imply a sample size beyond what's stated. A separate Tatari page titled '65 Advertisers Share How They're Planning for Black Friday' m | Nearly 6 in 10 (60%) of surveyed retail/DTC advertisers say they will increase BFCM TV ad budgets in 2026 vs 2025 | Tatari survey, via Modern Retail 2026 | Confirmed Đã xác nhận | ↗ |
| Adobe Analytics reported that AI-referred traffic — visits from a shopper clicking a link inside a generative-AI chatbot/assistant answer — to U.S. retail sites rose 805% year-over-year on Black Friday 2025 (Nov 28) alone. 805% is the single-day Black Friday figure. It is genuinely a different number from the other figures in 'conflicts', not a misquote of one: Cyber Monday alone was +670% (confirmed via news.adobe.com/news/2025/12/adobe-cyber-monday-hits-record, same source als | +805% YoY (Black Friday 2025 single day, Nov 28, vs Black Friday 2024) | Adobe Analytics / Adobe Digital Insights (as reported by Fox Business, Digital Transactions, Afrotech citing Adobe) 2025-11-29 | Confirmed Đã xác nhận | ↗ |
| Adobe's final 2025 holiday recap (published Jan 7, 2026) reported that traffic to U.S. retail sites from generative-AI tools was up 693.4% year-over-year across the full Nov. 1 - Dec. 31, 2025 window. This is the FINAL, full-season number, confirmed by direct fetch of Adobe's own newsroom release. It is not in real conflict with li#58's 760%: that figure is Adobe's own season-to-date-as-of-Dec-1 running total (confirmed via news.adobe.com/news/2025/12/adobe | +693.4% YoY (full 2025 holiday season, Nov 1 - Dec 31, US) | Adobe Newsroom (Adobe Analytics holiday recap) 2026-01-07 | Confirmed Đã xác nhận | ↗ |
| Adobe Analytics/Adobe Digital Insights has repeatedly reported that shoppers arriving at U.S. retail sites from an AI source convert at a higher rate than non-AI traffic, with the lift growing over time: +38% on Black Friday 2025, +31% across the full 2025 holiday season, +42% in March 2026, +40% across the full Prime Day 2026 event (June 23-26), and +60% in Adobe's July-2026 data (released Aug 19, 2026) — its 11th straight month of AI traffic out-converting other channels. All five windows in the claim are independently confirmed and each belongs to a distinct reporting date, not a single re-quoted figure — this is a genuine trend, not a misquote. Revenue-per-visit 'X% higher than non-AI' conflicting figures partially resolve by | +38% Black Friday 2025 (confirmed); +31% full holiday 2025 (confirmed); +40% Prime Day 2026 (June 23-26, confirmed); +42% March 2026 (confirmed); +60% in the Aug 19, 2026 Adobe cut, which covers July 2026 data (confirmed) | Adobe Analytics / Adobe Digital Insights (via Fox Business, Digital Transactions, TechCrunch, Digital Commerce 360) 2026-08-19 | Confirmed Đã xác nhận | ↗ |
| Adobe Digital Insights' Q1 2026 AI traffic report found that in March 2026, shoppers arriving at U.S. retail sites from AI sources spent 48% longer on-site and viewed 13% more pages per visit than shoppers from non-AI channels. Confirmed exactly as claimed via direct fetch. The li#41 (Yotpo/Eli Weiss) '45% more time on-site, 33% lower bounce' figure is unsourced and close-but-not-identical — likely a paraphrase drifted from a different Adobe window: Adobe's Jan 2026 holiday recap sep | 48% longer time on site; 13% more pages per visit (March 2026, US retail, AI-referred vs non-AI) | Adobe Digital Insights (Q1 2026 AI Traffic Report, via TechCrunch) 2026-04-16 | Confirmed Đã xác nhận | ↗ |
| Adobe Digital Insights reported that AI-referred traffic to UK retail sites grew 329% year-over-year across the 2025 holiday season (Nov 1 - Dec 31), with Boxing Day alone up 685% versus 2024 — well below the equivalent US season figure of +693.4%, and Adobe itself noted UK AI-referral volumes 'remained small relative to other channels.' Confirmed. UK total holiday online spend was £26.9bn, up 4.1% YoY. The claim's framing (lower than the ~693-805% US figures) is correct and Adobe's own commentary explicitly flags the small base — good context to keep attached to this stat, since 329% growth f | +329% YoY (UK, full 2025 holiday season, Nov 1 - Dec 31); Boxing Day alone +685% | Adobe Digital Insights (UK holiday 2025 online spending report) 2026-01-20 | Confirmed Đã xác nhận | ↗ |
| Using its AI Content Visibility Checker, Adobe scored how much of a retail site's content large language models can actually read, finding category averages ranging from 47% (Furniture & Home, lowest) to 63% (Cosmetics, highest), with Electronics at 56%, Sporting Goods and Apparel at 51%, and Grocery at 48%. Confirmed exactly — the 47-63% range in the claim matches Adobe's category-level scores precisely (Furniture & Home 47% to Cosmetics 63%). By page TYPE (not category) scores run higher: homepages ~75%, category pages ~74%, product pages ~66% per a related sear | 47-63% machine-readable, by retail category (Furniture & Home 47% lowest, Cosmetics 63% highest) | Adobe (AI Content Visibility Checker data, via Digital Commerce 360) 2026-06-17 | Confirmed Đã xác nhận | ↗ |
| Salesforce reported that generative AI and agents influenced $14.2 billion of the roughly $79 billion spent online globally on Black Friday 2025, including $3 billion of the U.S. total; Salesforce defines 'influenced' as sales touched by AI-driven personalized recommendations or conversational service, not sales autonomously purchased by an agent. Confirmed the $14.2B/$3B figures and that $79B is Salesforce's GLOBAL ONLINE (not omnichannel) Black Friday spend figure — resolving one conflict. However, a second figure genuinely does mix scopes as the conflicts note suspected: Salesforce's own Dec 5, 2025 | $14.2B global / $3B US, AI-and-agents-influenced sales on Black Friday 2025 (out of $79B global / $18B US total online+omnichannel Black Friday spend) | Salesforce (Shopping Index / Cyber Week data via Analytics Insight, corroborated by Salesforce's Dec 5, 2025 Cyber Week press release) 2025-12-01 | Confirmed Đã xác nhận | ↗ |
| Shopify reported that in Q1 2026, AI-chatbot-referred sessions to its merchants' stores grew more than 8x year-over-year, AI-referred orders grew nearly 13x, AI-referred sessions converted at nearly 50% higher rates than organic search, and AI-referred orders carried 14% higher average order values than organic search. Confirmed exactly via direct fetch. The conflicting figures in the corpus belong to different periods/framings and are each independently plausible rather than being simple misquotes: li#13's '15x growth in orders' (end-2025) predates this Q1 2026 data and cou | AI-referred sessions +8x YoY; AI-referred orders +13x YoY; conversion ~50% higher than organic search; AOV +14% vs organic search — Shopify Q1 2026 | Shopify (enterprise blog, Q1 2026 AI search insights) 2026-05-11 | Confirmed Đã xác nhận | ↗ |
| Salesforce's 2026 holiday predictions (published July 20, 2026) forecast that 20% of all 2026 holiday ecommerce traffic will come from AI chat agents — a mix of consumer-facing chatbots, autonomous backend agents, and competitor price-scraping bots, not solely human-directed AI shopping assistants — and that 1 in 3 ecommerce sites will have their own branded shopper agent live by Cyber Week 2026. Confirmed exactly via direct fetch, including the x#58 caveat about the 20% including scrapers and backend agents (not just shopper-facing bots) — this is Salesforce's own explicit wording, not a misreading. The 'retailers running their own agents grew sales 5 | 20% of 2026 holiday ecommerce traffic predicted from AI chat agents (includes bots, autonomous agents, and price-scrapers — not just human-directed shopping agents); 1 in 3 (~33%) ecommerce sites predicted to have a live branded shopper agent by Cyber Week 2026 | Salesforce (2026 Holiday Retail Predictions blog) 2026-07-20 | Confirmed Đã xác nhận | ↗ |
| Salesforce data (via Retail Dive, Nov 29, 2025) showed that on Black Friday 2025, U.S. online order volume fell 1% year-over-year and units per transaction fell 2%, while average selling price rose 7% — meaning the 3% YoY rise in total Black Friday dollar sales ($18 billion) came from higher prices, not more items or more orders, per Salesforce's own director of consumer insights, Caila Schwartz. Confirmed exactly as claimed via direct fetch, with an on-the-record Salesforce quote attached. Note this $18B figure is Salesforce's own US Black Friday total (per the conflicts note, likely omnichannel/in-store+online, distinct from Adobe's $11.8B US ONLINE- | US Black Friday 2025: online order volume -1% YoY; average selling price +7% YoY; units per transaction -2% YoY (against $18B total US Black Friday spend, +3% YoY) | Salesforce (Shopping Index data, via Retail Dive) 2025-11-29 | Confirmed Đã xác nhận | ↗ |
| Apptopia found ChatGPT referrals to U.S. retailer mobile apps grew 28% year-over-year over Black Friday weekend 2025, with Amazon capturing 54% of those referrals (up from 40.5% in 2024) and Walmart 14.9% (up from 2.7%) -- but such referrals were still just 0.82% of all ChatGPT app sessions on Black Friday, up from 0.64% a year earlier. The 30-second-window definition is a meaningful methodology detail: it's an inferred/proxy referral, not a tagged UTM or app-attribution link, so some of this 'referral' traffic could be coincidental app-switching rather than causal. Matches E-S16 claim closel | ChatGPT referrals to retailer mobile apps grew 28% YoY over Black Friday weekend 2025 (Thu-Sun). Amazon's share of ChatGPT referrals rose to 54% (from 40.5% in 2024); Walmart's rose to 14.9% (from 2.7% in 2024). ChatGPT referral sessions to e-commerce apps were 0.82% of all ChatGPT sessions on Black Friday 2025 (vs 0.64% on Black Friday 2024). | TechCrunch (Sarah Perez), citing Apptopia mobile-panel data December 2, 2025 | Confirmed Đã xác nhận | ↗ |
| Across the same 94 e-commerce brands, Visibility Labs/180 Marketing found ChatGPT-referred revenue was just 1.48% of non-branded-organic revenue for full-year 2025 ($474K vs $32.1M), rising to 2.2% in H2; organic sessions outnumbered ChatGPT sessions 70-to-1 for the year, narrowing to 47-to-1 by Q4 as ChatGPT traffic grew 1,079% (vs +17% for organic). All figures in the task's claim match the source exactly. Fits the pattern flagged in the corpus as 'best traffic, almost none of it' -- ChatGPT converts and monetizes reasonably well per-session but remains a rounding error in absolute revenue next to organic | Same 94-brand Visibility Labs/180 Marketing dataset, full year 2025: ChatGPT-referred revenue $474K vs non-branded-organic revenue $32.1M (ChatGPT = 1.48% of organic revenue for the full year, rising to 2.2% in H2 2025); organic sessions were 70x ChatGPT sessions across the full year, narrowing to 47x in Q4 2025; ChatGPT traffic grew 1,079% over the 12 months (1,544 visits in Jan 2025 to 18,202 in Dec 2025) vs non-branded organic traffic's +17% over the same period. | 180 Marketing (formerly Visibility Labs) blog Data Jan-Dec 2025; post bylined/updated February 23, 2026 | Confirmed Đã xác nhận | ↗ |
| Etsy CEO Kruti Patel Goyal told investors that agentic/AI traffic remained under 1% of Etsy's total site traffic through Q2 2026, even after growing roughly 15x year-over-year in Q4 2025 -- a small but consistently higher-AOV slice of traffic. Task doesn't specify which 2026 quarter -- both Q4 2025 and Q2 2026 disclosures say '<1%', so the claim holds across the year to date; use Q2 2026 if pairing with other Sept-2026-dated stats in this report. | Etsy CEO Kruti Patel Goyal has said agentic/AI traffic is under 1% of Etsy's total site traffic in each of the two most recent quarters disclosed: Q4 2025 (Feb 19, 2026 call) -- 'agentic traffic in Q4 increased roughly 15x YoY, though it still accounts for less than 1% of total traffic'; Q2 2026 (Aug 12, 2026 call) -- 'agentic experiences account for less than 1% of overall traffic, but they result in higher average order value.' | CustomerExperienceDive / PYMNTS, reporting Etsy Q4 2025 and Q2 2026 earnings calls February 19, 2026 (Q4 2025 call); August 12, 2026 (Q2 2026 call) | Confirmed Đã xác nhận | ↗ |
| John Lewis managing director Peter Ruis said AI-agent-driven product searches reached 2.5% of the retailer's searches in 2026, up from 0.3% a year earlier -- an eightfold rise, with around 60% of John Lewis orders now placed online. Directly confirmed with an exact quote from a named executive via a Reuters-bylined article; matches the task's stated value exactly. | Peter Ruis, managing director of John Lewis department stores, told Reuters that searches from AI agents had risen to 2.5% of product searches, from just 0.3% a year earlier -- roughly an eightfold increase over one year, describing the growth as 'exponential, and it's all age groups.' | Reuters (Sarah Young) September 3, 2026 | Confirmed Đã xác nhận | ↗ |
| Similarweb's clickstream-based 'Downstream Impact of AI Visibility' study (US desktop panel, 6 months, finance/travel/beauty verticals) found consumers were 2.5x more likely to visit an AI-recommended brand's site within 7 days, and that 55.9% of that AI-influenced traffic showed up in analytics as ordinary (branded) search rather than a direct AI referral -- meaning most of this influence is invisible in standard referral reporting. The task attributes this to 'Similarweb (compiled by Vauntly.ai)' -- I found no trace of 'Vauntly.ai' as a source or compiler in any result this session; the primary and all corroborating sources point directly to Similarweb's own report. Recommend dropping th | Similarweb's 'The Downstream Impact of AI Visibility' study (US desktop panel, six months of clickstream data, three verticals: finance, travel, beauty; head-to-head brand pairs like Amex/Capital One, Skyscanner/Kayak, Sephora/Ulta) found users were 2.5x more likely to visit an AI-recommended brand's website within the following 7 days than a non-recommended competitor, and 55.9% of that downstream traffic arrived via (branded) search rather than a direct click-through from the AI platform. AI-influenced visitors also showed deeper engagement (~12.0 pages / 11.8 min vs 6.5 pages / 5.6 min for non-AI-influenced visitors). | Similarweb ('The Downstream Impact of AI Visibility' report/blog); corroborated by Search Engine Journal Published/covered June 21-23, 2026 | Confirmed Đã xác nhận | ↗ |
| Bain & Company's 2026 Holiday Shopping Outlook (fielded with ROI Rocket, n=1,105 US online shoppers in Aug 2026 vs n=950 in Oct 2025) found 24% of shoppers now say they'll start their holiday shopping journey on a gen-AI platform such as ChatGPT, Claude or Gemini, up from 17% a year earlier, on a multi-select question. 17%->24% is +7pp, a ~41% relative increase, not the '50% increase' some social posts claim (li#459 in the conflicts field). Bain separately reports 13% plan to use a retailer's own AI agent (e.g. Walmart Sparky) -- additive to, not part of, the 24% gen-AI-plat | Gen-AI platforms (ChatGPT/Claude/Gemini) as a starting point for holiday shopping: 17% (Oct 2025, n=950) -> 24% (Aug 2026, n=1,105). Retailer/brand websites: 51% -> 60%. Amazon: 73% -> 71%. Question is multi-select. | Bain & Company, 2026 Holiday Shopping Outlook (infographic) 2026 (Bain's 2026 Holiday Shopping Outlook; fielding Aug 2026 vs Oct 2025 baseline) | Confirmed Đã xác nhận | ↗ |
| Narvar's 2026 Holiday Shopping Report, based on a survey of 1,348 US consumers and 100 senior US retail decision-makers published August 24, 2026 (with eTail Insights), found 65% of shoppers plan to use AI for at least one part of their holiday shopping, versus just 8% of retailers who feel very confident using AI to improve the experience. All sub-figures (43/33/29/78/8) match the report as covered by CPA Practice Advisor and corroborated by Forbes (Sharon Edelson, Sep 9 2026) and Chain Store Age. The 65% and 8% are different populations answering different questions -- describing the gap as a s | 65% of 1,348 US shoppers surveyed plan to use AI for at least one part of 2026 holiday shopping (43% gift discovery, 33% compare products/summarize reviews, 29% budgeting, 78% would use it if more personalized). Only 8% of 100 retail decision-makers are 'very confident' using AI to improve the shopping experience. | Narvar (2026 Holiday Shopping Report, with eTail Insights) 2026-08-24 | Confirmed Đã xác nhận | ↗ |
| PwC's Holiday Outlook 2026, based on a survey of 4,093 US consumers fielded June 15-26, 2026, found 29% of consumers plan to use AI somewhere in their holiday shopping this year, up from 22% in 2025, mostly to research products (76%), compare prices (55%) and stay on budget (25%); in apparel/footwear, 64% use AI for ideas but purchase from other retailers. pwc.com's own Holiday Outlook page (https://www.pwc.com/us/en/industries/consumer-markets/library/holiday-outlook-trends.html) returned a TLS certificate error on fetch this session, so this is confirmed via a syndicated article (Fibre2Fashion content re-run o | 29% of US consumers plan to use AI somewhere in 2026 holiday shopping, up from 22% in 2025. Among AI users: 76% research products, 55% compare prices, 25% stay on budget. In apparel/footwear, 64% use AI for ideas but buy elsewhere. | PwC, Holiday Outlook 2026 (US) syndicated 2026-09-09 (article), reporting PwC's Holiday Outlook 2026 survey fielded June 15-26, 2026 | Confirmed Đã xác nhận | ↗ |
| Google's 2026 holiday retail guide (Think with Google, distributed August 21, 2026), drawing on a Google-commissioned Ipsos Global Consumer Journeys survey (~14,305 AI-using and ~21,881 non-AI-using respondents across 25 countries, fielded December 2025), found 61% of US holiday shoppers used AI in their 2025 purchase journey, that AI users engaged with 12.1 touchpoints on average versus 4.2 for non-AI shoppers, and that 94% of AI users rated the tools valuable while 76% said AI saved them time summarizing product details. 12.1 vs 4.2 touchpoints computes to ~2.88x -- close enough to the claimed '2.8x' to be a rounding choice, not an error. Correctly supports the framing that AI lengthens exploration (more touchpoints) rather than collapsing the funnel; posts claiming AI 'collap | 61% of US holiday shoppers used AI platforms/capabilities during their 2025 purchase journey. AI-assisted shoppers engaged ~2.9x more touchpoints (12.1 vs 4.2, not exactly 2.8x). 94% of AI-using shoppers call AI tools valuable; 76% say AI saves time by summarizing product details/reviews. | Google / Think with Google (Ipsos-commissioned research) 2026-08 (guide distributed 2026-08-21; underlying Ipsos fieldwork Dec 2025) | Confirmed Đã xác nhận | ↗ |
| Across five separate 2026 studies, consumers consistently draw a line between AI as a research aid and AI as a buyer: ChannelEngine's Marketplace Shopping Behavior Report 2026 (4,500 shoppers across five countries, published Jan 7, 2026) found only 17% comfortable completing a purchase through AI; Ipsos's 'Shopping with AI' research found 27% of AI-aware consumers use AI for product research versus 9% who'd let it purchase autonomously; Syndigo's State of Product Experience 2026 (8,736 consumers across six countries, June 2026) found 32% have or would let an AI agent buy on their behalf; Accenture's Consumer Pulse research on grocery AI agents found about a third would let an agent finalize a purchase decision but only 9.5% want fully hands-off buying; and The Harris Poll's Algorithmic Aisle study found 72% of consumers comfortable with AI help only if they retain the final decision. All five figures line up consistently around the same theme (roughly 1/5 to 1/3 comfortable with AI completing/deciding a purchase, rising for research-only use). Ipsos and Accenture figures were confirmed via WebSearch snippets quoting the reports rather than | ChannelEngine (4,500 marketplace shoppers, US/UK/FR/DE/NL, Jan 2026): 17% comfortable completing a purchase through AI. Ipsos 'Shopping with AI': 27% of AI-aware consumers use AI for product research, 9% currently allow autonomous AI purchases. Syndigo State of Product Experience 2026 (n=8,736, US/UK/DE/FR/BR/AU, June 2026): 32% have or would let an AI agent buy on their behalf. Accenture Consumer Pulse (grocery): ~1/3 would let an agent make the final purchase decision, 9.5% ready for fully hands-off grocery purchasing. The Harris Poll Algorithmic Aisle: 72% comfortable with AI helping only if they make the final call. | ChannelEngine; Ipsos ('Shopping with AI'); Syndigo; Accenture; The Harris Poll 2026-01-07 (ChannelEngine); 2026 (Ipsos, Syndigo Aug 19 2026, Accenture Jul 15 2026, Harris Poll Jun 2026) | Confirmed Đã xác nhận | ↗ |
| The Harris Poll's 'The Algorithmic Aisle' study, fielded in June 2026 among 3,222 adults across the US, UK, Brazil and India covering beauty/wellness, personal care and household goods, found 78% of consumers assume brands pay to be recommended by AI, 72% are comfortable with AI help only if they keep the final say, 45% want paid or sponsored AI recommendations disclosed (rising to 52% in the UK), and that the UK is the most AI-skeptical market, with 38% wanting no AI involvement in shopping at all versus roughly 12% in India. Claim's stated sample of '3,000 consumers' rounds down from the actual 3,222 -- immaterial. All other figures (78%, 72%, 45%, UK-most-skeptical) confirmed directly from the Harris Poll's own article. | 78% of consumers assume brands pay to be recommended by AI; 72% comfortable with AI helping only if they make the final call; 45% want AI platforms to disclose paid/sponsored recommendations (52% in the UK); UK is the most AI-skeptical of the four markets (38% want no AI involvement at all, more than 3x India's rate). | The Harris Poll, 'The Algorithmic Aisle' 2026-06 (fieldwork); article published 2026 | Confirmed Đã xác nhận | ↗ |
| PYMNTS Intelligence's 'Global Digital Shopping Index: Merchant Edition' (commissioned by Visa Acceptance Solutions, surveying 1,185 retail merchants across the US, Brazil and the UAE) found only 23% of merchants can connect AI-driven traffic to the purchases it produces, a further 21% can see the traffic but not tie it to a sale, and 61% of merchants agree AI-generated results will influence purchases more than traditional search. The 61% 'expect AI results to influence purchases more than traditional search' figure comes from a related PYMNTS merchant article ('AI Takes the First Step in Shopping While Consumers Keep the Buy Button') rather than the same single page as the 23%/21% figu | 23% of merchants can connect AI-driven traffic to the purchases it produces; another 21% can see the AI-driven traffic but cannot connect it to a completed sale; 61% of merchants agree AI-generated results will influence purchases more than traditional search. | PYMNTS Intelligence (Global Digital Shopping Index: Merchant Edition, commissioned by Visa Acceptance Solutions) 2026 | Confirmed Đã xác nhận | ↗ |
| Akamai's 'Securing the Agentic Storefront' report (published July 15, 2026) found that nearly half (47.9%) of all AI-bot traffic on Akamai's global network between July and December 2025 hit the commerce vertical specifically. Radware's 2026 E-Commerce Bot Threat Report (June 23, 2026) found bad bots made up 43% of e-commerce traffic during the 2025 holiday season, up from 31% a year earlier, nearly matching the 46% share from human shoppers. Veriff's Identity Fraud Report 2026 (December 11, 2025) found e-commerce had a 19.2% net fraud rate in 2025 -- about five times the global average -- with impersonation behind more than 85% of all online fraud attempts. Correction/clarification: the claim's phrasing 'AI bots 47.9% of commerce traffic' is easy to misread as 47.9% of ALL commerce traffic; Akamai's own release specifies this is 47.9% of AI-BOT traffic that landed in the commerce vertical, a narrower and correctl | AI bots were 47.9% of all AI-bot traffic in the commerce vertical on Akamai's network, July-December 2025 (H2 2025) -- not 47.9% of all commerce traffic. Bad bots were 43% of e-commerce holiday-season traffic in 2025, up from 31% the prior year (Radware). E-commerce net fraud rate was 19.2% of verification attempts in 2025, ~5x the global average, with impersonation behind over 85% of observed fraud attempts (Veriff). | Akamai (State of the Internet: Securing the Agentic Storefront); Radware (2026 E-Commerce Bot Threat Report); Veriff (Identity Fraud Report 2026) 2026-07-15 (Akamai); 2026-06-23 (Radware); 2025-12-11 (Veriff) | Confirmed Đã xác nhận | ↗ |
| Profound's own analysis (published as 'ChatGPT 5.6 has transformed Shopping mode' on tryprofound.com, and reported by Search Engine Journal's Matt Southern on September 9, 2026) found that the share of ChatGPT Shopping product picks it classifies as feed-integrated jumped from 8.26% to 61.54% in a single day, July 10, 2026 -- the day after OpenAI's July 9 rollout of GPT-5.6 -- based on 1,757,723 tracked prompt runs that month. Over the same shift, references to the top 10 stores rose from 22.5% to 41.8% of picks, and the number of unique merchants ChatGPT referenced fell from 13,524 to 10,607. Of 687 merchants Profound tracked closely, 517 saw visibility swings of at least 33% (450 down, 67 up), and a least-squares regression relating each merchant's loss of web-search retrieval to its gain in feed retrieval explained 83% of the variance in those visibility changes. This is Profound's proprietary classification of its own customers' tracked prompts, not a randomized or independently audited sample of all ChatGPT Shopping traffic -- appropriate to cite as 'Profound found,' not as an OpenAI-confirmed figure; OpenAI has not | Feed-integrated ChatGPT Shopping picks jumped from 8.26% to 61.54% on July 10, 2026 (across 1,757,723 tracked prompt runs in July). Top-10 store share rose from 22.5% to 41.8%. Unique merchants referenced fell from 13,524 to 10,607 (-21.6%). Of 687 tracked stores, 517 saw visibility swings of >=33% (450 lost significant visibility, 67 gained); a regression model of web-search-loss vs. feed-gain explained 83% of the variation in those 517 stores' visibility changes. | Profound (tracking/analytics firm), corroborated by Search Engine Journal (Matt Southern) 2026-09-03 (Profound blog post); 2026-09-09 (Search Engine Journal coverage) | Confirmed Đã xác nhận | ↗ |
| ReFiBuy's Q2 2026 AI1000 index, co-developed with Digital Commerce 360 and published August 18, 2026, found that 972 of the 1,000 ranked retailers changed position quarter over quarter, the median retailer moved 35 spots, 641 moved at least 25 positions, the new #1-ranked retailer (Nixon) ranks only #722 by actual online sales, and just 10 of the 100 largest online retailers by sales appear anywhere in the AI1000's top 100. Fully confirmed by direct fetch -- every number in the corpus's claim (972/1000, median 35, Nixon #722, 10-of-100) matches exactly. This is a vendor-built readiness index (ReFiBuy scores retailers 0-100 on bot-friendliness, AI-source traffic, source diversity | ReFiBuy's AI1000 (Q2 2026, published Aug 18, 2026): of 1,000 ranked retailers, 972 changed rank; median move was 35 spots; 641 moved at least 25 positions; Nixon took the #1 AI1000 spot despite ranking #722 by online sales; only 10 of the 100 largest online retailers by sales appear in the AI1000's top 100. | ReFiBuy / Digital Commerce 360 AI1000 index, via ChannelX reporting 2026-08-18 | Confirmed Đã xác nhận | ↗ |
| A Bain & Company brief published August 8, 2025, citing Sensor Tower's opt-in mobile-panel data, found that ChatGPT click-through events roughly tripled from about 100,000 in March 2025 to about 300,000 in June 2025, average click-through rate rose from 2.2% to 5.7%, total ChatGPT prompt volume grew nearly 70% over the first half of 2025, and shopping-related prompts rose from 7.8% to 9.8% of all ChatGPT prompts (a 25% relative gain on top of the 70% overall volume growth) in the same window. Fully confirmed by direct fetch of Bain's own page -- every figure in the corpus's claim matches. Important methodology caveat also confirmed directly: Sensor Tower's numbers come from 'a panel of opt-in Android and iOS users' running utility apps (not a censu | Bain & Company brief (Aug 8, 2025), citing Sensor Tower panel data: ChatGPT click-through events roughly tripled, ~100K (March 2025) to ~300K (June 2025); average CTR rose from 2.2% to 5.7%; total ChatGPT prompt volume grew ~70% in H1 2025; shopping prompts rose from 7.8% to 9.8% of all prompts over the same period. | Bain & Company, 'How Customers Are Using AI Search' (Sensor Tower data) 2025-08-08 | Confirmed Đã xác nhận | ↗ |
| Lowe's CEO Marvin Ellison told analysts on the company's Q1 2026 earnings call (reported May 21, 2026) that customers who use the Mylow AI shopping assistant convert at three times the rate of customers who don't use it -- the same 'almost 3X' figure separately cited by Neelima Sharma, Lowe's SVP of Omnichannel & E-commerce Technology (not 'SVP Digital & Tech' as loosely stated), in trade-press interviews. Corrects/strengthens the attribution: the corpus cites only Neelima Sharma (SVP), but the same ~3x figure was independently stated by Lowe's own CEO on a public earnings call (May 20, 2026), which is a more authoritative company-official source than a trade-co | ~3x: Lowe's says customers who engage with its MYLOW/Milo AI shopping assistant convert at about 3x the rate of customers who don't use it. | Customer Experience Dive (reporting Lowe's Q1 2026 earnings call remarks by CEO Marvin Ellison) 2026-05-21 | Confirmed Đã xác nhận | ↗ |
| McKinsey (Oct 20, 2025 research) estimates agentic commerce could orchestrate $3 trillion to $5 trillion of global consumer commerce and as much as $1 trillion of US retail revenue by 2030; Morgan Stanley (Dec 9, 2025) put US agentic e-commerce at up to $385 billion by 2030 (10-20% of US online spend); Bain & Company (Dec 17, 2025) forecasts $300-500 billion, or 15-25% of US e-commerce, by 2030; and Deloitte's Global Retail Industry Outlook says AI agents could influence -- not necessarily complete -- up to 25% of global e-commerce sales by 2030. Confirmed McKinsey and Morgan Stanley figures via direct fetch of Digital Commerce 360's reporting (which quotes the firms verbatim); confirmed Bain's $300-500B/15-25% figure and its definition of 'agentic commerce' (purchases initiated/influenced/completed by | McKinsey: agentic commerce could orchestrate $3-5T globally and up to ~$1T in US retail revenue by 2030 (Oct 2025 research). Morgan Stanley: agentic shoppers could drive up to $385B in US e-commerce sales by 2030, 10-20% of US e-commerce spend (Dec 9, 2025 report). Bain: US agentic commerce could reach $300-500B by 2030, ~15-25% of US e-commerce (Dec 17, 2025). Deloitte: AI agents could influence up to 25% of global e-commerce sales by 2030. | McKinsey, Morgan Stanley, Bain & Company, Deloitte (via Digital Commerce 360 reporting + Bain's and Deloitte's own pages, all fetched directly) 2025-12-17 | Confirmed Đã xác nhận | ↗ |
| Visa's December 18, 2025 press release 'Visa and Partners Complete Secure AI Transactions' states Visa Intelligent Commerce has more than 100 ecosystem partners worldwide, over 30 actively building in its sandbox, and over 20 agents/agent enablers integrating directly, and that Visa 'predicts that millions of consumers will use AI agents to complete purchases by the 2026 holiday season'; separately, Visa's Business-to-AI (B2AI) Report, fielded with Morning Consult January 29-February 6, 2026 among 512 US business decision-makers, found 53% would allow AI agents to negotiate prices or terms directly with other AI agents on their behalf. Both figures confirmed by direct fetch of Visa's own newsroom pages (releaseId.21961.html for the partner/holiday-season numbers, releaseId.22266.html for the B2AI survey). This is the strongest-sourced stat in the batch: primary-source press releases fetched | Visa Intelligent Commerce (Dec 18, 2025 press release): 100+ ecosystem partners, 30+ building in the VIC sandbox, 20+ agents/agent enablers integrating directly; Visa predicts millions of consumers will use AI agents to complete purchases by the 2026 holiday season. Visa's B2AI Report (with Morning Consult, fielded Jan 29-Feb 6, 2026, n=512 US business decision-makers + 2,000 US adults): 53% of US businesses would let their AI agents negotiate prices/terms with other AI agents. | Visa Inc. (official newsroom press releases) 2025-12-18 | Confirmed Đã xác nhận | ↗ |
| Triple Whale's own Dec 2025 BFCM recap press release states brands using its platform generated $2.88 billion in revenue over the BFCM weekend, representing 19.7% of all Shopify merchant sales. The $2.88B/19.7% figures match the panel's value (which also cites a rounded '$2.9B' seen elsewhere — same number, rounded). Confirmed this is platform-wide merchant GMV share, NOT evidence that Moby (the AI assistant) or any AI agent caused that revenue — Tri | $2.88B in revenue generated by Triple Whale brands over BFCM 2025 weekend (Nov 28 - Dec 1, 2025), = 19.7% (~20%) of all Shopify merchant BFCM sales. | Triple Whale — 'Triple Whale Reports Record BFCM Engagement as Brands Turn to AI to Drive Smarter Growth' (PR Newswire) 2025-12-04 | Confirmed Đã xác nhận | ↗ |
| Triple Whale's own BFCM 2025 recap states its conversational AI assistant Moby and its suite of AI Agents 'handled nearly 25,000 real-time questions across five languages' over the BFCM weekend. Number confirmed verbatim against Triple Whale's own release. The panel's arithmetic observation stands: nearly 25,000 questions against >50,000 brands served is well under 1 query/brand on average if every brand were counted, which is a fair thing to raise on | Nearly 25,000 real-time questions handled by Moby (and Triple Whale's AI Agents) across five languages over BFCM 2025. | Triple Whale — 'Triple Whale Reports Record BFCM Engagement as Brands Turn to AI to Drive Smarter Growth' (PR Newswire) 2025-12-04 | Confirmed Đã xác nhận | ↗ |
| Klaviyo's own Dec 2, 2025 newsroom release, 'Klaviyo Breaks Records with First AI-powered BFCM,' states Klaviyo 'delivered more than 22.7B messages (up 25% YoY)' generating 'more than $3.8B in KAV,' and that usage of AI-driven product recommendations 'jumped 45% year-over-year' while 'revenue from those messages' surged '71% year-over-year' — i.e., the 71% is Klaviyo's own year-over-year revenue comparison for AI-recommendation messages, not a within-period lift of AI-recommendation messages over non-AI messages. All figures confirmed against Klaviyo's own release. The panel's conflict note about the 71% figure's baseline is well-founded and should be resolved as: it is a YEAR-OVER-YEAR revenue comparison for AI-recommendation-message revenue, NOT a same-period lift of | 22.7B+ messages (+25% YoY); $3.8B+ Klaviyo Attributed Value/KAV (+27% YoY); AI-driven product-recommendation usage +45% YoY, with revenue from those AI-recommendation messages +71% YoY; 'nearly 20,000' brands had their best sales day ever. | Klaviyo — 'Klaviyo Breaks Records with First AI-powered BFCM' (Klaviyo Newsroom) 2025-12-02 | Confirmed Đã xác nhận | ↗ |
| Meta executives stated on the Q4 2025 earnings call that GEM plus sequence-learning improvements drove a 3.5% lift in Facebook ad clicks and a >1% gain in Instagram conversions, a new runtime model drove a 3% conversion-rate increase, and the latest incremental-attribution model rollout drove a 24% increase in incremental conversions versus the standard attribution model. Verified verbatim against the official Meta IR transcript PDF (fetched, then extracted with pdftotext locally): 'the GEM and sequence learning improvements together drove a 3.5% lift in ad clicks on Facebook and a more than 1% gain in conversions on Instagram | GEM+sequence learning: +3.5% Facebook ad clicks, >1% Instagram conversions (Q4 2025). New runtime model (IG Feed/Stories/Reels): +3% conversion rate (Q4 2025). Incremental attribution feature's latest Q4 rollout: +24% incremental conversions vs standard attribution model. | Meta Platforms Q4 2025 Earnings Call Transcript (official IR) 2026-01-28 | Confirmed Đã xác nhận | ↗ |
| PYMNTS Intelligence, in its Global Digital Shopping Index: Merchant Edition (commissioned by Visa Acceptance Solutions), reports that 23% of merchants can clearly identify both AI-driven traffic and purchases today, while another 21% can recognize agentic traffic but not connect it to completed purchases. Figures match the claim exactly (23% / 21%). Flag for the report: this is a vendor-commissioned survey (Visa Acceptance Solutions paid for it), so cite it as 'PYMNTS Intelligence, commissioned by Visa Acceptance Solutions' rather than as an unqualified indepen | 23% of merchants can clearly identify both AI-driven traffic and purchases; another 21% can recognize agentic/AI traffic but cannot connect it to completed purchases. | PYMNTS Intelligence — 'How 23% of Merchants Captured Retail's Next Agentic Commerce Advantage' (Global Digital Shopping Index: Merchant Edition) 2026 | Confirmed Đã xác nhận | ↗ |
| Narvar's 2026 Holiday Shopping Report, based on a survey of 1,348 consumers and 100 senior retail decision-makers, found only 8% of retailers 'very confident' in using AI to improve the shopping experience, while 65% of shoppers plan to use AI for at least one part of their holiday shopping. Figures match the claim exactly (8% / 65%, n=1,348 / n=100). This is a vendor (Narvar, a post-purchase/delivery-tracking SaaS platform) self-commissioned survey used for PR/thought-leadership — the framing ('retailers unprepared for AI-savvy shoppers') promote | 8% of retail decision-makers are 'very confident' in their ability to use AI to improve the shopper experience, vs 65% of shoppers who plan to use AI for at least one part of their holiday shopping. | Narvar — 2026 Holiday Shopping Report ('Delivering the 2026 Holiday Shopping Season') 2026-08-24 | Confirmed Đã xác nhận | ↗ |
| Adobe Analytics reported U.S. shoppers spent $20 billion via Buy Now, Pay Later over the full 2025 holiday season (up 9.8% YoY); BNPL crossed $1 billion for the first time on a single day on Cyber Monday ($1.03B, +4.2% YoY); on Black Friday, BNPL spend was actually $747.5 million (about 6.3% of Black Friday digital sales) — the $761.8 million figure was Adobe's pre-season FORECAST for Black Friday BNPL, not the confirmed actual. The $761M in the claim is Adobe's October 2025 pre-season forecast for Black Friday BNPL (+11% YoY projected), not the confirmed actual of $747.5M reported after Black Friday. Season total ($20.0B/+9.8%) and Cyber Monday (>$1B, record) both check out exactly a | Season BNPL: $20.0B, +9.8% YoY (confirmed). Cyber Monday BNPL: $1.03B, +4.2% YoY, a record day (confirmed, 'over $1 billion'). Black Friday BNPL actual: $747.5 million (~6.3% of Black Friday digital sales) — NOT $761M. | Adobe (Adobe Digital Insights); actual Black Friday BNPL figure via Digital Commerce 360 citing Adobe Analytics 2026-01-07 | Corrected Đã sửa lại | ↗ |
| Salesforce's pre-Cyber Week analysis of Oct 1-Nov 15, 2025 found average selling prices up 7% YoY in the U.S. and 5% YoY globally; its final Cyber Week 2025 (Nov 25-Dec 1) results reported average selling price up 6% YoY globally. There is no Salesforce figure stating Cyber Week prices rose 5% in 2024 vs 7% in 2025 -- the claim appears to conflate the US-vs-global split of the pre-season forecast with a year-over-year comparison. Fetched both the Nov 20, 2025 pre-Cyber-Week predictions page (verbatim: 'The average selling price (ASP) rose 5% YOY globally and 7% in the U.S.', covering Oct 1-Nov 15, 2025 vs the same period a year earlier) and the full text of the Dec 5, 2025 final result | Salesforce pre-Cyber-Week trend data (Oct 1-Nov 15, 2025, YoY): ASP +7% in the US, +5% globally. Separately, Salesforce's final Cyber Week 2025 (Nov 25-Dec 1) results: ASP +6% YoY, reported as one global figure with no US/global split and no stated comparison to 2024's price-growth rate. | Salesforce 2025-11-20 | Corrected Đã sửa lại | ↗ |
| Charm.io estimated TikTok Shop exceeded $14 billion in total US sales in 2025, with its top category, beauty and personal care, generating $2.7 billion in US revenue from 147 million items sold at an $18.57 average unit price, led by Medicube ($98.9M) as the top beauty shop. The claim conflates two different Charm.io datasets. (1) The platform-wide full-year total is real but the precise figure and framing don't match: multiple outlets citing Charm.io say TikTok Shop 'exceeded $14 billion' in 2025 US sales (not a precise $14.4B), | TikTok Shop US full-year 2025: Charm.io puts total US sales at 'more than $14 billion' (rounded estimate; competing estimates from other trackers range $14B-$15.8B). Beauty & Personal Care specifically: $2.7B revenue, 147 million items sold, $18.57 average unit price (Charm.io, Jan 1-Dec 31 2025) — Medicube was the #1 BEAUTY shop at $98.9M, not the #1 shop platform-wide. | Charm.io 2026-01 | Corrected Đã sửa lại | ↗ |
| Haus's Cyber Week Incrementality Report found that brands ending ad-spend tests at least three weeks before BFCM saw a 105% lift once delayed (post-treatment-window) conversions were counted, versus 75% for brands testing closer to BFCM, and that CTV's post-treatment lift jumped 344% during Cyber Week; separately, Haus's Meta Report (640 experiments) found Meta Advantage+ over-reports revenue by about 12 percentage points versus manual campaigns, and its YouTube study (190 tests) found YouTube delivers 3.4x more incremental lift than Google Ads reporting credits it with. 3 of 4 sub-claims confirmed as stated; the 'brands that ramped spend early: 75% efficiency improvement' sub-claim has the number attached to the wrong group. Haus's own wording is: brands that STOPPED their test (ended treatment) 3+ weeks before BFCM saw the H | Haus 2025 Cyber Week Incrementality Report: brands that ended their incrementality test at least 3 weeks before BFCM saw +105% post-treatment-window lift, vs. +75% for brands whose test ran closer to BFCM (the claim reverses which group gets which number); CTV showed the largest median post-treatment-window lift increase at +344% during BFCM (confirmed); separately, Haus's 'Meta Report' (640 experiments over 18 months) found Meta Advantage+ over-reports revenue by ~12 percentage points vs. manual campaigns (confirmed); and Haus's YouTube report (190 tests) found YouTube drove 3.4x more incremental DTC lift than Google Ads platform reporting showed (confirmed) | Haus (incrementality vendor) 2025-12-11 | Corrected Đã sửa lại | ↗ |
| The Conference Board reported that its Consumer Confidence Index fell 6.8 points to 88.7 in November 2025 (Expectations Index down 8.6 points to 63.2), which Chief Economist Dana Peterson described as confidence tumbling 'to its lowest level since April' after several months of moving sideways. The two headline numbers (index ~88, expectations ~63) are confirmed almost exactly (88.7 and 63.2). However, 'largest drop since April' is a materially different claim from what The Conference Board actually said, which is that the resulting LEVEL is the lowe | Conference Board Consumer Confidence Index, November 2025: 88.7 (down 6.8 pts from October's 95.5); Expectations Index: 63.2 (down 8.6 pts); Present Situation Index: 126.9 (down 4.3 pts). The Conference Board's own characterization is that the INDEX LEVEL fell to its 'lowest since April' 2025 — not that the size of the monthly drop itself was 'the largest since April.' | The Conference Board 2025-11-25 | Corrected Đã sửa lại | ↗ |
| Omnisend's holiday 2025 survey found that between 14% and 21% of shoppers (by country) plan to use Buy Now Pay Later for holiday purchases, and that a large majority (75-86% depending on country) plan to use AI in some way while holiday shopping. The 40-45% BNPL figure in the claim directly contradicts Omnisend's own published 14-21% BNPL figure — likely either misattributed from a different survey (e.g., the PayPal-commissioned BNPL survey found in searches, 'From AI to BNPL, PayPal Survey Reveals How | Omnisend's 2025 holiday survey of ~4,000 consumers (US/Canada/UK/Australia) found 14-21% of shoppers (by country) plan to use Buy Now Pay Later, mostly for larger purchases — NOT 40-45% as claimed. A separate/later Omnisend blog (same broad survey program) reports 75-86% of shoppers across those countries plan to use AI in some form for holiday shopping tasks, far above the claimed '~1 in 4'. The $250-499 most-common-spend-bracket and the specific 1-in-3/1-in-5 shopping-start-timing breakdown, and the 45-56% 'rising prices' figure, could not be located in any Omnisend material checked. | Omnisend 2025 | Corrected Đã sửa lại | ↗ |
| Deloitte projects 2026-27 holiday retail sales of $1.70-1.71 trillion (+4.0-4.8%) and e-commerce sales of $316.1-318.9 billion (+7.5-8.4%), driven partly by disposable income growth of 4.5-5.2%; separately, Bain & Company's Consumer Lab Holiday Survey 2026 (with ROI Rocket, US n=1,105) found 24% of online shoppers plan to start holiday shopping on generative-AI platforms, up from 17% in its 2025 survey (n=950) — a 7-point increase, not '17%'. The dollar/percent retail and e-commerce figures match Deloitte's own forecast exactly. But the AI-adoption figure attached to this claim is misattributed to Deloitte and is actually Bain's (duplicating A-S54, as the conflicts field already flagged). The paren | Deloitte: total US holiday retail sales +4.0-4.8% YoY to $1.70-1.71T; e-commerce +7.5-8.4% YoY to $316.1-318.9B; disposable personal income +4.5-5.2%, all for Nov 2026-Jan 2027. The '24% plan to use AI to start search, up 17% from 2025' is NOT Deloitte data — it is Bain & Company's figure (24% in Aug 2026 vs 17% in Oct 2025), a 7-percentage-point rise, not a 17% increase. | Deloitte (holiday retail forecast) + Bain & Company (Consumer Lab Holiday Survey 2026) 2026-09-10 (Deloitte); 2026-08 (Bain) | Corrected Đã sửa lại | ↗ |
| McKinsey & Company's ConsumerWise 'State of the US Consumer' update (nationwide US sample of 4,000+, fielded July 29-Aug 5, 2026) found 47% of consumers plan to spend about the same this holiday season, 23% more and 21% less; 45% plan to start shopping by the end of October; and 46% say they will probably or definitely use AI tools for holiday shopping (mainly to compare products/prices, find deals, or get gift ideas). Net spending intent was reported weakest for home decor and furniture (around -45 net), with accessories also negative but the precise category-level numbers in the claim (-32 home decor, -30 accessories) do not match what secondary reporting quotes from the chart. mckinsey.com itself timed out on every direct fetch attempt; figures cross-checked via two independent secondary write-ups (MediaPost and Times of Oman/ANI wire) that both quote McKinsey directly and agree with each other on 47/23/21/9, 45% October, and 46% AI | McKinsey ConsumerWise State of the US Consumer (n>4,000, fielded July 29-Aug 5, 2026): 47% plan to spend about the same, 23% more, 21% less, 9% undecided (47+23=70% same-or-more, matching the claim); 45% plan to start holiday shopping by end of October; 46% say they will probably/definitely use AI tools for holiday shopping (of whom ~49% for comparing products/prices, 48% for deals/discounts, 47% for gift ideas). Net purchase intent is reported lowest for home decor (-45), then accessories, then furniture (also reported at -45 in one source) — not '-32' and '-30' as claimed. | McKinsey & Company, ConsumerWise (via SIVO Insights survey partner) 2026-08-28 (published); survey fielded 2026-07-29 to 2026-08-05 | Corrected Đã sửa lại | ↗ |
| Shopify's official BFCM 2025 release states shopping peaked at $5.1 million per minute at 12:01pm EST on Black Friday, but the '81+ million customers' figure it reports is for the full Black Friday-Cyber Monday weekend, not Black Friday alone; no Shopify-published 'shoppers on Black Friday' figure of ~30 million could be found. The peak-sales-per-minute figure is confirmed and correctly tied to Black Friday specifically. But the '30M+ shoppers that day' claim does not match anything in Shopify's official materials: the press release's only shopper count (81+ million) explicitly cover | $5.1M/minute peak confirmed for Black Friday (12:01pm EST); no official Shopify figure of '30M+ shoppers' for Black Friday alone -- the only published shopper count is 81+ million customers for the ENTIRE BFCM weekend | Shopify (official investor press release) 2025-12-01 | Corrected Đã sửa lại | ↗ |
| Retail Economics' 'ZigZag UK Returns Benchmark 2025' (produced with ZigZag Global) forecasts the value of UK non-food returns will fall to £25.1 billion in 2025, down from £26.7 billion in 2024, as the overall returns rate eases from 21% to 19.5%. Value (£25.1bn) confirmed, but the attribution in the corpus ('unattributed, UKP Worldwide, logistics vendor') is wrong or at best a secondary re-citation -- the actual originating research is Retail Economics x ZigZag Global's 'UK Returns Benchmark 2025', not | £25.1bn forecast value of UK non-food returns in 2025 (down from £26.7bn in 2024), with the overall return rate easing from 21% to 19.5% | Retail Economics / ZigZag Global ('UK Returns Benchmark 2025') 2025 (report edition covering 2025 forecast) | Corrected Đã sửa lại | ↗ |
| A field experiment on Instacart covering 840,000+ customers (Knight & Mitrofanov, 'Disclosing Low Product Availability,' Management Science) found that disclosing low-inventory warnings produced a 5.7% increase in long-run order frequency and a 5.8% increase in customer spending, driven by a 3.7% reduction in refunds and a 2.9% reduction in product replacements -- the order-frequency figure matches the claim exactly, but the refund reduction is 3.7%, not the 2.6% cited. RETHINK Retail's 'State of US Omnichannel 2026' report (the stated proximate source) could not itself be located (see B-S27), so it's unconfirmed whether RETHINK Retail cited this exact academic study or a different one. The +5.7% order-frequency match is a st | Closest identifiable public study: order frequency +5.7% (matches claim exactly); but refunds -3.7% (not -2.6% as claimed), plus a -2.9% drop in product replacements and +5.8% spend, not cited in the claim | Knight, B. & Mitrofanov, D., 'Disclosing Low Product Availability: An Online Platform's Strategy for Mitigating Stockout Risk,' Management Science 2024 (Management Science publication; underlying field experiment date not independently confirmed here) | Corrected Đã sửa lại | ↗ |
| Strike Social's holiday ad-cost analysis states that Halloween week shows 15-34% lower ad costs than the Black Friday-through-Cyber Monday window, and that the most expensive days of the year fall 7-10 days before Christmas Eve — the 15-34% figure should be attributed to Strike Social (Lee Baler), not to eHouse. Correction: I fetched the Strike Social article and it contains the exact 15-34% figure and the 7-10-days-before-Christmas-Eve peak-timing claim verbatim, attributed to Strike Social/Lee Baler — not to eHouse. I searched extensively for an eHouse (Shopify Plat | Halloween-week ad costs run ~15-34% lower than Black Friday-through-Cyber-Monday costs — but this figure traces to Strike Social (Lee Baler, VP Sales & Strategy), not to 'ehouse, Shopify Platinum agency' as the corpus attributes it. | Strike Social — 'Christmas Advertising Costs: When to Scale Holiday Spend' (author: Lee Baler) 2026 | Corrected Đã sửa lại | ↗ |
| Klaviyo's official BFCM 2025 release states revenue from repeat customers grew 13.5% year over year, outpacing new-buyer revenue growth of +9% YoY. Klaviyo's own press release/SEC filing says 13.5%, not 13.6% — the task's figure is off by 0.1pp, likely from a secondary summary rounding differently. The +9% new-buyer figure is not in the primary press release text itself but is repeated consistently across | Repeat-buyer revenue +13.5% YoY vs. new-buyer revenue +9% YoY | Klaviyo, "Klaviyo Breaks Records with First AI-powered BFCM" (press release), corroborated by secondary trade coverage 2025-12-02 | Corrected Đã sửa lại | ↗ |
| "Black Friday itself saw the highest email volume globally, with 204.1M emails sent." ... Cyber Monday: "172M" emails sent. Sunday, Nov 24: "email sends surging 15% compared to the day prior." Found a real, fetchable primary source with actual printed (not just charted) numbers for 2 of the 4 cited days: Black Friday 204.1M (corpus said ~203M -- close, ~0.5% off) and Cyber Monday 172M (corpus said ~171M -- close, ~0.6% off). These are near-matches, | ~204.1M emails on Black Friday (Nov 29); ~172M on Cyber Monday (Dec 2). Nov 24 (~80M) and Dec 1 (~115M) figures NOT found in any fetchable Omnisend source -- only that Nov 24 sends were 'surging 15% compared to the day prior' (no absolute number given). | Omnisend Blog, "Early BFCM Sales: A New Norm in Ecommerce" 2024-12-09 | Corrected Đã sửa lại | ↗ |
| Adobe Analytics reported US online spending of $10.8 billion on Black Friday 2024 (up 10.2% YoY) and $13.3 billion on Cyber Monday 2024 (up 7.3% YoY), as part of $41.1 billion in total Cyber Week online spend. The dollar figures themselves are confirmed exactly, but attribution should be corrected: this is an Adobe Analytics/Adobe Digital Insights figure, not an Omnisend-originated statistic — Omnisend (and the social post) is merely relaying Adobe's number. Report | Black Friday 2024 US online spend: $10.8B (+10.2% YoY); Cyber Monday 2024: $13.3B (+7.3% YoY) | Adobe Digital Insights / Adobe Analytics 2024-12-03 | Corrected Đã sửa lại | ↗ |
| Omnisend's 2025 holiday survey (fielded by Cint, ~4,000 consumers across the US, Canada, UK, and Australia) found that only 21% of shoppers planned to use Buy Now, Pay Later on Black Friday and 17% on Cyber Monday — 'nearly half' actually describes the share who said they prefer to pay in full upfront, the opposite framing. About 1 in 4 shoppers (20-28% depending on country) reported using AI for gift ideas/deals in the prior six months, and large majorities (64-75% depending on country) said they plan to shop earlier than the traditional October-Cyber Monday window. The BNPL figure in the task ('nearly half plan BNPL') is a misreading/mix-up: the primary source's 'nearly half' figure describes consumers who prefer paying upfront, not BNPL adopters, and actual BNPL intent is 21% (BF) / 17% (CM). A separate ~43-45% 'BNPL pl | Survey (Omnisend/Cint, Aug 2025, n=4,000, US/Canada/Australia/UK): 21% plan to use BNPL on Black Friday and 17% on Cyber Monday (NOT 'nearly half'); ~20-28% (roughly '1 in 4') have used AI for gift ideas/deals in the past 6 months; majorities (75% US, 70% UK, 71% Canada, 64% Australia) plan to shop early, well before the typical Oct-Cyber Monday window | Omnisend 2025 holiday shopping survey (via Cint) 2025-08 | Corrected Đã sửa lại | ↗ |
| Martin Goodson's 2014 Qubit research paper found that when testers check results daily and stop as soon as they see statistical significance (p<0.05) — a common but statistically invalid practice — at least 80% of the resulting 'winning' tests are false positives; the paper does not study Black Friday/Cyber Monday specifically, and the claim that 'during BFCM that number climbs even higher' is an unsupported extrapolation added later by secondary sources, not a Qubit finding. This is an 11-year-old methodology paper about statistical practice generally (any A/B testing tool, any industry), not an ecommerce- or BFCM-specific study, and it predates the 'BFCM' framing entirely. The report should cite the 80% figure only with its actua | The '80%' figure is real and traceable to Martin Goodson's Qubit white paper 'Most Winning A/B Test Results Are Illusory' (Jan 2014): 'at least 80% of the winning results are completely worthless' — but this specifically describes tests that are monitored daily and stopped as soon as p<0.05 is hit (classic 'peeking'/early-stopping), not A/B testing in general, and has no connection to BFCM or peak season. | Qubit Research / Martin Goodson (primary paper, 2014); Chronos/SplitBase-style secondary sources add the BFCM framing 2014-01 | Corrected Đã sửa lại | ↗ |
| SplitBase's own blog states that most A/B tests should run for three to four full weeks and reach at least 100 conversions per variation (alongside hitting the required sample size and 95% statistical significance) before being trusted — the claim's 'at least 4 weeks' overstates SplitBase's actual minimum, which starts at three weeks. The conversions figure (100+ per variation) is confirmed verbatim. The duration figure is a minor but real correction: source says '3 to 4' weeks, the claim rounds up to 'at least 4 weeks.' The conflict note's underlying point stands independently of this corr | SplitBase's own published guidance says run tests for 'three to four full weeks' (not a flat 'at least 4 weeks') plus at least 100 conversions per variation, required sample size, and 95% significance — all four conditions together, not just duration and conversions. | SplitBase (self-reported agency guidance) undated | Corrected Đã sửa lại | ↗ |
| Haus's Olivia Kory published a 2025 year-in-review summary confirming YouTube drives 3.4x more incremental lift than Google Ads reports, TikTok delivers 68% more lift in the post-campaign window (1.9x with Amazon/retail impact included), and CTV efficiency improves 344% during Cyber Week when delayed conversions are counted; no Haus source found ties a 75% efficiency gain specifically to brands that ramped spend early in Q4. Three of four figures confirmed exactly, each independently cross-checked against a separate Haus blog post with its own disclosed sample size — unusually well-sourced for this corpus. The 75%-early-ramp claim appears to be a conflation: the only "75%" in Haus | YouTube drives 3.4x more incremental lift than Google Ads reports (confirmed); TikTok delivers 68% more lift in the post-campaign window and 1.9x more value counting Amazon/retail impact (confirmed); CTV efficiency improves 344% during Cyber Week when delayed conversions are counted (confirmed). The fourth sub-claim — "brands that ramped spend early in Q4 saw 75% efficiency improvements" — is NOT supported by any source found; the only matching "75%" figure in Haus's published research is a different metric (overall efficiency improving "more than 75%" fleet-wide when latent/delayed impact is counted, unrelated to early-ramp timing). | Olivia Kory (X/Twitter, Haus year-in-review) + Haus blog posts 2025-12-28 | Corrected Đã sửa lại | ↗ |
| The 0.76%/0.65% CTR gap is genuinely from the Taboola-sponsored Columbia/Harvard/CMU/TU Munich study of 500M+ impressions, but that study concluded AI and human ads perform comparably after controls, with no significant conversion penalty for AI ads; the -8%/-14% conversion-rate-drop-by-price-tier figures are from an unrelated Digital Applied "AdBeacon" benchmark, and the two datasets were combined into a single graphic implying one unified finding that AI ads convert worse, which is not what either underlying study actually found together. This is the clearest case of a misleading composite in the whole batch: the source image (posted by @HedgieMarkets) cites both sources correctly in its fine print, but visually and rhetorically presents them as one study proving AI ads 'stop converting' — wher | The CTR figures (0.76% AI ads vs. 0.65% human ads, from 500M+ impressions / 3M+ clicks) are real and correctly attributed to the Taboola/Columbia/Harvard/CMU/TU-Munich study — but that study's own conclusion is that AI-generated ads perform COMPARABLY to human ads once statistical controls are applied, with no reduction in downstream conversion found. The conversion-rate-drop-by-price-tier figures (-8% over $100, -14% over $500, no difference under $100) come from a completely separate, unrelated dataset (Digital Applied's "AdBeacon" benchmark, ~50,000 ad variations across Meta/Google/TikTok, published 2026-07-22) that has nothing to do with the Taboola study. The viral graphic ("AI Ads Win Clicks. Human Ads Win Sales.") merges these two unconnected studies into one visual, producing a conclusion neither source supports and that directly contradicts Taboola's own stated finding. | Columbia Business School / Taboola press release (CTR study); AdBeacon / Digital Applied blog (conversion-by-price-tier benchmark) 2026-01-28 (Taboola study); 2026-07-22 (AdBeacon/Digital Applied benchmark) | Corrected Đã sửa lại | ↗ |
| Northbeam's Media Buyer Index for the week of June 29-July 5, 2026 shows Meta down across CVR (-8.18%), ROAS (-2.39%), and budget share change (-2.98 points, ending at 49.13% share), while the social-media post relaying this data (by @ecommcowboy, quoting Bryan Bumgardner's Northbeam Weather Report) incorrectly narrates "Meta is up across the board" — the two directly contradict each other within the same post. Verified by directly viewing the attached table image, which is dated and fully legible. This resolves the corpus's open question ("resolve which week the quoted +18%/+14% refers to") — it does not appear to refer to any different week; it is simply a wrong re | The table figures are accurate for the week of 2026-06-29 to 2026-07-05 vs. the previous week: Meta CVR -8.18%, ROAS -2.39%, budget share 49.13% (share change -2.98pp); Google CVR -5.59%, ROAS +0.90%, share 17.21%; TikTok CVR +27.72%, ROAS -1.16%, share 5.58%; AppLovin CVR -2.82%, ROAS -7.46%, share 1.63%. The claim that the same post's text says "Meta is up across the board: conversion rate +18%, ROAS +14%, share of spend +6%" is CONFIRMED to be a real, verified self-contradiction: the poster's own attached table shows Meta DOWN on every one of those three metrics, not up. | Northbeam, "The Media Buyer Index" (table image, relayed by Ecomm Cowboy / Bryan Bumgardner) 2026-07-17 (post date; table covers week of 2026-06-29 to 2026-07-05) | Corrected Đã sửa lại | ↗ |
| Billo's own blog post states its Q4 Performance Calculator/benchmark is built on '80,069 sales-objective Meta video ads that ran between July and December 2025, spanning 14 industry verticals' -- the sample size and vertical count in the claim are exactly right, but the claim's framing as 'Q4 2025' is broader than the actual Oct-Dec quarter (it spans H2 2025). Minor but real correction: 'Q4 2025' should read 'July-December 2025 (H2 2025)' when this stat is cited, since the vendor's own wording defines the window that way. Nick Shackelford's disclosed #ad promotion of this tool is consistent with what's on Billo's si | 80,069 sales-objective Meta video ads across 14 verticals, but the underlying window is July-December 2025, not strictly 'Q4 2025' (Oct-Dec) | Billo (vendor blog) 2026 | Corrected Đã sửa lại | ↗ |
| Adobe Analytics (based on over 1 trillion visits to US retail sites, 100M+ SKUs, 18 categories) reports US Cyber Week 2025 online sales of $44.2 billion (+7.7% YoY), with Black Friday at $11.8 billion and Cyber Monday at a record $14.25 billion (+7.1% YoY). The dollar figures in the claim are exactly right and correctly labeled as 2025. Separately, publicly available TikTok Shop affiliate statistics describe affiliate/creator-driven sales as roughly 60-70% of TOTAL US TikTok Shop GMV (not a BFCM-specific figure, and not specifically defined as 'creators the brand does not employ'). The dollar totals are solid and should be cited as Adobe Analytics 2025 data (correcting the corpus's own worry that these needed 'pinning to a year and source' -- they are 2025, and they are Adobe's, not the Josh Snow post's own data). The 60% TikTok Shop fig | $44.2B Cyber Week (Thanksgiving-Cyber Monday, +7.7% YoY); $11.8B Black Friday; $14.25B Cyber Monday (+7.1% YoY) -- all confirmed for 2025. The 60% TikTok Shop creator-routed sales figure is NOT confirmed as a BFCM-specific or 'not employed by the brand' figure. | Adobe Analytics (2025 Cyber Week data); TikTok Shop affiliate stats aggregators (for the 60% figure, unconfirmed) December 2025 | Corrected Đã sửa lại | ↗ |
| Adobe's pre-holiday-2025 survey of more than 5,000 U.S. consumers (published Oct 8, 2025) found that when asked how they planned to use generative AI for holiday shopping, 53% cited research, 40% product recommendations, 36% finding deals, and 30% gift inspiration; the 72%/47%/43%/35% figures attributed to 'Adobe surveys' via NewsNation could not be traced to a fetchable Adobe report this session and do not match any Adobe survey wave found. This claim is a genuine correction, not just a window mismatch. The corpus's 72%/47%/43%/35% traces to NewsNation ('AI is transforming holiday shopping as consumers use it for deals and gift ideas'), which I could not fetch directly (HTTP 403). No other outlet | Adobe's own pre-season intent survey (5,000 US consumers, Oct 2025): 53% research; 40% product recommendations; 36% finding deals; 30% gift inspiration — NOT 72%/47%/43%/35% | Adobe (pre-season consumer intent survey, via Digital Commerce 360) 2025-10-08 | Corrected Đã sửa lại | ↗ |
| Sensor Tower's December 2025 analysis found Rufus made up 40% of Amazon mobile-app sessions on Black Friday 2025 (vs 30% on Nov 1), and Sensor Tower SVP Ian Simpson has separately said Rufus sessions convert roughly 3.5x non-Rufus sessions across 10 distinct shopping behavior patterns — but the widely-repeated '66% of Black Friday purchases' figure and '100,000-session panel' size do not appear in Sensor Tower's own published post and could not be traced to any Sensor Tower primary statement this session. Selection effect: Sensor Tower's own text says 'the jury is still out on how much of Rufus' contribution can be written off as a spurious correlation ... versus how much is truly determinative' -- heavy/high-intent shoppers may simply be more likely both to us | Sensor Tower's own Dec-2025 write-up says Rufus was 40% of Amazon APP sessions on Black Friday 2025 (up from 30% on Nov 1, easing to 38% by month-end) — it does not state a 66%-of-purchases figure or a 100,000-session panel size anywhere in its text. A Sensor Tower SVP (Ian Simpson), reported separately via a Feb-2026 newsletter interview, says Rufus sessions convert ~3.5x non-Rufus sessions across 10 distinct shopping patterns, with an explicit correlation-vs-causation caveat. | Sensor Tower blog (Lucy Greider, 'Retail Embraces AI: How Did Rufus Impact Amazon's November Sales?') + Retailgentic newsletter interview with Sensor Tower SVP Ian Simpson December 2025 (Sensor Tower blog); February 5, 2026 (Retailgentic) | Corrected Đã sửa lại | ↗ |
| Per Sensor Tower's own December 2025 post, Rufus's share of Amazon app sessions rose from 30% on Nov 1, 2025 to a 40% peak on Black Friday, easing to 38% by month's end -- Sensor Tower indexes its purchase-growth chart to Nov 1 (not Oct 1), and does not publish the '+90% Rufus vs +8% non-Rufus' session-growth figures that circulate online; those numbers, along with other inconsistent variants (+100%/+20%, +75%/+35%) found elsewhere, appear to be readers' estimates off an unlabeled chart rather than a number Sensor Tower stated in text. Three different secondary sources give three different Rufus-vs-non-Rufus growth-multiple pairs (+90%/+8%, +100%/+20% per TechCrunch quoting Sensor Tower, +75%/+35% per a viral X post) for what appears to be the same underlying chart -- strong evidence these a | Sensor Tower's Dec-2025 chart is indexed to Nov 1 (not Oct 1) as baseline: Rufus share of Amazon app sessions rose from 30% (Nov 1) to a 40% peak on Black Friday (a 33% relative increase), settling at 38% by month-end. The companion purchase chart is explicitly described as a relative/normalized comparison (Nov 1 = baseline) rather than absolute session counts, and Sensor Tower's own text states no '+90%'/'+8%' (or any other) percentage growth figures -- it only mentions a '91-point day-over-day increase' (in indexed points) in purchase-session activity on Black Friday itself. | Sensor Tower blog (Lucy Greider, Dec 2025) December 2025 | Corrected Đã sửa lại | ↗ |
| Amazon's own reported numbers, by quarter: Q3 2025 -- 250M Rufus customers, ~60% more likely to complete a purchase, ~$10B incremental annualized sales 'on track'; Q4 2025 -- 'more than 300 million' Rufus customers for the year and (per earnings coverage) nearly $12B in actual incremental annualized sales; Q2 2026, after Rufus's May-2026 rename to Alexa for Shopping -- 'more than 350 million' customers in the trailing 12 months and U.S. users spending 40% more per order. The '3.5x vs. a keyword search' conversion claim is not a Jassy/Rufus figure found in any of these calls; it traces instead to a separate AWS retailer-facing product launch. Conflict note in the task says a 'Buy For Me' feature, not Rufus, is the true source of the '60% more likely' figure -- this is NOT supported: Jassy's own Q3 2025 and Q4 2025 quotes both attribute the ~60% figure directly to Rufus ('Customers who used Rufus ar | Q3 2025 call (Oct 30, 2025): Jassy said Rufus had 250M active customers in 2025, monthly users +140% YoY, interactions +210% YoY, users '~60% more likely to complete a purchase,' and Rufus 'on track to deliver over $10 billion in incremental annualized sales.' Q4 2025 call (Feb 5, 2026): Jassy raised this to 'more than 300 million customers used Rufus' in 2025, reaffirmed the ~60%-more-likely figure, and outside reporting on that same call cites 'nearly $12 billion' in 2025 incremental annualized sales (i.e., the $10B was a Q3 projection later updated to an actual ~$12B figure, not two competing claims). Q2 2026 call (Aug 3, 2026), after Rufus was folded into/replaced by 'Alexa for Shopping' in May 2026: 'more than 350 million customers' used it in the trailing 12 months (not '300M' -- that was the prior, Rufus-era, full-2025 figure), active users 'nearly doubling' in Q2, interactions 'up over 5x YoY,' and US customers who use it 'spend 40% more per order on average' than non-users (not '>40%,' and it's a US-only comparison). No Amazon earnings-call transcript found states a '3.5x vs. keyword search' figure for Rufus/Alexa for Shopping; a 3.5x conversion claim does exist but attaches to a separate, B2B-facing AWS product ('Agentic Shopping Assistant' for retailers, launched ~May 2026), not to Jassy's consumer-facing Rufus/Alexa-for-Shopping commentary. | Amazon Q3 2025, Q4 2025 and Q2 2026 earnings call transcripts (Andy Jassy remarks); aboutamazon.com Q2 2026 recap; Retail Dive October 30, 2025; February 5, 2026; July 30 / August 3, 2026 | Corrected Đã sửa lại | ↗ |
| Per Brainlabs' analysis of 54 of its own agency clients (Jan 2025-Apr 2026, via Digiday, Aug 25, 2026), AI-platform referral sessions rose 163% to about 200,000/month on average while the clients' organic sessions fell 10.5% overall (140.1M to 125.4M), with a sharper single-month drop from ~20M to below 15M around September 2025 when Google's AI Overviews reached roughly 30% of U.S. search results pages. The task attributes this to 'an unattributed agency dataset cited by Adrian Luna, Webscale' -- I could not find this dataset on Webscale's blog or in 'The Chai' newsletter archive (checked webscale.com/blog and webscale.com/newsletters directly; no matching po | The figures are essentially accurate but the attribution in the task is wrong. This is Brainlabs' (a real, named marketing agency) study of 54 of its own clients (24 US, 24 UK, 6 other; 29 are large advertisers with 1,000+ staff), Jan 2025-Apr 2026: AI-platform referral sessions grew 163% over the period to 'roughly 200,000 a month' on average; organic sessions fell 10.5% overall (140.1M to 125.4M across the full 14 months) for 46 of 54 clients, and around the September-2025 inflection point (when Google AI Overviews reached at least 30% of U.S. search result pages), organic sessions for the cohort fell from about 20 million to below 15 million in a given month. AI referrals also converted about 1.5x higher than organic for this cohort. | Digiday (Sam Bradley), reporting a Brainlabs agency study August 25, 2026 (study period January 2025-April 2026) | Corrected Đã sửa lại | ↗ |
| Walmart CEO John Furner reported a rising Sparky order-value lift over two earnings calls in 2026: 35% higher average order value for Sparky users on the Q4 FY2026 call (Feb 19, 2026), growing to 40% more spend per order by the Q2 FY2027 call (Aug 24, 2026) as Sparky adoption grew roughly 70% year-over-year. The task's conflict note treats 40% vs 35% as two different, disputed figures possibly reflecting a grocery/repeat-purchase vs browse-and-discover split -- I found no evidence for that specific explanation. The simpler, verified explanation is that it's the sa | Both figures are real Walmart CEO John Furner quotes, but from two different, sequential earnings calls describing the same underlying metric (Sparky users' order-value lift vs non-users) -- not two disagreeing concurrent numbers, and not a 'Sparky spend' vs 'AI-influenced-session AOV' split as the task implies. Q4 FY2026 call (Feb 19, 2026): 'Shoppers using Sparky had a 35% higher order value than average shoppers.' Q2 FY2027 call (Aug 24, 2026): 'Customers using Sparky spend 40% more per order than those who don't,' with Sparky adoption up 70% YoY and half of app users having tried it by Q4 FY26. | CIO Dive / Robert Hu blog (Q4 FY26 call); Yahoo Finance / Benzinga (Q2 FY27 call), both quoting CEO John Furner February 19, 2026 (Q4 FY2026 call); August 24, 2026 (Q2 FY2027 call) | Corrected Đã sửa lại | ↗ |
| Similarweb's State of Ecommerce 2026 report (Sept 10, 2026) found that when consumers use AI in shopping research, they still use search alongside it 89% of the time, and that AI+search journeys convert best of any combination -- but the specific claim that AI-assisted journeys have grown from under 5% to over 11% of all journeys over two years is not substantiated in the press release itself and may be confusing a separate, unrelated Similarweb conversion-rate stat (ChatGPT referrals converting at 11.4% vs 5.3% for organic, from the prior year's 'State of Ecommerce 2025' report). Correctly flags in the task's own conflict field that this is 'much lower than survey-based use-AI figures (29-77%)' -- confirmed as a real methodological split: a separate March-2026 Similarweb blog post (by Shai Belinsky) cites a SURVEY figure of '35% of US | Confirmed directly from Similarweb's Sept-10-2026 'State of Ecommerce 2026' press release: 'consumers who use AI in their shopping research also use search' 89% of the time (matches 'search appears in 9 of 10 AI journeys'), and AI-recommended brands get 'in some cases... a 2-to-1 advantage over competitors' with complex (AI+search) journeys converting best. NOT confirmed: the specific '>11% of journeys now vs <5% two years ago' figure -- this could not be found in the Sept-2026 press release text I fetched. A widely-repeated '11.4%' figure that looks similar is actually a DIFFERENT metric from a DIFFERENT, earlier Similarweb report ('State of Ecommerce 2025', cited in an April-2026 Practical Ecommerce article): 'traffic to ecommerce sites from ChatGPT converted at roughly 11.4%, compared to 5.3% from organic search' -- a conversion-rate comparison, not a journey-share-over-time comparison. These two distinct '11%'-ish numbers appear to be getting conflated. | Similarweb Investor Relations press release ('State of Ecommerce 2026'); Practical Ecommerce (citing 'State of Ecommerce 2025') September 10, 2026 (State of Ecommerce 2026); April 19, 2026 article on State of Ecommerce 2025 | Corrected Đã sửa lại | ↗ |
| Five separate 2025-2026 holiday surveys report different AI-adoption figures because they ask different questions of different bases: PayPal/Talker Research's 77% and Zeta Global's 83% are both conditioned on prior AI use (past-year users and weekly users respectively, not all shoppers); Visa's 47% (Morning Consult, Oct 2025) is 'at least one shopping task' among all surveyed; KPMG's 15% (rising to 30% among 25-34-year-olds) comes from KPMG UK's Consumer Pulse survey of 3,000 UK consumers, not a US sample; McKinsey's holiday-AI figure (reported elsewhere as roughly 46%, i.e. 'about half') could not be confirmed against a McKinsey page fetched this session. Corrected on two points: (1) KPMG's 15%/30% figures are from a UK panel, not US -- mislabeling it under a 'US unless noted' claim is a geography error worth flagging in the report. (2) McKinsey's '~half intend to use AI for price checks/deal hunting' could not | PayPal/Talker Research: 77% of US adults who have used-or-considered AI for a purchase in the past year (n=1,000, fielded Sep 5-9 2025, released Oct 22 2025) plan to use AI as a shopping assistant this holiday season -- NOT 77% of all shoppers. Zeta Global: 83% of 2,000 US adults who use AI at least weekly (Oct 2025) say they'll let AI help pick gifts -- base is weekly AI users, not general holiday buyers. Visa: 47% of 1,000 US adults (Morning Consult, Oct 14-16 2025) have already used an AI tool for at least one shopping task. KPMG: 15% of 3,000 UK consumers (OnePoll for KPMG UK, fielded Sep 15-22 2025, published Oct 29 2025) will use AI chatbots to find Black Friday deals, rising to 30% of 25-34-year-olds -- this is a UK survey, not US as the claim's 'US unless noted' framing implies. McKinsey: could not confirm a specific 'price checks/deal hunting' figure via a fetched primary source; secondary coverage (not independently fetched this session) puts overall 2026 holiday AI-use intent at 46% ('probably or definitely' use AI tools), among whom top uses are comparing products/prices, finding deals and gift ideas -- 46% rounds to '~half' but is not a figure specific to price-checking alone. | PayPal/Talker Research; Zeta Global; Visa/Morning Consult; KPMG UK; McKinsey (unconfirmed) 2025-10-22 (PayPal); 2025-10-28 (Zeta); 2025-12-02 (Visa press release, survey Oct 14-16 2025); 2025-10-29 (KPMG UK) | Corrected Đã sửa lại | ↗ |
| "nearly 9 in 10 Reddit users verify AI product recommendations with real people on Reddit before they buy." Methodology on the same page: Attest panel, n=1,000, US, A18+ monthly social media users, fielded February 2026, comparing Reddit against Facebook, Instagram, X, TikTok, Pinterest, Discord, Snapchat, YouTube. WebFetch could not load business.reddit.com directly (blocked, same as first pass), but a reader-proxy fetch (r.jina.ai) of the same URL this session successfully retrieved the live page text, which contains the '9 in 10' figure verbatim plus full methodology. | Nearly 9 in 10 Reddit users verify AI product recommendations with real people on Reddit before buying -- CONFIRMED verbatim. Sample size is n=1,000 (not n=1,873 as the corpus states). The 'nearly 7 in 10 trust Reddit gifting advice over AI' and '3 in 4 willing to pay more for Redditor-recommended gifts' figures could NOT be found on the source page and are unconfirmed. | Reddit for Business, "Moments" / Holiday Insights page (business.reddit.com) 2026-02 (survey fielded); page appears to be Reddit's ongoing 2026 holiday marketing hub | Corrected Đã sửa lại | ↗ |
| The figures attributed to Peec AI (43.2% site: operator, 26.5% year mentions, 21.9% 'official', 5.4% 'best/top', 3.0% pricing/cost) come from a July 22, 2026 analysis by Peec AI researcher David Konitzny of ChatGPT 5.6's query fan-outs in general -- not a study isolated to purchase/buying questions -- based on a single account in Dubai sampled that month and weighted toward software and AI-tool queries, a scope and sample size the corpus's 'sample size and period not disclosed' note undersells (the period is disclosed; the narrow, single-account, vertical-skewed sample is the bigger caveat). Could not locate or fetch the original Peec AI blog post/LinkedIn post containing this exact breakdown directly (Peec's own 'Patterns we see in ChatGPT query fanouts' post, fetched directly, has different numbers -- e.g. 'best' at 24.3%, year mentions at 5.44% | These exact percentages (43.2% site:, 26.5% year, 21.9% 'official', 5.4% best/top, 3.0% pricing/cost) trace to a Peec AI analysis by researcher David Konitzny, published ~July 22, 2026, of ChatGPT 5.6 query fan-outs -- but that study covers ChatGPT fan-outs GENERALLY (all topics), not specifically 'buying questions', and was run on a single account based in Dubai, sampled in July 2026, weighted toward software/AI-tools queries. | Peec AI (David Konitzny), via secondary citation on Buffy (heybuffy.com) 2026-07-22 | Corrected Đã sửa lại | ↗ |
| OpenAI launched Instant Checkout in ChatGPT on September 29, 2025 with Etsy and (soon after) Shopify merchants, charging a 4% transaction fee on completed sales; by March 2026 (OpenAI's own reversal reported March 5-6, 2026, with its official pivot post 'Powering Product Discovery in ChatGPT' published March 24, 2026) OpenAI wound the feature down after fewer than ~30 Shopify merchants ever went live, and Walmart EVP Daniel Danker said checkout completed inside ChatGPT converted at about one-third the rate of sending the same shoppers to Walmart's own site to finish the purchase. CNBC's own articles (cnbc.com/2026/03/20 and /03/24) returned 403 to direct fetch, as did openai.com/index pages and help.openai.com -- verified via WebSearch snippets that reproduce exact CNBC/OpenAI wording plus one successful direct fetch of OpenAI's still- | OpenAI launched Instant Checkout Sept 29, 2025 (Etsy first, Shopify merchants following) with a 4% merchant transaction fee; it officially wound the feature down in March 2026 (reporting broke ~Mar 5-6; OpenAI's own pivot post published Mar 24, 2026), replacing it with a 'product discovery' experience that routes purchases to merchants' own checkout/apps. Walmart's EVP said checkout inside ChatGPT converted at roughly 1/3 the rate of sending shoppers to Walmart's own site. | OpenAI (developers.openai.com Agentic Checkout Spec, fetched directly) + CNBC (Mar 20 and Mar 24, 2026 reporting, via search-verified snippets) + PYMNTS (4% fee) 2026-03-24 | Corrected Đã sửa lại | ↗ |
| Accenture's Consumer Pulse Research 2026 report 'Talk to my AI agent: the new rules of brand value' (surveying 25,590 consumers across 16 countries) states that 26% of active generative-AI users have already bought a MORE expensive item because AI increased their confidence in the purchase, and that the same 26% saw their overall basket size increase -- the opposite direction from the 'traded down to cheaper items' framing given in a Supermarket News piece bylined by Accenture's own Global Retail Industry Lead, Kelly Askew (published Sept 4, 2026), which separately cites 57% weekly AI usage and 68% comfort with AI agents on commerce tasks that do not appear verbatim in the public PDF and could not be independently confirmed. This is a genuine, load-bearing discrepancy, not a rounding issue: the Accenture PDF's own text (verified by extracting the actual PDF, not just search snippets) says active genAI users traded UP (bought a MORE expensive item) and that the SAME 26% cohort saw | Accenture's 'Talk to my AI agent' Consumer Pulse Research 2026 (25,590 consumers, 16 countries -- confirmed by direct PDF text extraction) states 26% of active genAI users have already bought a MORE expensive item because AI increased their purchase confidence, and the SAME 26% (not a separate 21%) saw their basket size increase; 61% want an agent that shops across multiple grocery retailers on their behalf (this one matches the corpus exactly). A companion Supermarket News piece bylined by an Accenture exec (not the PDF itself) states the opposite direction -- '26% ... a LESS expensive grocery item, and another 21% ... increased basket size' -- plus '57% weekly AI users' and '68% comfortable with AI agents on commerce tasks,' neither of which appears in the public PDF. | Accenture, 'Talk to my AI agent: The new rules of brand value' (Consumer Pulse Research 2026) 2026 | Corrected Đã sửa lại | ↗ |
| In E-Commerce Bench (Qwen team, arXiv 2608.30730), a deterministic SIMULATED marketplace where 18 frontier models each ran five independent 365-day store episodes, 10 of the 90 episodes ended in bankruptcy, and across 8,647 repeat purchases from the same honest supplier 16 of the 18 models showed no clear sign of pushing the purchase price down over the year. Verified directly from the paper HTML on 2026-09-14 after two agents disagreed (15 vs 16 of 18). Paper says 16. Also: "Four of the eighteen models slip into bankruptcy on part of their runs." Do not cite the live GitHub leaderboard (21 models/105 episodes) as | 10 of 90 episodes ended in bankruptcy (18 models x 5 episodes); the top-earning model ranked 16th of 18 on fraudulent-supplier spend; across 8,647 repeat purchases 16 of 18 models showed no sign of bargaining the price down over the year | E-Commerce Bench (Qwen team), arXiv:2608.30730v1 2026-08-31 | Corrected Đã sửa lại | ↗ |
| Adobe itself only characterizes AI's traffic contribution qualitatively, saying 'the base of users remains modest' even as YoY growth is triple-digit; no Adobe release states AI's percentage share of total retail traffic. Separately, Apptopia/TechCrunch data shows that only 0.82% of ChatGPT sessions on Black Friday 2025 referred users onward to a retailer's app — the inverse population from 'what share of retail traffic is AI-driven,' and not a like-for-like number. The claim as written ('less than 1%... share of total retail-site traffic coming from AI') is directionally plausible given how Adobe describes the base as 'modest', but is unsourced and conflates two different populations if pointed at the Apptopia number: 0. | No primary source states AI's share of TOTAL retail-site traffic on Black Friday 2025 as a clean percentage. A distinct, real metric exists but measures something different: ChatGPT referrals to e-commerce apps were 0.82% of ALL ChatGPT sessions on Black Friday 2025 (up from 0.64% in 2024), per mobile-analytics firm Apptopia, as reported by TechCrunch. | Adobe (qualitative statement); Apptopia via TechCrunch (0.82% ChatGPT-session figure) 2025-12-02 | Needs context Cần ngữ cảnh | ↗ |
| Salesforce reported that AI and digital agents 'influenced' $14.2 billion of Black Friday 2025's global online sales (about $3 billion of it in the U.S.), and $7.5 billion of Thanksgiving 2025's global online sales ($1.5 billion in the U.S.) — a combined roughly $21.7 billion across the two days. Salesforce defines 'influenced' as sales where AI played a role via personalized recommendations or conversational assistance, not sales autonomously transacted by an AI agent. Could not locate a single Salesforce page in this session that states the Black Friday/Thanksgiving day-specific $14.2B/$3B and $7.5B/$1.5B figures verbatim in one place — Salesforce's Dec 5, 2025 Cyber Week press release (fetched directly) discusses the WEEK | Salesforce: AI and agents 'influenced' $14.2B in global online sales on Black Friday 2025 ($3B of it in the US); $7.5B globally on Thanksgiving 2025 ($1.5B US). Combined Thanksgiving+Black Friday ≈ $21.7B globally (rounds to the claimed ~$22B). | Salesforce (Salesforce Shopping Index / Cyber Week data), corroborated by CNN Business, Digital Commerce 360, and PaymentsJournal citing Salesforce 2025-12-05 | Needs context Cần ngữ cảnh | ↗ |
| Northbeam's Cyber Week 2025 recap reports that across its customer base, ad spend rose just over 9% year over year, revenue grew more than 13%, first-time CAC climbed about 8%, and MER improved as a result -- but the article does not give a specific MER percentage, and none of the figures are given to two-decimal precision. Fetched the full text of Northbeam's own Dec 22, 2025 blog post (its primary Cyber Week 2025 recap). Directionally the claim matches: spend +9.23% rounds to 'just over 9%', revenue +13.35% rounds to 'more than 13%', and CAC +8.03% rounds to 'about 8%' -- consi | Northbeam's own recap (Cyber Week 2025, across its customer base, YoY): ad spend 'rose by just over 9%'; revenue grew 'more than 13%'; MER 'improved' (no percentage given); first-time CAC 'climbed about 8%' | Northbeam 2025-12-22 | Needs context Cần ngữ cảnh | ↗ |
| AI-adoption figures for holiday shopping range from 13% to 77% across Bain, KPMG, Omnisend, PwC, McKinsey, Klaviyo, Narvar, and PayPal/Talker Research surveys because each asks a different question (delegating to a retailer AI agent vs. starting discovery on a GenAI platform vs. using AI for 'at least one' task vs. broad 'planning to use AI') on different populations (KPMG's 15%/30% figure is UK, not US) — no single number should be quoted as 'the' AI-adoption rate without naming the surveyor and the exact question asked. This is a compilation stat; the conflicts field's own caution ('do not quote a single AI adoption number without the wording') is correct and should stay in the report. Two line items in the original list ('nearly half used or plan to use, unnamed news report' | Each figure is real but answers a different question: Bain 13% plan to use retailer-specific AI shopping agents (2026); KPMG 15% plan to use AI chatbots to find Black Friday deals — this is a UK survey (n=3,000), not US, climbing to 30% among ages 25-34; Omnisend's ~25% could not be pinned to one exact figure (closest confirmed Omnisend numbers are 81% who'll use AI for 'at least one' holiday task, and 28% of US shoppers who used AI for gift ideas in the last 6 months, or 19% who use chatbots specifically to hunt/compare deals — no exact '~25%' line item found); Bain 24% plan to start discovery on AI platforms (2026, up from 17% in 2025); PwC 29% plan to use AI (up from 22%); McKinsey 46% probably/definitely will use AI tools; Klaviyo 56% of US consumers plan to use AI shopping assistants for BFCM 2025 (global study, 7,000 consumers); Narvar 65% plan to use AI for at least one part of holiday shopping (2026, n=1,348 US); PayPal/Talker Research 77% planning to use AI for holiday shopping (2025, n=1,000 US, question wording covers 'have used or considered using AI... within the last year'). | Bain & Company; KPMG UK; Omnisend; PwC; McKinsey; Klaviyo; Narvar; PayPal/Talker Research 2025-2026 (varies by survey, see each id) | Needs context Cần ngữ cảnh | ↗ |
| Cordial's Holiday 2026 Consumer Research (n=1,000 US consumers) found free shipping (61%) leads bigger discounts (45%) as the top factor in where consumers choose to buy this holiday season; a separate generational free-shipping-importance breakdown (reportedly 75% of Boomers vs 44% of Gen Z) could not be located in the same or any other Cordial publication. The 61% vs 45% figure is confirmed verbatim from Cordial's own page. The 75%-Boomers-to-44%-Gen-Z generational drop-off was searched for directly (including generation-specific query terms) and not found anywhere in the Cordial report content that was fetched, | 61% of consumers cite free shipping as a top factor in where they buy this holiday season vs 45% for a bigger discount — CONFIRMED. The generational breakdown (75% Boomers to 44% Gen Z) could NOT be confirmed from any fetched source. | Cordial, Holiday 2026 Consumer Research Data 2026 (exact publish/fielding date not stated on the page) | Needs context Cần ngữ cảnh | ↗ |
| Klaviyo's 2025 BFCM Recap Report found that average discount depth across its customer base fell from 29.1% in 2024 to 26.2% in 2025, a 10% year-over-year decline, even as overall demand and revenue for Klaviyo brands grew. Confirmed exact figures directly from Klaviyo's own page: 29.1% (2024) -> 26.2% (2025) = a 10% relative drop, resolving the corpus's ambiguity ('unclear if 10% is relative or percentage points') — it is relative, not points (29.1-26.2=2.9 points, which is ~10% | Average BFCM 2025 discount across Klaviyo-brand ecommerce sends: 26.2%, down from 29.1% in 2024 — a 10% relative (percentage) drop, not a 10-percentage-point drop. | Klaviyo, 2025 BFCM Recap Report 2025-12-02 | Needs context Cần ngữ cảnh | ↗ |
| A UK-market Meta marketing summit (an invite-only Manchester event covering 2025 holiday strategy, corroborated independently via a UK agency's recap) recommended allocating 20%+ of ad budget to Partnership Ads and noted AI now recommends up to 70% of all Meta content — but the additional claims that Q5 (Dec 26–mid-Jan) delivers +20-30% incremental revenue and that 81% of users don't click ads during peak could not be confirmed from any public source and should not be printed without that caveat. Treat as UK/EMEA-specific guidance, not a global stat — matches the corpus's own flag that the underlying data may be UK/EMEA 2023-24, not current/global. Do not print '81% of users don't click ads' or '+20-30% Q5 revenue' as confirmed Meta figures — no public | 20%+ of Meta ad budget recommended for Partnership Ads and 70%+ of content now AI-recommended are corroborated for a UK Meta 'summit' event covering 2025 peak-season strategy. The +20-30% Q5 incremental-revenue figure and the '81% don't click ads during peak' figure could not be independently verified — the corpus's own notes say these came from a photographed, partly-unreadable presentation slide. | Launch (UK agency) recap of a Meta peak-season summit, plus general Q5/'Fake Friday' industry usage 2026 | Needs context Cần ngữ cảnh | ↗ |
| Google's August 2026 holiday retail guide, drawing on an Ipsos survey, states that 94% of US holiday shoppers who used AI last season found it valuable and that AI-journey shoppers engage with 2.8x more touchpoints than non-AI shoppers (per a separate Ipsos Global Consumer Journeys study of ~13,189 shoppers in 25 countries, Dec 2025); the guide also cites a TransUnion marketing-mix-model meta-analysis (40 studies, 800+ models, Q1 2024–Q4 2025) finding that running Google Search and YouTube together drove 37% higher ROAS than all other media combined. I could not access the Google PDF guide directly (it was emailed to advertisers, not posted publicly); ppc.land's article quotes it extensively verbatim and I treat that as sufficiently reliable trade-press reporting of a real document, but flag that as a limi | 94% of US holiday shoppers who used AI last season rated it valuable (Google/Ipsos survey, cited in Google's Aug 21 2026 'AI-powered holiday season 2026' retail guide — this is 94% of AI-users, not of all shoppers). AI-journey shoppers interact with 2.8x more touchpoints than non-AI shoppers (Ipsos Global Consumer Journeys survey, ~13,189 online shoppers across 25 countries, fielded Dec 2025 — flagged as possibly reflecting touchpoints within Google's own surfaces rather than the full cross-retailer journey). Running Google Search + YouTube together drove 37% higher ROAS than all other media in aggregate, per TransUnion marketing-mix-model meta-analysis (40 studies, 800+ models, Q1 2024–Q4 2025) cited in the same Google guide. | Google — 'Supercharge your business this season: the retail marketers' guide to the AI-powered holiday season 2026' (distributed to advertisers Aug 21, 2026), as reported by ppc.land 2026-08-21 | Needs context Cần ngữ cảnh | ↗ |
| Josh Fletcher's 'at least double the volume' rule for a 30-40% sitewide discount only holds exactly at the low end of that range (a 60%-gross-margin business needs 2x volume at 30% off) — at 40% off, the same business needs roughly 3x volume, not 2x, so the single 'double, minimum' figure understates what's needed toward the top of the stated discount range. The corpus's own 'conflicts' note already flags margin-dependence ('holds for ~60% GM at 30% off') — my recomputation confirms that caveat is necessary, not optional: at 60% GM, 30% off needs exactly 2x volume, but 40% off (the other end of the '30-40%' range | Doubling order volume exactly offsets a blanket discount ONLY at specific margin/discount combinations. At a 60% gross margin, a 30% sitewide discount requires exactly 2x volume to hold baseline gross profit — but at the same 60% margin, a 40% discount requires 3x volume, not 2x. The general rule is: volume multiple needed = original margin % ÷ (original margin % − discount %). | Direct recomputation of gross-profit breakeven math under a sitewide discount 2026-09-14 | Needs context Cần ngữ cảnh | — |
| Klaviyo reports that across its customer base, email and text drove 42% of total revenue during BFCM 2025, rising to 43% on peak days — a platform-wide average of Klaviyo Attributed Value (KAV) as a share of total revenue, not an incremental or GMV-independent figure. The 42% is Klaviyo's own attributed-revenue metric (KAV/total revenue), not incremental lift vs a no-email counterfactual — exactly the ambiguity the task flagged. It is a platform-wide average; individual-brand dashboards in the corpus (20.43%, 25.2%, 54.90%) | 42% of total revenue (43% on peak days) | Klaviyo BFCM 2025 Recap (press release) 2025-12-02 | Needs context Cần ngữ cảnh | ↗ |
| Klaviyo reports that revenue generated by messages containing AI-powered product recommendations rose 71% year over year during BFCM 2025, while usage of AI-driven recommendations across customers grew 45% YoY. The 71% is YEAR-OVER-YEAR GROWTH in revenue from AI-recommendation-powered messages, not a 'lift' versus a non-AI control group within the same period — the baseline ambiguity the task flagged is real. Also confirms the task's warning about mix-up risk: a sepa | +71% YoY revenue from AI-driven product recommendation messages (usage of the feature itself grew +45% YoY) | Klaviyo BFCM 2025 Recap (press release) 2025-12-02 | Needs context Cần ngữ cảnh | ↗ |
| Omnisend reports that during BFCM week 2024 (Nov 24-Dec 2), its merchants sent 1.13 billion emails (27.94% open rate, 8.22% click-to-conversion rate) and 19.1 million SMS messages (3.64% click-to-conversion rate). Numbers confirmed across multiple independent secondary quotations, but I could not load Omnisend's original page directly to pin the exact definition of 'click-to-conversion rate' for this specific report (the canonical-looking URL omnisend.com/blog/bfcm-emai | Email: 1.13B sent, 27.94% open rate, 8.22% click-to-conversion rate; SMS: 19.1M sent, 3.64% click-to-conversion rate; (Push: 15.2M sent, 4.66% click-to-conversion, not in original claim) | Omnisend BFCM/holiday email, SMS & push statistics report 2024-2025 (BFCM 2024 period: Nov 24-Dec 2, 2024) | Needs context Cần ngữ cảnh | ↗ |
| Omnisend's BFCM data shows the welcome automation had the highest click-to-conversion rate during BFCM at 63.66%, with abandoned cart second at 44.57% — click-to-conversion rate here means the share of people who CLICKED the message who went on to purchase, not the share of all recipients or all sends who purchased. This resolves the red flag the task raised: 63.66%/44.57% are implausible as an order-per-recipient or order-per-send conversion rate, but are plausible as a click-to-conversion rate (orders as a % of clicks only) — highly engaged clickers on a welcome or cart | Welcome automation: 63.66% click-to-conversion rate (highest of any automation); abandoned cart automation: 44.57% click-to-conversion rate (second highest) | Omnisend BFCM/holiday email, SMS & push statistics report 2024-2025 (BFCM 2024 period) | Needs context Cần ngữ cảnh | ↗ |
| Chronos Agency's BFCM 2025 playbook sets internal deliverability guardrails of spam complaints under 0.01%, unsubscribes under 0.3% and bounces under 1% per send, tripping any of which triggers suppression of non-engagers — considerably stricter than Gmail's own published bulk-sender threshold of a 0.3% spam-complaint rate. The exact numbers in the claim (spam <0.01%, unsub <0.3%, bounce <1%) are verbatim from Chronos's own blog post, so the extraction is accurate to its source. But this is a self-reported agency internal target, not a platform standard — Google's and Yahoo's act | Chronos Agency's own internal BFCM guardrails are: spam complaints <0.01% (target; Postmaster Tools "ideal" is <0.1%), unsubscribes <0.3%, bounces <1%; the article separately notes Gmail's actual per-send rule is complaint rate <0.3%. | Chronos Agency blog (self-reported agency benchmark); Google Workspace Admin Help (bulk sender guidelines) for the platform comparison 2025 (Chronos article, undated exact day); Google guidelines effective Feb 2024, updated periodically | Needs context Cần ngữ cảnh | ↗ |
| Contentsquare's customer case study reports that after Lovehoney moved its three most-viewed categories out of a filter dropdown and onto its Black Friday landing pages (a change made within days of the pages going live on Nov 19), on-page conversions rose 30%, exit rate fell 20%, and bounce rate fell 17%. Numbers match the claim exactly and are directly attributable to Contentsquare's own site (not blog-to-blog). However, this is a vendor-published case study promoting Contentsquare's own product, with no independent audit, no disclosed statistical significance | +30% on-page conversions, -20% exit rate, -17% bounce rate — confirmed verbatim on Contentsquare's own published Lovehoney case study. | Contentsquare (vendor case study) undated (references a Nov 19 Black Friday page launch) | Needs context Cần ngữ cảnh | ↗ |
| AfterShip reports, via Digital Commerce 360's coverage of Dr. Squatch, a 31.89% click-through rate on tracking-page product recommendations generating roughly $32,978 in attributed revenue in Q1 2026, a 94% on-time delivery-date accuracy rate across 99.98% of orders using its AI prediction model, a 25% reduction in 'where is my order' tickets, exception-resolution times cut by more than 58%, and — the only figure tied specifically to BFCM — a 62.69% month-over-month jump in Cyber 5 2025 shipment volume handled at close to 100% completion. Numbers are consistent across two independent trade-press write-ups (Digital Commerce 360 and itbrief.co.uk), which is stronger corroboration than a single vendor blog, but both trace back to the same AfterShip press briefing — this is still fundamentally a ve | Confirmed via Digital Commerce 360 and corroborating trade coverage: 31.89% tracking-page recommendation CTR, ~$32,978 attributed revenue (Q1 2026, ~766 orders), 94% AI delivery-date accuracy across 99.98% of orders, 25% fewer WISMO tickets, 58% faster exception resolution, and Cyber 5 2025 shipment volume +62.69% MoM with ~99.8-100% delivery completion. | Digital Commerce 360 (trade press) reporting AfterShip's (vendor) self-reported customer metrics; corroborated by itbrief.co.uk coverage of the same AfterShip Intelligence launch 2026-09-02 | Needs context Cần ngữ cảnh | ↗ |
| Taylor Holiday (CEO, Common Thread Collective) reported, from an unspecified sample of CTC's own client portfolio, that average Meta attribution-window value capture is ~85% within 7 days / 15% over 28 days at baseline, rising 47% in delayed (28-day) share in the two weeks before Black Friday, then falling 33% by Cyber Monday. Matches the claim exactly, including the 33% Cyber Monday figure the corpus flagged as unconfirmed elsewhere in the cluster. This is an agency's own client-base analysis, not an audited or peer-reviewed study — no sample size, vertical mix, or explicit year is | Baseline ~85% of Meta-attributed value captured within 7-day click / ~15% over the remaining 28-day window; in the two weeks before Black Friday, 28-day (delayed) value share rises 47%; by Cyber Monday, 28-day contribution drops 33% vs. the pre-BF peak. | Taylor Holiday (X/Twitter post) 2025-11-07 | Needs context Cần ngữ cảnh | ↗ |
| Olivia Kory, Haus's Chief Strategy/Marketing Officer, stated ahead of a Sept 2026 AppLovin/Ridge/HexClad masterclass that across Haus's ~5 years of BFCM incrementality testing, delayed lift exceeded immediate lift in 73% of experiments, video campaigns drove 286% more delayed lift than search, and delayed lift rises 79% during Q4 versus the same channels' evergreen-period tests. Exact match to the claim, sourced directly to Haus's own CSO rather than a byline-anonymous compilation. Haus sells incrementality/geo-lift measurement, so it has a direct commercial interest in publicizing findings that platform/click attribution undercounts | Delayed lift exceeded immediate lift in 73% of Haus experiments; video (YouTube/CTV) drove 286% more delayed lift than search; delayed lift increases 79% during Q4 vs. evergreen periods. | Olivia Kory (X/Twitter post, Haus CSO) 2026-09-10 | Needs context Cần ngữ cảnh | ↗ |
| Olivia Kory (Haus) wrote that experiments leading up to BFCM "on average drive over 2x the value after the treatment ends," implying daily efficiency targets could be cut roughly in half during the leadup, and that video/view-based channels (YouTube, CTV) see the most delayed lift, social channels (Pinterest/TikTok/Meta) cluster around 1.2-1.23x, and Search/PMax/AppLovin see less of a gap at 1.06-1.12x. Matches the claim, including the "drop daily efficiency targets by around half or more" framing the corpus correctly identified as a derived recommendation rather than a separate measurement — confirmed it is Kory's own phrasing, in the same breath as the 2x f | Haus BFCM-leadup experiments drove on average over 2x their measured value after the treatment window ended; by channel, Pinterest/TikTok/Meta cluster at 1.2-1.23x post-treatment value while Google Search/PMax and AppLovin sit at 1.06-1.12x. | Olivia Kory (X/Twitter post, Haus CSO) 2026-09-12 | Needs context Cần ngữ cảnh | ↗ |
| Haus's July 2025 report on 640 of its Meta incrementality experiments since 2024 found Advantage+ over-reports revenue by 12 percentage points versus Manual campaigns, with Manual campaigns cannibalizing less existing organic demand and delivering higher incremental ROAS in 58% of head-to-head brand tests. Confirmed exactly via a directly-fetched, methodology-disclosing Haus blog post (the most rigorously documented source in this cluster) — this is the strongest-sourced stat in the batch. Flag regardless: Haus is a vendor that sells incrementality/geo-lift test | Meta Advantage+ campaigns over-report revenue by 12 percentage points relative to Manual campaigns (i.e., for every $100 of platform-reported revenue, Manual delivers ~$12 more real revenue than Advantage+, on average); manual campaigns also showed higher post-treatment-window lift (+32% vs. +17%) and 58% of brands saw higher incremental ROAS on Manual head-to-head. | Haus, "The Meta Report: Lessons from 640 Haus Incrementality Experiments" 2025-07-28 | Needs context Cần ngữ cảnh | ↗ |
| Triple Whale co-founder Maxx Blank published a report analyzing 12 months of data across a 755-shop AppLovin cohort using three measurement methods (multi-touch attribution, a proprietary Foundation MMM, and real geo holdout tests), finding AppLovin outperformed the top ad-channel benchmark on all three lenses while still representing only 7.7% of the cohort's ad spend. Confirmed exactly by directly fetching the actual report page this session (via a text-extraction proxy, since triplewhale.com blocks direct fetch/curl with a 403; the resolved link came from the corpus's own captured tweet https://x.com/aryehMaxx/status/20959 | Across 755 AppLovin-spending shops over 12 months: 2.90 lifetime Triple Attribution (last-click) ROAS for AppLovin vs. 2.08 for an established social-advertising benchmark (61% of qualifying shops higher); Foundation MMM (trained on ~6,300 shops) found 66% of AppLovin advertisers had a higher expected incremental return per dollar than their own benchmark on other platforms; 5 of 7 geo holdout tests were statistically significant, all five positive, averaging +8.3% revenue lift (range +3.5% to +13.5%); AppLovin was 7.7% of the cohort's combined ad spend. | Triple Whale, "The AppLovin Ads Report" (blog, by Maxx Blank) 2026-09-03 (last updated 2026-09-11) | Needs context Cần ngữ cảnh | ↗ |
| Northbeam's Brayden Cruz reported that across Northbeam's own customer base, Cyber Week 2025 ad spend rose just over 9% year over year, revenue grew more than 13%, MER improved even as first-time customer CAC climbed about 8%. Directionally and materially confirmed via direct fetch of the Northbeam blog post; the rounded language ("just over 9%", "more than 13%", "about 8%") matches the claim's more precise decimals closely enough to treat them as the same underlying dataset, though | Cyber Week 2025 YoY across Northbeam's customer base: ad spend "just over 9%" higher, revenue "more than 13%" higher, MER improved, first-time customer CAC up "about 8%." | Northbeam blog, "Cyber Week 2025: A Data-Driven Look at Performance and Strategy" (Brayden Cruz) 2025-12-22 | Needs context Cần ngữ cảnh | ↗ |
| Bryant Garvin (Operator in Residence, Triple Whale) published a Meta ad-spend heatmap covering Nov 28-Dec 2, 2024, built from an analysis of $585.8M in Meta spend, showing Thanksgiving as the lowest-spend day and hourly intensity peaking in the early-to-mid afternoon on Cyber Monday, on a color scale running roughly $2M-$10M per hour. Verified by directly viewing the chart image itself (not just the tweet text): the described pattern (Thanksgiving lowest, scale ~$2M-$10M, peak concentrated midday on Cyber Monday specifically, slightly more than Black Friday/Sat/Sun) matches the claim well, | Chart-confirmed: Thanksgiving (Nov 28) is the lowest-intensity day on the heatmap; intensity rises sharply from Nov 29 onward and the darkest (highest-spend) cells appear around midday on Dec 1 (Sun) and especially Dec 2 (Mon, Cyber Monday); the chart's color scale runs from about $2.0M/hour (darkest blue/purple) to $10.0M/hour (darkest red). | Bryant Garvin (X/Twitter post + attached heatmap image) 2025-11-26 | Needs context Cần ngữ cảnh | ↗ |
| Incrementality consultant Shishir Acharya described a single, unnamed brand's BFCM Meta campaign ($5M spend, 4.5x platform-reported ROAS) where a geo-lift test measured 2.7x incremental ROAS, implying about $9M of reported sales were not actually incremental. Confirmed to be an accurate quote of the primary post — the numbers as stated are exactly what the corpus attributed. But this is a single anonymous-brand anecdote from a consultant with an interest in selling incrementality-testing services (the entire post's | As posted: one unnamed brand spent $5M on Meta for BFCM (year unspecified in the post itself, but posted Aug 2026, so likely referencing BFCM 2025), platform reported 4.5x ROAS, a geo-lift test measured 2.7x incremental ROAS, implying roughly $9M of platform-credited sales would have happened anyway. | Shishir Acharya (LinkedIn post) 2026-08-17 | Needs context Cần ngữ cảnh | ↗ |
| Triple Whale co-founder Maxx Blank posted that the BFCM ad-spend map shifted: Meta and Google both slipped slightly in spend share (67.81%→67.6% and 23.14%→22.65% respectively) while TikTok and AppLovin absorbed the difference with 24% and 36% year-over-year increases in spend share. Verified exactly via the corpus's own screenshot of the tweet. No methodology, cohort size, or time window ("BFCM" vs. full year) is disclosed in this post, and it is Triple Whale's own book of business, not an industry-wide census — a vendor with a direct int | Confirmed as posted: Meta's share of ad spend moved 67.81% → 67.6%, Google's moved 23.14% → 22.65%, while TikTok's share grew 24% YoY and AppLovin's grew 36% YoY off much smaller bases. | Maxx Blank / Triple Whale (X/Twitter post) 2025-12-03 | Needs context Cần ngữ cảnh | ↗ |
| Media buyer Phil Kiel posted (and shared a dashboard screenshot showing) that one Meta ad account reached 11.1M unique people year-to-date in 2025 with 29.8% net-new / 70.2% previously-served, that the week of Dec 8 alone reached 632.2K people (171.5K / 27.1% net-new) on $25,442 spend, and that DPA carousel ads ran 17% net-new vs. 27% net-new for video in one specific week. Fully confirmed by directly viewing the dashboard image itself, not just the tweet text — every number in the claim (11.1M/29.8%/70.2%, the Dec 8 week's 171.5K/27.1%/632.2K/$25,442) is visible verbatim on the chart. This is exactly the 'single-brand dashboard | Confirmed via the actual dashboard screenshot: 11.1M total cumulative unique people reached YTD 2025, of which 29.8% (11.1M... net-new label matches total, a known dashboard quirk) were net-new and 70.2% (26.1M) previously served; for the week of Dec 8, 2025: 171.5K net-new (27.1%) of 632.2K total period reach, previously served 460.7K (72.9%), spend $25,442. Separately, by asset type in one week: DPA carousel 17% net-new on ~$10k spend, video 27% net-new on ~$30k spend. | Phil Kiel (X/Twitter post + attached dashboard screenshot) 2025-12-15 | Needs context Cần ngữ cảnh | ↗ |
| Could not locate Dara Denney's original September 2025 YouTube video recommending 10-25 ads per ad set. A different, independently-found summary of Dara Denney's Meta ads guidance describes a much smaller structure: '2-3 ad sets maximum per campaign with 5-6 ads per ad set.' This is a real conflict, not just the one already noted in the corpus (li#217): one secondary source attributes 5-6 ads per ad set to Dara Denney, directly contradicting the 10-25 figure attributed to her via Jake Abrams' summary. Either her guidance is contex | 10-25 ads per ad set (images 20-25, videos 10-15) vs 5 previously | withbaker.com (secondary summary of Dara Denney's guidance; original video not located) 2025/2026 | Needs context Cần ngữ cảnh | ↗ |
| Motion (Creative Analytics) and HexClad have a real, publicly documented partnership -- a Motion case study states HexClad 'increased top-of-funnel spend by 60%' using Motion's 5-step creative testing workflow (14-day test-and-report cycle) -- but the specific figures in the claim (75 to 25 assets, add-to-cart +56%, conversion +23%, time-on-page 1.5 to 3.5 minutes, tied to a Mother's Day sale) were not found in the accessible Motion case-study content. The 60%-top-of-funnel-spend figure is a DIFFERENT metric from the 75-to-25-assets/ATC/CVR figures in the claim -- they may come from a separate Motion webinar or podcast episode (a live Kyra Richards session) not indexed by search, or the claim may be conflati | Assets 75 to 25; ATC +56%; CVR +23%; time on page 1.5 to 3.5 min | Motion (Creative Analytics) -- partnership confirmed; specific figures not located 2025/2026 | Needs context Cần ngữ cảnh | ↗ |
| Salesforce's final 2025 holiday recap (published Jan 8, 2026, based on $1.29 trillion in global online holiday sales) stated that AI and agents 'drove 20% of all retail sales' worth $262 billion globally across Nov. 1 - Dec. 31, 2025 — a dollar-value share of sales, which many secondary sources loosely rephrase as '20% of orders.' The $262B/20% figure is CORRECT for the full-season window but needs context: (1) it is a sales-dollar share per Salesforce's own wording ('driving 20% of all retail sales'), not literally a share of order COUNT as some rephrasings ('one in five orders') imply | $262B / 20% of all retail SALES (dollar-value share), full global 2025 holiday season Nov 1 - Dec 31 — not literally '20% of orders' | Salesforce Newsroom (2025 Holiday Shopping Data) 2026-01-08 | Needs context Cần ngữ cảnh | ↗ |
| Visibility Labs/180 Marketing (94 U.S. e-commerce brands, Jan-Dec 2025, 9.46M non-branded-organic vs 135K ChatGPT sessions) found ChatGPT converts 31% higher than NON-BRANDED organic search (1.81% vs 1.39%) with 10.3% higher revenue per session but lower AOV. Kaiser & Schulze (973 sites, $20B revenue, Aug 2024-Jul 2025) found ChatGPT converts about 13% WORSE than organic search generally, and 86% worse than affiliate links -- a different, larger site sample over a different window. The two studies' headline conversion comparisons are not necessarily the same measurement: Visibility Labs explicitly strips out branded/homepage traffic from its organic baseline, while published coverage of Kaiser & Schulze does not state whether its broader 'organic search' includes branded search -- this distinction, not just sample differences, is the most likely reason the two studies disagree, but the K&S paper's own definition could not be confirmed from any source fetched this session. li#75's '86% conversion problem' framing is a likely misattribution: in Kaiser & Schulze, 86% is the AFFILIATE-vs-ChatGPT gap, not the organic-vs-ChatGPT gap (which is ~13%). A report citing 'Hamburg study, 86% conversion problem' for ChatGPT-vs-organic-search | Visibility Labs (now rebranded '180 Marketing') study: 94 seven/eight-figure e-commerce brands, GA4 data Jan-Dec 2025 (though the report itself cautions data wasn't statistically reliable until late April 2025); 9.46M non-branded-organic sessions vs 135K ChatGPT-referral sessions; conversion rate 1.81% (ChatGPT) vs 1.39% (non-branded organic) = +31%, ChatGPT ahead in 10 of 12 months; revenue/session $3.65 vs $3.30 = +10.3%; full-year AOV $204 (ChatGPT) vs $238 (organic), though the report flags H2-only AOV ($208 vs $238) as more reliable. All figures match the task's stated values (task's '$204 vs $238' AOV matches the full-year, less-reliable figure; if using the report's own preferred H2 number it's $208, not $204). Separately, Kaiser & Schulze (SSRN, Oct 2025): 973 sites, $20B combined revenue, Aug 2024-Jul 2025, 50,000+ ChatGPT transactions vs 164M traditional-channel transactions -- found ChatGPT converts ~13% worse than 'organic search' overall, and 86% worse than affiliate links specifically. Coverage of the K&S paper never specifies whether its 'organic search' bucket includes branded search (unlike Visibility Labs, which explicitly excludes homepage/branded traffic to isolate 'non-branded organic'); SSRN itself returned 403 to direct fetch this session, so the paper's own methodology section on branded-vs-non-branded organic could not be verified. | 180 Marketing (formerly Visibility Labs) blog; Kaiser & Schulze SSRN working paper via Digiday/Digital Commerce 360 coverage 180 Marketing post updated/bylined Feb 23, 2026 (Jeff Oxford) covering Jan-Dec 2025 data; Kaiser & Schulze posted to SSRN Oct 8, 2025, covering Aug 2024-Jul 2025 | Needs context Cần ngữ cảnh | ↗ |
| Anthropic's Claude Commerce Agents launch post (claude.com/blog/claude-for-commerce-agents, Sept 2, 2026) states retailers running shopping agents on Claude have seen carts up to 35% larger and shoppers 60% more likely to complete a purchase, but names no specific partner, sample size, or methodology for the claim. Fetched the primary post directly. It generalizes to 'retailers' (plural) with no partner named and no sample size/methodology disclosed alongside the headline claim -- exactly the self-reported, unaudited framing the corpus's conflict note flags (li#55/#109 s | Anthropic's own Sept 2, 2026 launch post says: "Retailers running shopping agents on Claude have seen carts up to 35% larger and shoppers 60% more likely to complete a purchase." | Anthropic (Claude.com blog) 2026-09-02 | Needs context Cần ngữ cảnh | ↗ |
| Digital Commerce 360 (Sept 2, 2026), citing Dr. Squatch staff (Joan Park, Senior CX Analyst; Will Edwards, Senior Strategic Operations Manager; Clara Kim, Senior Technical Program Manager), reports Dr. Squatch achieved a 25% reduction in 'where is my order' tickets, a 94% on-time delivery accuracy rate, and a 31.89% click-through rate on tracking-page product recommendations 'during Q1 2026' using AfterShip's AI post-purchase agent, and separately reports a 99.83% delivery-completion rate plus 62.69% month-over-month shipment growth during Cyber 5 2025. All numeric values in the panel's claim are individually accurate, but combining them under one 'BFCM/Cyber 5 2025' banner is misleading: the WISMO/accuracy/CTR numbers describe Q1 2026 performance (a later, different period than BFCM), while only the completi | Confirmed exact figures, but from TWO different time periods presented as one story: -25% WISMO tickets, 94% on-time delivery accuracy, and 31.89% tracking-page recommendation CTR are all Q1 2026 results. Only the ~100% (precisely 99.83%) Cyber 5 completion rate and +62.69% MoM shipment growth are actually from Cyber Five/BFCM 2025. | Digital Commerce 360 — 'Dr. Squatch drives revenue through post-purchase tracking page adjustments' 2026-09-02 | Needs context Cần ngữ cảnh | ↗ |
| Sierra's own LinkedIn post states: 'Minted saw a 387% month-over-month increase in conversation volume during Black Friday-Cyber Monday,' with no resolution/containment/CSAT figures given in the same post. Sierra's separate, undated general customer page for Minted (sierra.ai/customers/minted) states the agent achieves 'case resolution rates exceeding 65%... and over a 95% CSAT rating' — but that page does not tie those two figures to the same BFCM period as the 387% volume spike. The panel's conflict note is accurate and important: this specific 387% figure comes with NO resolution/containment/CSAT rate attached, so a reader cannot tell whether the extra conversation volume was actually handled well or just absorbed. The 65% resolution | +387% month-over-month increase in Minted's AI agent conversation volume during BFCM 2025 — confirmed as Sierra's own claim. No resolution rate, containment rate, or CSAT is disclosed alongside this specific figure. | Sierra (official LinkedIn post; and sierra.ai/customers/minted) approx. December 2025 (post reported '9 months ago' relative to a Sept 2026 fetch, consistent with a BFCM 2025 recap post) | Needs context Cần ngữ cảnh | ↗ |
| Anthropic's Sept 2, 2026 blog post 'Building Commerce Agents with Claude' opens with: 'Retailers running shopping agents on Claude have seen carts up to 35% larger and shoppers 60% more likely to complete a purchase,' naming no partner and no methodology. Trade press (e.g. a Sept 7, 2026 article and Post Radar, Sept 8, 2026) subsequently quoted Angela Jiang, Anthropic's head of product for the Claude platform, clarifying that the ~30-35% cart-size growth reflects results from 'one partner,' explicitly 'not an industry-wide average,' while coverage of the 60% purchase-completion-likelihood figure is less consistent — some articles tie it to the same single partner, others frame it as a broader 'early tests' finding about AI-referred traffic converting at a higher rate than other traffic sources. The number itself matches the panel's claim exactly, but the panel's own conflict notes are correct and important: li#31/li#150 wrongly generalize 'one partner' to 'retailers' (plural); li#135 misquotes the 60% purchase-likelihood figure as a 60-point 'purchas | Carts up to 30-35% larger, shoppers ~60% more likely to complete a purchase — confirmed verbatim as Anthropic's own published claim, but the cart-size figure is explicitly from ONE unnamed retail partner (per Anthropic's own head of product), not an industry-wide average; no baseline, sample size, time period, or methodology is disclosed anywhere. | Anthropic — 'Building Commerce Agents with Claude' (claude.com blog) 2026-09-02 | Needs context Cần ngữ cảnh | ↗ |
| This is Justin Brooke's own self-reported marketing description of his product/methodology (via podcast show notes), not an independently audited case study; the NYT figure it leans on is real but measured something different — a Digiday report on the New York Times confirms The Times built a synthetic-audience panel from its own subscriber database (642,000 readers) and validated it with a 10% internal holdout group, reaching a 0.918 (~92%) accuracy score on choosing among four candidate PODCAST NAMES — an editorial/content-preference test, not a purchase-prediction or e-commerce validation test, and not a comparison to traditional human focus groups in the market-research sense. CONFIRMS the panel's own flagged conflict: the NYT/92%-accuracy claim is real but is being borrowed from an unrelated domain (editorial content naming) to imply validation of e-commerce purchase-prediction, which the NYT case does not test. Additionally the pa | Justin Brooke's own podcast-episode description (Build With AI ep. 166, host Corey Ganim) states his 'predictive wear' framework runs ad copy/sales pages through 13 AI personas at '13 to 20 cents per run' (the panel's '13 cents' is the low end of a stated range, not the full figure), claims '$260K personally, 4 to 6X ROAS for clients' and 'a $36,000 offer launched using this exact process,' and asserts 'Harvard, Stanford, and the New York Times have all independently validated this approach,' citing that 'The New York Times uses the same synthetic audience process to test headlines and found 92% accuracy compared to their human focus groups.' | Podcast episode description: 'Build With AI' ep. 166, 'AI agent focus groups are the future of marketing' (host Corey Ganim, featuring Justin Brooke); cross-checked against Digiday's reporting on The New York Times' synthetic-audience research 2026-05-19 | Needs context Cần ngữ cảnh | ↗ |
| Hershey VP of Retail Stephanie Berman told Modern Retail that s'mores generate an estimated $200M-$250M in overall category sales during peak season (Easter through Labor Day, all brands combined), and separately that Hershey's own s'mores-related sales have grown double digits year-over-year to date; the AI display-placement tool (augmented reality + image recognition, deployed in 'several thousand stores') is described purely as a merchandising/analytics aid and is not credited by the source with causing either number. CORRECTION needed for the report: the $200-250M figure is the WHOLE US s'mores CATEGORY (all brands), NOT 'Hershey's peak-season sales' as the panel's claim implies by juxtaposition. Hershey's own sales growth (double-digit YoY) is a separate data point not ca | $200-250M is Hershey's own ESTIMATE of the TOTAL US s'mores CATEGORY's peak-season (Easter–Labor Day) retail sales — not Hershey's own sales, and not attributed to AI. Hershey's own s'mores-related product sales have grown 'double digits' YoY, stated as a separate, non-AI-attributed fact in the same article. | Modern Retail — 'Hershey starts using AI to strategize for s'mores season' 2026 | Needs context Cần ngữ cảnh | ↗ |
172 numbers checked against a primary source · 172 con số đã đối chiếu tới nguồn gốc. A further 38 were dropped because no source could be found · 38 con số khác bị loại vì không truy được nguồn.