Shoppers tapped and scrolled their way to a record Thanksgiving haul even as overall growth cooled, and the story behind the numbers reveals a market where mobile dominance and AI guidance rewrote playbooks long before the evening rush arrived. This roundup brings together the strongest reads from leading measurement firms to explain what happened, why it mattered, and how merchants can steer the rest of Cyber Week with precision.
The goal is simple: synthesize what the data providers saw, contrast the perspectives where they diverged, and turn their collective lessons into practical moves for the days ahead. Consider this a stitched-together briefing on the most consequential shifts—timing, channels, discounts, and payments—told through the lens of multiple sources.
Context that frames the record: how Thanksgiving sits inside Cyber 5 and why 2025’s growth cooled
Across datasets, Thanksgiving set a new record while signaling a moderation in momentum. Adobe estimates U.S. online sales reached $6.4 billion, up 5.3% year over year, while Salesforce measured a 3% U.S. increase and 6% globally. Both reads point in the same direction: elevated demand, slower pace versus last year. Shopify’s merchant telemetry corroborated the strength of the day while highlighting the same structural forces—mobile concentration and earlier buying—that shaped outcomes.
Why did a record arrive with a softer beat? Mobile crossed new thresholds, midday became the new prime time, and AI began to act like a true conversion engine. Those advances concentrated high-intent activity into targeted moments, which lifted conversion and average order value even as order volumes grew more slowly. In other words, shoppers bought bigger and earlier, but fewer times.
What follows summarizes the timing shift to late-morning peaks, names the category winners, dissects discount depth by vertical, and reconciles mixed BNPL reads. It also sets expectations for the next two anchors of Cyber 5, where mobile remains the default and AI’s referral power continues to climb.
What powered the record—and what tempered the tempo
The dominant themes were clear across providers: mobile’s majority share, AI’s outsized ability to convert, and promotions that pulled demand forward without breaking average discount rates. These combined to elevate spend per order and spotlight specific categories, particularly large appliances, consoles, and fitness gear.
Tempering the tempo were softer order counts and uneven financing signals. Even with a higher share of purchases closing on phones, growth decelerated versus last year. BNPL’s picture looked steady in one dataset and softer in another, implying a complex mix of financing preferences and retailer portfolios at work.
Midday, on a phone: mobile crosses 60% as peak demand moves to late morning
Mobile finally cleared a symbolic bar. Adobe places mobile at 61.6% of Thanksgiving spend, translating to an estimated $3.9 billion and up 9.4% in dollars. Salesforce goes farther on mix, attributing 80% of traffic and 75% of orders to mobile on the day. The channel didn’t just gain share—it redefined when people buy. Adobe identified a 10 a.m.–2 p.m. peak, while Shopify’s network saw an 11 a.m. ET crest, both a break from last year’s late-evening surge.
That temporal shift mattered for conversion. Small screens now benefit from more ergonomic cart flows, app-level speed, and autofill that reduce friction at precisely the hours when shoppers have a few minutes to act. With inventory promises and delivery windows visible up front, midday mobile sessions turned into fast, decisive purchases rather than idle browsing. The result: cleaner conversion funnels and higher basket values, especially when paired with timely push or SMS nudges.
However, this isn’t a story of unbounded acceleration. Growth slowed despite better mobile conversion because order frequency lagged. In practical terms, merchants saw higher intent in compressed windows; the task now is to expand those windows or replicate them on adjacent days without eroding margins.
Agents at the cart: AI referrals explode and deliver higher-intent shoppers
Referral traffic from AI sources moved from curiosity to force multiplier. Adobe tallied a 725% surge in AI-originated visits and a 54% conversion lift for those users relative to other referrals. Salesforce observed even sharper advantages around the two days before Thanksgiving: agentic AI traffic converted 600% better globally and 300% better in the U.S. Compared with social or generic referral channels, AI delivered users who had already narrowed choice sets and budget ranges.
Retailers running on-site agents saw that intent translate into outperformance. Salesforce reports that merchants with embedded agents grew 8.7% year over year at the start of the week, compared with 6.4% for those without. Third-party consumer agents—ChatGPT, Perplexity, and peers—sent rising volumes, underscoring a new top-of-funnel where shopping lists, deal tracking, and comparison logic increasingly originate outside brand domains.
Yet the AI tide raised new questions. Attribution chains grew more complex as handoffs between agents, apps, and web sessions multiplied. Content quality and margin control became flashpoints when AI steered shoppers toward the deepest deals. And brand visibility inside third-party agent answers remained contested terrain, pushing retailers to optimize feeds and metadata so agentic systems can surface relevant, profitable products.
Promotions with a purpose: selective depth pulls buying forward and reshapes baskets
Discounts didn’t just get louder; they got smarter by category and earlier on the calendar. Salesforce recorded some of the steepest pre-Thanksgiving markdowns in beauty and apparel—roughly 40% in makeup, 30% in skin care, 28% in apparel—while average global and U.S. rates stayed broadly flat versus last year on those days. Adobe’s read aligns: stronger-than-expected day-of dealmaking helped capture impulse buys on mobile and nudge decisive purchases during daytime windows.
That approach shaped the basket. Tight, high-visibility deals on known favorites prompted quick taps, while larger-ticket buys cleared during the late-morning peaks when shoppers could compare delivery dates and warranty terms. The mix favored higher spend per order, a natural outcome when targeted depth intersects with a mobile-first funnel that minimizes friction.
The risk is familiar—margin cliffs and competitive leapfrogging. The opportunity is orchestration: pacing depth where elasticity is proven, staggering windows to avoid channel cannibalization, and pairing agent-driven guidance with curated bundles that protect unit economics even when headline discounts look aggressive.
Payments and proof points: BNPL steadies, baskets swell, and methodologies matter
Signals on financing were mixed but instructive. Adobe estimates $447.7 million of Thanksgiving purchases used BNPL, up 4.1% year over year, suggesting steady engagement on the day. Salesforce saw BNPL usage flat globally and down 5% in the U.S., attributing the softness to convenience-led choices and alternative credit options. Both reads imply BNPL remained meaningful without driving disproportionate upside.
These financing patterns tied neatly to the “higher spend per order, softer order counts” theme. Big-ticket categories—consoles, large appliances, exercise equipment—over-indexed, inflating average order value even as orders grew modestly. When delivery promises were clear and checkout was one tap away, shoppers chose certainty over volume.
Methodology helps explain the numeric gaps without negating the direction. Adobe’s panel spans more than a trillion visits across wide category coverage, while Salesforce aggregates telemetry from its merchant base across commerce and marketing platforms. Different retailer mixes and windows produce different absolute numbers, yet both frameworks pointed to the same narrative: durable demand, moderated growth, and a financing mix that remains stable rather than explosive.
What to do next: practical moves for the rest of Cyber Week and beyond
Sources converge on one immediate priority: design the experience for the midday mobile spike. Page speed, image compression, inventory surfacing, and clear delivery promises should take center stage, with push and app moments landing between 9 a.m. and 1 p.m. local. Checkout should require the fewest taps possible, with wallet options prioritized and friction screens removed or deferred.
AI deserves dedicated treatment rather than a generic add-on. On-site agents should guide discovery, explain tradeoffs, and assemble bundles that balance value and margin. Clean, structured feeds need to flow to third-party agents so products can be recommended accurately. Measurement must evolve in lockstep: tag agent-originated cohorts distinctly and compare their lifetime value and return behavior to other referrals.
Promotions work best when elasticity is respected. Keep depth where responsiveness is proven—beauty and apparel are prime candidates—seed selective big-ticket offers, and stagger windows to avoid race-to-the-bottom dynamics. On payments, surface BNPL clearly on high-ASP items, pre-approve returning customers when possible, and test alternative wallets to lift one-tap conversion on small screens.
Looking ahead to Black Friday and Cyber Monday: sustained momentum, smarter execution
The near-term setup remains solid: record spend with moderated growth, mobile as the default, AI as a real conversion engine. Adobe projects $11.7 billion for Black Friday, up 8.3% year over year, and $14.2 billion for Cyber Monday, up 6.3%; mobile should hold a majority share on Black Friday. The expectation across sources is a continuation of targeted promotions rather than blanket markdowns, with depth concentrated where it moves the needle.
Practical focus areas follow naturally. Schedule mobile interventions for late morning, align merchandising with the categories shoppers already chose—consoles, large appliances, fitness, toys, and activewear—and ensure dynamic recommendations reflect trending searches and agent queries. For measurement, unify referral tags for AI sources, separate agentic traffic from social, and benchmark mobile AOV and conversion by hour to sharpen daypart strategy.
For readers seeking more than a replay, three threads offered the clearest path forward: lean into the late-morning, phone-first reality; treat AI as a primary acquisition and conversion channel with dedicated measurement; and calibrate discount depth by category to maintain margin while sustaining velocity. For deeper context, consult holiday trackers and periodic dashboards from leading analytics platforms, examine merchant cohort studies that isolate agent impact, and review hourly conversion analyses to refine dayparting. Taken together, these steps tied the record to repeatable practice, grounded expectations for the weekend, and positioned execution to benefit from the mobile and AI currents that had already reshaped the holiday.