The past week produced holiday spending forecasts ranging from down 10% to up 5.3%, depending on which research firm you ask. Consumer confidence metrics hit levels not seen since 1997, yet actual spending data from Bank of America cards shows growth continuing at 2.0% year-over-year in September.
When predictions conflict this sharply, watching where companies actually put their money becomes more useful than parsing survey intentions. This week's data shows clear patterns in capital allocation, format decisions, and technology deployment that reveal how retailers are positioning for the next 12-24 months.
The AI and Automation Investment Surge
NVIDIA's State of AI in Retail survey published this week found 89% of retailers are now actively using AI or assessing projects, up from 82% in 2023. More importantly, 87% report positive revenue impact from their implementations and 94% achieved operational cost reductions. These aren't pilot projects anymore. The survey covers retail executives across major chains, though NVIDIA doesn't disclose the exact sample size or company list.
The specificity of reported use cases matters here. Inventory management leads adoption, followed by marketing content generation at 60%, predictive analytics at 44%, and personalized marketing at 42%. These aren't speculative applications. They're operational systems with measurable outcomes.
Honeywell's survey of 450 retail executives conducted in May found 85% have developed AI capabilities, with 60% actively expanding implementations and 89% having investment underway or planned within 1-2 years. Executives cited ROI realization within 12-18 months for most AI projects.
Walmart's announcement of 90 million battery-free IoT sensors from Wiliot represents the scale at which this technology is moving. The sensors harvest energy from radio waves, light, motion, and heat, eliminating battery replacement costs. Currently active in 500+ locations, the deployment will expand to 4,600 Supercenters and Neighborhood Markets plus 40+ distribution centers in 2026, covering 90 million pallets.
The sensor data feeds directly into Walmart's AI systems for real-time inventory tracking and dynamic stock management. This isn't a test. It's enterprise-scale deployment across the largest U.S. retailer.
What this means for retailers: The NVIDIA finding that 87% report positive revenue impact provides evidence that retail AI has moved beyond experimental to demonstrably value-generating at scale. The 89% adoption rate suggests implementation is becoming table stakes. Companies not yet deploying AI face growing operational disadvantage as competitors achieve measurable efficiency gains.
The Format Consolidation and Micro-Fulfillment Bet
Amazon's actions this month reveal a clear strategic pivot. The company announced closures of multiple Amazon Fresh locations across four states in October, including four Southern California stores and additional locations in Washington, Virginia, New York, and Los Angeles. Only 60 Amazon Fresh stores remain operational in the U.S., down from a peak of 70+. The company also closed 14 UK locations and converted five UK stores to Whole Foods format.
At the same time, Amazon opened a hybrid Whole Foods format in Plymouth Meeting, Pennsylvania on October 8. This store combines traditional Whole Foods shopping with a 10,000 square foot automated micro-fulfillment center built within the existing store's back-of-house area. Customers can scan QR codes while shopping in-store to add items for pickup or delivery, accessing 12,000+ unique items from both Whole Foods and Amazon brands.
The format uses Fulfil's grocery robotics platform to prepare orders within minutes. Free pickup on orders of any size, with $9.95 delivery for Prime members. Amazon described this as testing to "refine and expand this offering to additional stores over time."
The pattern is clear. Close formats that haven't proven demand (Fresh stores, Just Walk Out technology removed in spring 2025). Invest in formats that integrate digital and physical (micro-fulfillment centers, hybrid shopping experiences). Amazon also announced Whole Foods Daily Shops, a small-format concept at 7,000 to 14,000 square feet versus traditional 40,000 sq ft stores, with a Manhattan location scheduled for 2025.
Reformation opened its New Orleans location on October 30 featuring its "Retail X" technology concept. Touchscreens throughout the store allow customers to browse full inventory and digitally order items to dressing rooms. The store includes an in-store recycling program where customers can return old Reformation pieces for credit.
None of these format experiments have released performance metrics yet. All opened within the past 30 days or remain in test phase. The absence of published data suggests either results are too early to assess or companies view the metrics as competitively sensitive.
What this means for retailers: The investment pattern suggests successful small-format strategies require digital augmentation rather than simply shrinking product selection. Pure format downsizing without capability enhancement appears less viable. Capital allocation favors stores reimagined as fulfillment-enabled customer touchpoints rather than traditional product display spaces.
The Channel Migration Numbers Show Acceleration
The U.S. Buy Online, Pick Up In-Store market hit $129.36 billion in 2024, according to market research from Debutify analyzing multiple industry sources. Projections show the market reaching $509.4 billion by 2033, representing 16.45% compound annual growth rate.
Capital One Shopping Research found 67% of Americans used BOPIS in the last 6 months, with 68% making multiple purchases through the channel. Approximately 97.2 million Americans regularly use BOPIS, representing 34.2% of U.S. consumers. Grocery dominates the channel. About 78.7% of click-and-collect sales are grocery items, with 84% of online grocery shoppers using BOPIS at least once monthly.
The BOPIS adoption rate dwarfs curbside pickup. Consumers are 99.1% more likely to use BOPIS versus curbside for click-and-collect, according to the Capital One data. The research doesn't explain why this gap exists, though consumer preferences for seeing items before taking them home (77% cite this reason) may explain part of it.
Adobe Analytics projects mobile commerce will hit 56.1% of all online holiday sales in 2025, representing $142.7 billion. This marks the first full calendar year mobile exceeds 50% of e-commerce spend. Adobe bases this projection on analysis of over 1 trillion visits to U.S. retail sites and over 100 million SKUs.
Digital wallets captured 54% share of e-commerce transactions in 2025, with 5.6 billion digital wallet users worldwide and 70% of U.S. online adults using digital or mobile payments in the past 3 months. Credit cards fell to 31% of online transactions, down from 40%+ in prior years.
Buy Now, Pay Later shows similar growth trajectory. Adobe projects $19.8 to $20.4 billion in BNPL spending for November-December 2025, up 9-12% year-over-year and representing approximately 8% of all online holiday sales. Cyber Monday will become the first single day to cross $1 billion in BNPL spend at $1.05 to $1.08 billion.
The global BNPL market hit $560.1 billion in 2025 according to CoinLaw's analysis of multiple industry reports, with $122.26 billion in North America. About 91.5 million U.S. consumers are expected to use BNPL in 2025, with average transaction value of $142.
What this means for retailers: The BOPIS projection from $129 billion to $509 billion by 2033 represents nearly 4x growth in nine years. Physical stores' future role may be primarily fulfillment nodes for digital orders rather than traditional shopping destinations. The 54% digital wallet share indicates payment infrastructure has fundamentally shifted, with implications for checkout experience design and loyalty program integration.
The Confidence-Spending Gap Persists, But Cracks May Be Forming
The Conference Board Consumer Confidence Index fell to 94.6 in October, down from 95.6 in September. The Expectations Index dropped to 71.5, down 2.9 points from September's 74.4. This represents the ninth consecutive month below the 80 threshold that historically signals recession within a year.
The Present Situation Index rose to 129.3, up 1.8 points from 127.5. This creates a stark divergence where consumers view current conditions as strong but future prospects as increasingly bleak. The Conference Board survey draws from a panel of 36+ million consumers, with monthly samples of approximately 3,000 respondents.
The University of Michigan Consumer Sentiment Index registered 53.6 in October, down from 55.1 in September and approximately 19% below year-ago levels. This represents near historic lows, only modestly above pandemic-era readings. Year-ahead inflation expectations hit 4.5% while long-run expectations ran at 3.4%, both elevated relative to actual measured inflation of 2.7%.
Yet Bank of America card transaction data for September shows 2.0% year-over-year growth, marking the fourth consecutive monthly gain. The Redbook Index tracking 9,000+ major U.S. retailers registered 5.7% year-over-year growth for the week ending November 1, up from 5.0% the prior week.
This disconnect between sentiment and behavior has persisted throughout 2024-2025. Consumers have consistently expressed more pessimism than their actual spending patterns justify. Whether this time the pessimism proves predictive remains the key unknown.
Holiday spending forecasts from major research firms released this week show unprecedented divergence:
The 15+ percentage point range in these forecasts (from -10% to +5.3%) reflects fundamental disagreement about whether consumer pessimism will finally translate to reduced spending.
Boston Consulting Group's November 3 study of 10,000+ consumers across 10 countries found 81% concerned about rising prices for everyday essentials and 71% bracing for tariff-related increases. Consumers expect at least 30% off to qualify as a "good deal," and 77% deliberately delay purchases earlier in the year to capture Black Friday promotions.
What this means for retailers: The wide forecast range suggests planning for multiple scenarios rather than anchoring to any single projection. The BCG finding that consumers expect at least 30% off indicates promotional intensity will likely increase during holiday periods. Companies with strong proprietary data on their specific customer segments have advantage over those relying primarily on industry surveys with conflicting signals.
Government Data Goes Dark, But Companies Keep Moving
The federal government shutdown beginning October 1 has suspended all major economic data releases. The Census Bureau, Bureau of Economic Analysis, and Bureau of Labor Statistics cannot publish new reports during the shutdown.
As of November 6, the following data releases remain unavailable:
- Census Bureau retail sales for September (scheduled October 16)
- BEA Personal Income and Outlays for September (scheduled October 31)
- BLS Employment Situation for October (scheduled November 1)
- All November data releases remain uncertain
The most recent available official data comes from August 2025. Census Bureau retail sales for August totaled $732.0 billion, up 5.0% year-over-year. BEA Personal Consumption Expenditures increased $129.2 billion in August (0.6% monthly rate), while the personal saving rate fell to 4.6% from 4.8% in July.
Federal Reserve officials acknowledged in late October statements that they are struggling to assess the economy's state without access to current government data. The Fed is relying more heavily on private-sector data sources and regional Fed surveys to compensate for the federal data gap.
The timing creates particular challenges because September and October represent the critical transition from summer spending patterns into holiday preparation. Retailers typically use September retail sales data to finalize November-December inventory, promotional, and staffing plans.
Yet retail format experiments, AI deployments, and channel migration investments are all proceeding. Companies are making decisions with proprietary data sources rather than waiting for government confirmation. Sensormatic Solutions published its forecast of the busiest in-store shopping days for the 2025 holiday season, with November 28 (Black Friday) at #1, December 20 (Super Saturday) at #2, and December 13 at #3.
The forecast notes in-store traffic follows the "50/20 rule," where 50% of the week's traffic and sales occur in the busiest 20 hours. This type of analysis relies on historical foot traffic data rather than government statistics.
What this means for retailers: Companies with superior internal analytics capabilities have unprecedented advantage. Those dependent on government statistics for demand planning face decision-making without September or October confirmation during the year's most critical quarter. The ability to make confident decisions with proprietary data becomes a competitive differentiator when public data disappears.
Following the Strategic Pattern
When forecasts diverge by 15+ percentage points and official data goes dark for months, capital allocation reveals strategy more clearly than survey intentions. This week's data shows retailers betting on:
Operational efficiency through AI and automation: 89% adoption rates with 87% reporting positive revenue impact suggests this is where companies see certain returns regardless of demand environment.
Integrated digital-physical formats: Micro-fulfillment centers, endless aisle technology, and format consolidation away from purely digital (Amazon Fresh closures) or purely physical toward hybrid models.
Channel migration infrastructure: BOPIS market projected to grow 4x by 2033, mobile crossing 50% of e-commerce, digital wallets capturing 54% share. Companies are investing in fulfillment and payment infrastructure for channels showing clear growth trajectories.
Scenario planning rather than single forecasts: The absence of consensus on holiday spending combined with government data blackout means companies with flexible operations and strong proprietary data have advantage over those committed to single demand scenarios.
The next 60-90 days will show whether consumer pessimism finally impacts behavior or whether the sentiment-spending disconnect persists through another holiday season. But retailers aren't waiting for that answer. They're placing bets on operational capabilities that deliver returns across multiple demand scenarios.