In its September 28 holiday forecast, Adobe projected $275.1 billion in U.S. online spending for November 1 through December 31, 2026, up 6.7% year over year. It also forecast a 130% rise in AI-referred traffic to U.S. retail sites, measured through shoppers clicking links. These are forecasts, not completed season results or a prediction that an individual merchant’s orders will grow at the same rate.
Define the visit that would help the customer
Start with a product or category the business can fulfill well. Identify the decision the page helps a buyer make and the facts needed to make it accurately. A visitor who reaches a vague page through a sophisticated tool still faces a vague page.
Review specifications, available variants, compatibility limits, delivery information and the return process. Assign someone who understands the product to check those details. Do not use generated descriptions to invent capabilities, certifications or experiences that the business cannot support.
Keep discovery information connected to stock
A product description can remain persuasive after the relevant variant has sold out. Decide who updates availability and how the customer is informed when an item cannot be supplied. Check the page and checkout together so that they do not offer conflicting answers.
For a hypothetical gift retailer, an AI-assisted shopper might arrive at a specific size or color rather than the general category. Staff should know how that variant is represented, which alternatives can legitimately be offered and what happens when inventory changes between the visit and the order.
Review the mobile task from beginning to end
Adobe’s forecast puts mobile at 57.4% of full-season online spending. That makes a practical phone review worthwhile, but a national share should not be substituted for the business’s own customer data. Use the devices and flows relevant to the store.
Check whether essential product details, the complete amount and delivery conditions remain understandable through checkout. Observe confusing labels, blocked controls or unexpected account requirements. Give the site maintainer a specific finding and a reproducible task rather than a request to make the site feel more modern.
Measure fulfilled value, not just the referral count
Choose an outcome the business can verify, such as a correctly fulfilled order or a useful inquiry. Keep the measurement definition stable and identify limitations in the available referral data. Do not attribute every increase in visits or revenue to AI without evidence.
Review cancellations, avoidable customer questions and returns alongside order value. A campaign that attracts poorly matched buyers may create work that is invisible in the traffic chart. Preserve the reason for an exception so that the next content change addresses the actual problem.
Keep the forecast in the planning column
The forecast can help an owner prepare for different discovery habits and a busy online season. It does not determine the right marketing budget, inventory order or discount for a particular business. Test those decisions against the store’s own capacity and contribution.
The most useful preparation is accurate product information and a dependable customer task. AI-assisted discovery can change how a shopper reaches the business; it does not remove the merchant’s responsibility to explain the offer clearly and deliver what was promised.
Eric Kuvykin
Publisher and editorial director covering payments, fintech, merchant services, banking, AI, technology and operating strategy for small and medium-sized businesses.
