Is your small business visible when someone asks an AI assistant to find the right gift, compare prices or discover a store they have never heard of? The new holiday storefront is increasingly a conversation.
AI is becoming the shopper standing between the customer and your product page
QuickBooks’ 2026 survey suggests AI-assisted holiday shopping has moved well beyond experimentation.
| What AI shoppers want help with | Share of AI users | What that means for merchants |
|---|---|---|
| Find a better price or discount | 64% | Your current price and sale price need to be clear and current. |
| Save time searching | 47% | The assistant must quickly understand what the product is and whether it fits the request. |
| Discover new products or retailers | 46% | Small brands can potentially enter a consideration set without being a household name. |
| Receive personalized recommendations | 39% | Attributes, audience and use cases matter more than vague lifestyle copy. |
| Feel more confident choosing a gift | 27% | Reviews, return policy and reliable factual product information can reduce uncertainty. |
The survey also found that 33% of consumers said their use of AI shopping tools increased from the previous year.
What shoppers are not yet doing in large numbers is handing over the entire transaction. Only 15% reported allowing an AI assistant to complete a purchase on their behalf, while half said they would not currently be comfortable doing that.
Nearly half say they are already trying to become easier for AI to find
QuickBooks found that 46% of surveyed small businesses had already taken steps to improve their visibility in AI-powered search or shopping tools, including systems such as ChatGPT and Google’s AI experiences.
Another 33% said they planned to.
The assistant needs facts it can compare
OpenAI says product recommendations can draw on structured metadata from merchants and third-party providers as well as other public web information.
For merchants appearing on a product detail screen, OpenAI says ranking can consider factors such as:
- Availability
- Price
- Product quality signals
- Whether you are the maker
- Whether you are the primary seller
- Merchant and product metadata
OpenAI’s direct product-feed specification makes the information requirements even clearer. Basic discovery records include stable product identity, title, factual description, URL, brand, seller name, product image, price and availability.
Merchants outside Shopify—or larger merchants wanting more direct control—can also look at OpenAI’s Agentic Commerce Protocol and product-feed program. Direct product-feed onboarding is currently available to approved partners.
10 things an AI shopping assistant needs to understand about your business
Exactly who you are
Use the same business and brand identity across your website, product listings, marketplace accounts, Merchant Center, Business Profile and social profiles.
An AI system should not have to decide whether “Smith Handmade,” “Smith Gifts LLC” and “SmithHomeStore” are three unrelated companies.
What the product actually is
Avoid product titles that sound clever but communicate little.
“Sunday Morning” may be a nice collection name. “Hand-poured 12-ounce soy candle — cedar and vanilla” tells a shopping system what the item actually is.
OpenAI currently recommends concise titles and factual product descriptions rather than vague marketing language.
Who the product is for
Gift shopping is usually expressed through intent rather than SKU numbers.
A shopper may ask for:
- A gift for a 10-year-old who likes science.
- A retirement gift for a fisherman.
- A hostess gift under $40.
- A durable travel bag for someone who flies weekly.
- A stocking stuffer for a coffee lover.
Useful factual attributes—age range, interests, material, dimensions, capacity, compatibility, style, skill level, occasion and intended use—help systems connect those requests with products.
What it costs right now
Price is particularly important this season because finding a better price or discount is the most common reason AI shoppers gave for using the technology.
Keep the visible product page, structured product data and merchant feeds synchronized.
If a sale begins Friday night, your feed should not still report Thursday’s price days later.
Whether it is actually in stock
OpenAI’s product-feed format explicitly requires an availability status such as in stock, out of stock, preorder or backorder.
Google Merchant Center likewise treats price and availability consistency as core product-data quality issues.
Which size, color or variation is available
Do not treat all variants as one indistinguishable product if the purchasable options differ.
OpenAI’s current feed guidance recommends unique variant identities and variant-specific titles, URLs, media, price and availability where those attributes differ.
Google Merchant Center also warns that missing or incorrect variant information such as size, color and item-group IDs can limit product eligibility or produce incorrect displays.
How much shipping costs—and whether it arrives in time
This is a major holiday issue. QuickBooks found 41% of shoppers expect higher shipping and delivery costs this season, up sharply from the previous year.
Make shipping thresholds, pickup options, processing times and realistic delivery information easy to find.
Google product experiences can use shipping information, and current OpenAI commerce specifications include fulfillment and shipping information beyond the basic discovery fields.
What happens if the recipient hates it
Returns become part of the product decision when an assistant is helping a shopper compare similar gifts.
State the return window, return method, fees, exclusions and warranty clearly rather than hiding them behind vague language.
Google supports merchant-return-policy information in structured data, while OpenAI’s richer commerce feed model can include returns and related seller-policy information.
Why a shopper should trust the product
OpenAI says product recommendation experiences may consider reviews and other third-party information, and merchant ranking can consider quality.
Google can also use legitimate product-rating information in product experiences.
This does not mean stuffing your own website with fake five-star testimonials. The valuable signals are real customer feedback, consistent product information, useful photography and evidence that the merchant is what it claims to be.
Where the machine can get the authoritative answer
Your product should have a stable, crawlable product page with a clear canonical destination rather than relying entirely on social posts, JavaScript popups or information available only after a login.
For Google, keep Merchant Center and relevant structured product data synchronized with the visible page.
For ChatGPT search visibility, OpenAI says sites should allow OAI-SearchBot. Blocking that crawler prevents pages from appearing as sourced content in ChatGPT search answers, although links can sometimes still surface through other discovery paths.
For commerce, product feeds can provide an even more direct structured catalog path.
Website + structured data + merchant feed beats trying to “write for robots”
| Layer | What it does | Small-business action |
|---|---|---|
| Product page | Human-readable source of truth | Keep title, specifications, price, stock and policies current. |
| Product structured data | Expresses product facts in machine-readable form | Use Product / Offer markup that matches the visible page. |
| Google Merchant Center | Supplies structured catalog information to Google’s shopping ecosystem | Maintain accurate products, identifiers, variants, shipping and availability. |
| OpenAI commerce feed | Supplies structured catalog data directly for ChatGPT commerce experiences | Eligible merchants can explore ACP/direct-feed participation. |
| Shopify Catalog | Can supply product data into ChatGPT for Shopify merchants | Keep Shopify’s actual catalog data clean and current. |
| Business identity | Helps establish who the merchant is | Keep Organization/OnlineStore and business-profile details consistent. |
| Crawler access | Lets search systems retrieve public pages | Check robots.txt, CDN and bot-security settings. |
Five “AI SEO” distractions
Do not obsess over
- Repeating every imaginable conversational keyword.
- Creating hundreds of near-identical gift-guide pages.
- Fake Reddit or forum mentions.
- Rewriting normal product copy into strange “AI-friendly” prose.
- Treating llms.txt as a magic Google ranking switch.
Spend the time on
- Clean product data.
- Accurate prices.
- Accurate inventory.
- Real variants and attributes.
- Shipping and return details.
- Strong original photos.
- Legitimate reviews.
- Crawlability.
Google’s own 2026 guidance specifically says there is no special AI markup required for its generative-search experiences and that Google Search does not use llms.txt as a visibility signal.
It recommends continuing to focus on crawlability, useful content, technical SEO, good images, Merchant Center data and Business Profile information.
Can the page answer a shopper’s constraints without guessing?
| Weak product page | Machine-friendly and customer-friendly version |
|---|---|
| “The perfect holiday companion.” | “Insulated 20-ounce stainless-steel travel mug with leak-resistant lid.” |
| “Available in beautiful colors.” | “Available in navy, sage, black and cream.” |
| “Great gift!” | “Gift box included; popular for teachers, coworkers and coffee drinkers.” |
| “Fast shipping.” | “Orders placed by 2 p.m. CT ship within one business day.” |
| “Easy returns.” | “Unused products may be returned within 30 days; holiday purchases through Dec. 24 returnable through Jan. 31.” |
Make “near me” understandable too
AI shopping is not limited to shipped e-commerce products. Google specifically recommends keeping Business Profile information current for generative AI search, and LocalBusiness structured data can communicate details such as location and opening hours.
A local merchant should make these facts unambiguous:
- Correct business name
- Street location
- Holiday opening hours
- Telephone number
- Local pickup availability
- Service or delivery area
- Whether the item is stocked locally
- Pickup timing
- Parking or access information where relevant
- Current Business Profile details
That query contains at least five facts a merchant needs to communicate clearly: ownership/identity, location, product category, price and local availability.
Try these five shopper questions before Black Friday
| Test prompt | What it exposes |
|---|---|
| “Find me a unique [your category] gift under $50.” | Price, product category and differentiation. |
| “Compare [your product] with similar options for [recipient/use].” | Attributes, use cases, reviews and competitive positioning. |
| “Can I get [your product] before December 20?” | Availability, fulfillment and shipping clarity. |
| “Which seller has the best price on [your product]?” | Current price, promotions and merchant identity. |
| “Find a small business selling [product] with easy holiday returns.” | Merchant identity and return-policy visibility. |
If you only have one afternoon, fix these first
| Priority | Fix | Why now |
|---|---|---|
| 1 | Price and sale-price accuracy | Deal finding is the leading AI-shopping use case. |
| 2 | Stock and variant accuracy | Do not send holiday traffic to unavailable merchandise. |
| 3 | Shipping cutoff / delivery information | Holiday shoppers care about arrival dates. |
| 4 | Clear factual titles and attributes | Helps assistants match products to intent. |
| 5 | Return policy | Reduces gift-purchase uncertainty. |
| 6 | Merchant feed / structured data consistency | Creates cleaner machine-readable product facts. |
| 7 | Crawler and indexing audit | A perfect product page is useless to search systems that cannot access it. |
QuickBooks survey data came from an August 2026 survey of 6,000 U.S. consumers participating in the holiday season and 1,002 U.S. small-business owners, managers and decision-makers at businesses with 0–99 employees. Survey-based spending figures are estimates based on respondents’ stated intentions, not actual completed holiday transactions.
Visibility in ChatGPT, Google AI features or any shopping system is not guaranteed by implementing these recommendations. Product discovery systems, ranking factors, data providers and merchant integrations continue to evolve.
Can an AI shopping assistant understand your store?
Check each statement that is reliably true for your important holiday products today—not what you plan to fix next week.
Check what is true today
Your AI-shopping visibility score will appear here.
This score measures whether your product information is ready for AI-assisted discovery. It cannot predict whether a particular product will be recommended or ranked by ChatGPT, Google or another shopping platform.
