In September, a Shopify agency sent a shopping bot to 200 online stores in the US and UK. One in seven shut the door on it. Home Depot, Best Buy, Macy’s, Tesco, Argos, and Next all turned it away twice. That’s a real risk with Black Friday on November 27. AI shopping agents are tools like ChatGPT, Google’s AI Mode, and Microsoft Copilot that search, compare, and sometimes buy products for a shopper.
To get picked, your store must let them in. It must also give them clean product data and a checkout they can finish. This guide is a checklist you can work through in the 11 days left. We cover access, product data, live pricing, agent checkout, and tracking. Each step comes with the numbers behind it.
Key Takeaways
- Adobe expects AI traffic to US retail sites to rise 130% this season, with a 95% jump on Black Friday itself.
- Bot protection, not robots.txt, is the most common reason agents are blocked. So security and marketing teams should review the rules together.
- Most stores include their share price and stock information in their schema. Few share shipping (22%), returns (20%), or barcodes (18%).
- Checkout for AI shopping agents is real but early. Shopify stores have the easiest path, while others should start with Google Merchant Center.
- Keep the hype in check: only 7% of US shoppers plan to start Black Friday searches in an AI chatbot.
What Are AI Shopping Agents, and How Big Is Black Friday 2026 for Them?
AI shopping agents are assistants that research products for a shopper and, in some cases, place the order. This Black Friday, they’ll send more traffic than ever. Still, they’re a small slice of total sales.
There are two kinds worth knowing. Research agents, such as Perplexity Shopping, compare options and cite sources. The shopper still checks out on your site. Execution agents go further and place the order through a checkout connection.
Interest in AI shopping agents is growing fast. According to Adobe’s 2026 holiday forecast, AI traffic to US retail sites will rise 130% year over year from November 1 to December 31. Adobe expects the biggest jumps on Thanksgiving (159%), Black Friday (95%), and Cyber Monday (88%).
Salesforce goes further. Its 2026 holiday predictions say 20% of holiday ecommerce traffic will come from AI chat agents. However, that figure also counts backend bots and rival price scrapers, so read it with care.
The traffic is valuable, too. Adobe says AI’s revenue-per-visit lead over other traffic will double this year. In addition, 69% of shoppers told Adobe they’re less likely to return an item bought with an AI assistant’s help.
Here’s what most guides miss, though. In an Omni Calculator survey of 899 US adults (September 2026), 59% planned to use AI this Black Friday. Yet only 7% will start their search in a chatbot. Amazon (31%) and Google or Bing (29%) still lead.
Our view: treat AI shopping agents as a fast-growing second front, not a swap for search. The fixes below are cheap, and most of them help your Google rankings as well.
Checklist Part 1: Make Sure AI Shopping Agents Can Get In
First, check that your robots.txt file and your bot protection both allow the agents you want. Most blocks happen in security software, not in robots. txt.
Blend Commerce’s 2026 AI Shopping Audit tested 200 UK and US stores from September 8 to 15. In total, 29 stores, or 14.5%, refused its agent twice. Large retailers did worse: 20% blocked it, against just 2.5% of emerging brands.
The odd part? All 24 stores that refused access but had readable robots.txt files welcomed AI search crawlers. Instead, the block came from firewall and bot-management rules. Blend says security teams often set these rules without input from the people who own revenue.
There’s a fair caveat. Blend used its own unbranded browser, not ChatGPT or Gemini, so named agents may fare better. Even so, it shows how easily a “block unknown bots” rule can catch an AI shopping agent working for a real customer.
Run this quick check:
- Open yourdomain.com/robots.txt. Make sure OAI-SearchBot isn’t blocked. OpenAI’s crawler guide says crawling stops at once when robots.txt says no.
- Let crawlers reach your product images, too. OpenAI warns that a 403 error on your image host can block feed processing.
- Email your CDN or firewall provider. Ask which AI agents are on the allow list.
- Keep cart, checkout, and account pages closed to crawlers.
| Crawler | Owner | What it does | Allow it? |
| OAI-SearchBot | OpenAI | Builds ChatGPT’s search and shopping index | Yes |
| ChatGPT-User | OpenAI | Fetches pages live when a user asks | Yes |
| GPTBot | OpenAI | Collects data to train models | Your call |
| Googlebot | Feeds Search and AI Mode | Yes | |
| Google-Extended | Controls Gemini training use | Your call | |
| PerplexityBot | Perplexity | Builds Perplexity’s index | Yes |
A common mistake is treating every AI bot the same. Blocking GPTBot to keep content out of training is a fair choice. But blocking OAI-SearchBot removes you from ChatGPT’s shopping results, and that costs sales.
Checklist Part 2: Give AI Shopping Agents Product Data They Can Read
AI shopping agents pick products from structured facts, not hero images. So add a complete product schema, including shipping, returns, and barcodes, and serve it in the page’s raw HTML.
Blend could read product pages at 171 stores. Here is the number of stores that shared each key data field:
| Data field | Stores (of 171) | Share |
| Product schema | 141 | 82% |
| Price | 130 | 76% |
| Stock status | 124 | 73% |
| Shipping details | 38 | 22% |
| Returns policy | 34 | 20% |
| GTIN (barcode) | 31 | 18% |
The bottom three rows matter most in late November. Picture a shopper who asks, “Will my order arrive by December 20, and can I send it back?” If your page can’t answer, an AI shopping agent picks a rival that can.
Fix it in this order:
- Shipping and returns: Add shippingDetails and hasMerchantReturnPolicy to your product markup. Please include your holiday return window if you decide to extend it.
- Barcodes: Add GTINs so agents can match your listing to reviews and rival prices.
- Raw HTML: View the page source. If a script loads the price or stock, some crawlers won’t see it. A late-2024 Vercel and MERJ study, summed up by Alhena, found major AI crawlers didn’t run JavaScript. That data is two years old, but it still matters.
- Plain facts: Replace “cozy, premium feel” with “80% merino, 300 gsm, true-to-size fit, machine washable at 30°C.”
That last point counts more than it seems. Capgemini argues that when AI tools shop, product fit and value for money beat brand appeal. Clever copy won’t save a page with missing facts.
Checklist Part 3: Keep Black Friday Prices and Stock in Sync
Your feed, product page, and schema must show the same price and stock at all times. During flash sales, a mismatch can make an agent drop you or quote the wrong deal.
Black Friday prices move fast. As Botify points out, they can change from hour to hour. Meanwhile, AI shopping agents read your data from three places: your product feed, your live page, and your structured data. When those disagree, the agent can’t trust any of them.
The platforms are speeding up to match. For example, OpenAI’s product feed spec accepts updates as often as every 15 minutes, according to Forrester. Google Merchant Center can also turn off products whose feed price doesn’t match the page.
Use these steps:
- Schedule sale prices. Load Black Friday prices with start and end dates in your feed. Don’t edit them by hand at midnight.
- Refresh feeds more often. Please consider transitioning from daily to hourly updates, or even more frequently, for Black Friday week.
- Update delivery cutoffs daily. A stale “arrives by December 24” is worse than no promise at all.
- Drop fake urgency. A countdown timer that resets on reload won’t fool a machine. Worse, it may make your data look unreliable.
Here’s the trade-off. Real-time feeds take developer time that many small stores don’t have in November. For them, scheduled sale prices and a higher refresh rate deliver most of the benefit. Most platforms let you set sale dates without writing any code. Save the bigger build for January.
Checklist Part 4: Open a Checkout Path for AI Shopping Agents
If you’re on Shopify, check your Agentic Storefronts settings today. If you’re not, Google’s Universal Commerce Protocol, set up through Merchant Center, is the main open route. However, Google is still rolling out access in phases.
| Route | Who it suits | Where products show | Status: Nov 2026 |
| Shopify Agentic Storefronts | Shopify stores | ChatGPT, Copilot, Perplexity, Google AI Mode, Gemini | On by default for eligible US merchants since March 2026 |
| Google UCP checkout | Merchant Center accounts | AI Mode in Search, Gemini | Early access, phased |
| OpenAI product feed | Any store via OpenAI’s merchant portal | ChatGPT | Feeds live; in-chat checkout scaled back |
Google unveiled the Universal Commerce Protocol on January 11, 2026. It was built with Shopify, Etsy, Wayfair, Target, and Walmart. Under UCP, you stay the merchant of record, so the customer relationship remains yours.
The platform gap is large. In Blend’s audit, Shopify stores averaged 91 out of 100 for agent readiness, while stores on other platforms averaged 59. Of the 38 stores with a live agent checkout endpoint, 37 ran on Shopify.
Now for the contrarian take. OpenAI scaled back its Instant Checkout plans this year. That’s a warning against spending your Black Friday budget on a custom agent checkout.
AI shopping agents will still direct most shoppers to your site, where they will complete their purchases. Therefore, guest checkout and express payments matter more right now. Blend found 148 of 171 readable stores already offer express payment on the product page.
Still choosing a platform for next year? Our comparison of the best ecommerce platforms for small businesses weighs agent readiness alongside cost.
Checklist Part 5: Track AI Traffic and Run Your 11-Day Sprint
Please establish a separate channel for AI referrals now, as doing so will allow you to measure Black Friday results. Then work through the fixes in order of impact.
Start with tracking. In Google Analytics 4, create a custom channel group for referrers like chatgpt.com, perplexity.ai, copilot.microsoft.com, and gemini.google.com. Next, check your server logs for visits from AI shopping agents.
Last year, HUMAN Security saw agent traffic to its ecommerce clients jump 144.7% from the pre-holiday baseline to the Black Friday weekend peak.
Then test like a shopper. Ask ChatGPT, Gemini, and Perplexity for your best-selling product type, such as “best waterproof hiking boots under $150 that ship to London by Friday.” Note whether your product appears and if the price and delivery date are correct.
Here’s a day-by-day plan:
- Nov 16–17: Audit robots.txt and firewall rules. Email your CDN provider.
- Nov 18–19: Fill schema gaps for shipping, returns, and GTINs.
- Nov 20: Please verify that the raw page source reflects the price and stock.
- Nov 21–22: Please load the Black Friday sale prices into feeds with the specified dates.
- Nov 23: Confirm your Shopify agent channels or Merchant Center status.
- Nov 24: Build the AI channel group in analytics.
- Nov 25: Run shopper-style prompts in three assistants and log errors.
- Nov 26–30: Please ensure code changes are frozen and verify feeds and logs daily.
Don’t skip the code freeze. A rushed theme update on Thanksgiving can break the schema you just fixed. As a result, every agent would read stale data on your most significant day.
Conclusion
Four moves matter most before Friday.
- First, ensure that your firewall isn’t turning away the assistants your customers use.
- Second, fill the data gaps that cost you comparisons: shipping times, return windows, and barcodes.
- Third, keep feed, page, and schema prices identical through every flash sale.
- Finally, please start measuring, as next year’s budget will depend on this year’s numbers.
AI shopping agents won’t decide Black Friday 2026 on their own. Most shoppers will still start on Amazon or Google. But the stores that agents can read cleanly will win a growing share of high-intent, low-return traffic. Meanwhile, the stores that block them won’t even know what they missed.
Your next step: open your robots.txt and your firewall dashboard today, and send that one email to your CDN provider.
Frequently Asked Questions
1. Should I block AI crawlers on my online store?
Block training crawlers like GPTBot if you like, but allow search crawlers such as OAI-SearchBot and PerplexityBot. Blocking search crawlers removes your products from AI shopping results. Furthermore, please verify the firewall rules, as Blend’s 2026 audit found that most blocks occur there.
2. How do I get my products to show up in ChatGPT shopping?
Allow OAI-SearchBot in robots.txt, then add a clear product schema with price, stock, and shipping. For richer listings, apply through OpenAI’s merchant portal and submit a product feed. Shopify stores can turn on ChatGPT through Agentic Storefronts.
3. Do AI shopping agents buy things without the shopper’s approval?
Rarely, for now. Most AI shopping agents conduct research and comparisons before directing the shopper to the checkout process. Capgemini found 38% of shoppers trust AI agents with routine purchases, but most still want to approve a buy before any money moves.
4. Does traffic from AI shopping agents convert better?
It does, according to Adobe. This holiday season, AI traffic will double its revenue-per-visit lead over other channels. Furthermore, 69% of shoppers say they’re less likely to return items bought with AI help, which protects your margins.
5. Do small online stores need to prepare for agentic commerce before Black Friday?
Yes, and small stores have an edge. In Blend’s 2026 audit, emerging retailers scored 75 for agent readiness, ahead of large chains at 63. Quick fixes like schema and robots.txt checks take hours, not weeks.

Senior Business Analyst / Prduct Owner with 12+ years of experience driving data-driven insights, optimizing business processes, and delivering strategic IT solutions.