Shopify Now Lets Merchants Track AI Sales. Here's What It Means.
Last updated: June 2026
AI distribution is the practice of making your product discoverable, quotable, and purchasable directly inside AI assistants like ChatGPT, Claude, and Gemini. In June 2026, Shopify shipped a set of tools that let millions of merchants measure exactly that: which AI platforms are driving sales, what shoppers are asking AI in their category, and where their product data falls short (Modern Retail, June 2026).
On its own, that reads like a routine analytics update. It is not. Shopify powers a large share of global e-commerce, and it just decided AI traffic was worth a dedicated measurement product. That is the clearest signal yet that AI has become a distribution channel companies are expected to track and optimize, the same way they track and optimize search.
For services companies, the signal comes with a catch. Shopify is built for products: things with a SKU, a fixed price, and a checkout. If you sell insurance, a loan, or any service where the price is worked out per customer rather than printed on a label, no Shopify is coming to hand you the same dashboard. The real question this launch raises is what AI distribution even looks like when you cannot just list a product at a price.
What Shopify actually launched
Shopify rolled out three tools inside its admin panel, all aimed at the same question its merchants kept asking: is AI sending us business, and can we see it?
| Tool | What it does |
|---|---|
| AI channel dashboard | Tracks orders, sales, and conversions from AI platforms, including ChatGPT, Google Gemini, Microsoft Copilot, and Shop, in one place |
| Search Intelligence | Shows the AI shopping queries common in a merchant's category and whether their products appear, then flags data gaps (missing descriptions, attributes) that limit visibility |
| Knowledge Base | Surfaces the questions AI assistants ask about a business that go unanswered, such as store locations, policies, or bulk ordering, so the merchant can answer them directly |
Shopify’s director of product, Aaron Glazer, put the motivation plainly: “One of the questions merchants have is, how does an agent discover my product?” (Modern Retail, June 2026).
Shopify's agentic sales dashboard: AI revenue and sessions by channel, plus the searches where products appeared. Source: Shopify (Spring '26 Edition).
Why this matters more than it looks
These tools were not a quiet update. They shipped in Shopify's Spring '26 Edition, announced on 17 June 2026 with more than 150 updates (Shopify Editions). Editions is Shopify's flagship showcase, where it stages its biggest bets. Putting AI measurement in that release, next to the headline features, says how seriously Shopify is taking the channel.
Shopify does not build a measurement product for a channel it considers a rounding error. The usage data behind the move is moving fast. AI chat interfaces still account for less than 1% of overall web traffic, but the trend line is steep: AI referrals now drive more than 1.5% of traffic to Walmart and Target, up from under 1% a year earlier (Modern Retail, June 2026). Adobe found AI-referred traffic to US retail sites grew more than 800% year over year heading into late 2025 (Adobe Analytics, 2025). On the consumer side, 41% of shoppers now use AI assistants for product research, 33% for reviews, and 31% for deals.
What stands out is the order of operations. Shopify already shipped the selling side: agentic storefronts went live for millions of merchants in March 2026, and it is building the Universal Commerce Protocol with Google. The measurement layer comes on top of all that. That sequence is the tell. You instrument a channel when you expect people to actively work it, not when you expect it to run itself.
So here is the shape of AI distribution Shopify is quietly confirming, and it holds for everyone, not just its own merchants. The AI platform handles discovery. The merchant handles the conversion, the customer relationship, and the job of proving any of it happened. The platform surfaces demand. Capturing it is your problem.
What this means for services without a SKU
Here is where it diverges for anyone selling insurance, banking, credit, or any complex, quote-based service.
Shopify’s tools work because product retail has a shape software can read. Every product is a SKU: a fixed item, with a fixed price, a buy-now path, and an order that either happened or did not. That structure is what makes the tools possible. Search Intelligence can check whether a specific SKU shows up for “best running shoes under 150.” The dashboard can count the order. The Knowledge Base can fill a static gap like store hours.
Services break that structure in the place that matters: the price. An insurer can list a dozen products and tiers, but it cannot list the price or bind the policy until it has assessed the specific risk. Home, health, a business loan: each one is individually rated and underwritten, so the number the customer actually pays is calculated from their situation, not printed on a label. Even the simple personal lines that bind in minutes today are never one click. There are risk questions, eligibility checks, and mandatory pre-contractual disclosure (the IPID, in the EU) before anyone is covered, and the more complex lines add underwriting and human follow-up on top. There is also no single countable order; there is a quote, then a qualified lead, then a sale that often closes days later by phone, on your own site, or through a broker.
Picture the actual moment. Someone asks ChatGPT, “I just rented a two-bedroom flat in Madrid, what home insurance do I need and roughly what would it cost?” There is no listing to return. A useful answer has to ask about the property, the contents, and the tenant, then produce a real estimate and explain what is covered. Shopify’s dashboard has nothing to count here, because nothing was listed and nothing was ordered. What happened was a conversation that ended in a qualified lead, or in nothing at all. The unit Shopify’s whole model rests on does not exist.
So every one of Shopify’s tools has a services equivalent, and not one of them looks the same:
| What Shopify gives products | Why it does not transfer to services | The services equivalent |
|---|---|---|
| Dashboard counts AI-driven orders | Services have no single countable order; the outcome is a quote and a lead, and the sale usually closes later by phone, email, or on your own site | Full-funnel analytics: conversations, quotes, leads, and conversions attributed back to the AI channel |
| Search Intelligence checks if your SKU appears for a query | Services are not retrieved as listings; the AI recommends a provider inside a conversation and describes your offer in its own words | Recommendation monitoring: synthetic buyers testing who the AI recommends, in what order, at what price, under which prompts |
| Knowledge Base fills static gaps like store hours | The questions are personalized ("what would this cost for my situation?") and cannot be answered from a fixed list | A conversational AI app that answers live, generates a personalized estimate, and qualifies the customer in the moment |
For services, in other words, the product feed becomes a conversation and the order becomes a lead. You cannot list your way into this channel. You need something that can hold the conversation, run the quote, and capture the lead, plus an analytics layer built for quotes and leads rather than orders.
That second half, the measurement, is the part most services companies get wrong first. Default web analytics badly undercount AI-sourced traffic, partly for the reason above: when the sale closes by phone or on your site, the last click looks like a phone call or a direct visit, and the AI conversation that started it never gets credit. Standard Google Analytics and UTM tracking pick up only a sliver of the channel, while asking customers directly how they found you surfaces several times more (WaniWani attribution analysis, 2026). A Shopify merchant now sees AI sales in a dashboard. A services company looking only at GA sees a channel that appears not to exist, and quietly concludes it is not worth the effort. That conclusion is expensive, because AI-sourced leads tend to convert better than traditional search (Morningstar / PR Newswire, 2026): by the time someone has been quoted inside an AI conversation, they are already informed, qualified, and looking at real numbers.
There is one signal that arrives even earlier: the bots. AI assistants reach you through crawlers that read your content and agents that call your app, and they identify themselves in your server logs, GPTBot, ChatGPT-User, PerplexityBot, ClaudeBot, Google-Extended, and the rest. Watching those hits tells you whether AI can even find and parse you, before a single human shows up. A service running its own AI app sees more here than a retailer does: not just that an agent appeared, but which questions it asked and where the conversation stalled. The catch is that not every agent is a customer. Some are competitors’ bots pulling your quotes to reverse-engineer your pricing, so for services, bot tracking is part measurement and part security.
What to do if you sell a service, not a product
Shopify just wrote the retail playbook down. Services do not get the same tools, but the lesson transfers, and most of it is worth doing no matter who you buy from.
- Treat AI as a channel you own, not one you wait on. Shopify just told millions of merchants to work their AI presence the way they work search. The same logic applies to services; the surface is just a conversation rather than a product feed.
- Build something that can hold a conversation. For quote-based products, the unit of distribution is an app that can answer real questions, generate a personalized estimate, and capture the lead, deployed where customers already ask: ChatGPT, Claude, Gemini. Discovery happens on the platform; the conversion has to happen somewhere you control. If you distribute through brokers or aggregators rather than direct, the same app can run discovery and qualification in the conversation, then hand the qualified lead to your existing network instead of closing the sale itself.
- Measure the channel before you judge it. If your only view is GA, AI will look negligible and you will under-invest in it. Track the full path, from the conversation to the quote to the lead, or you are guessing about the channel your fastest-moving competitors are already in.
- Build compliance in from the start. A retailer flagging a missing product description is not the same as an insurer running a demands-and-needs check and serving the right pre-contractual disclosures inside an AI conversation, the kind of thing the IDD already requires of any insurance distribution. For regulated services, owning the experience is not a nicety; it is a requirement.
This is the gap WaniWani was built to fill for services: the conversational app plus the measurement and compliance layer underneath it, the services counterpart to what Shopify just shipped for products.
So what does this actually mean for services?
Strip it back. Shopify just confirmed, across millions of merchants, that AI is a channel worth measuring and worth working. The tools it shipped happen to fit products. The conclusion underneath them fits anyone who sells.
For services there is no packaged version on the way. The provider that shows up inside the conversation with an accurate, personalized answer is the one that earns the lead. The provider that cannot be reached is not in the running, and over time it stops being mentioned, because assistants lean on the sources and tools that actually answer.
Retail is roughly twelve to eighteen months ahead on this curve, and what happens in retail keeps previewing what comes for services. The services companies that treat AI as a channel they own, and measure it honestly, will have built a position before most of their competitors notice the channel exists.
If you sell a quote-based service and want to see what owning that channel looks like on your own products, WaniWani can show you.
Frequently Asked Questions
What did Shopify launch for AI shopping?
Shopify launched three tools inside its admin panel in June 2026: an AI channel dashboard that tracks orders, sales, and conversions from ChatGPT, Google Gemini, Microsoft Copilot, and Shop; a Search Intelligence tool that shows the AI shopping queries in a merchant’s category and whether their products appear; and a Knowledge Base that surfaces unanswered questions AI assistants ask about a business. Together they let merchants measure and optimize AI as a sales channel.
Why does Shopify launching these tools matter?
It is a scale signal, and the timing says it too. Shopify put AI sales tracking in its Spring '26 Edition, its flagship release, not the place a company parks a side experiment. By adding a dashboard for AI sales on top of the selling tools it shipped in its previous Edition, Shopify is telling millions of merchants that AI is a real channel they are expected to work, the same way they work search. AI chat is still under 1% of web traffic, but AI referrals to retailers like Walmart and Target have roughly doubled in a year.
Do Shopify’s AI tracking tools work for services like insurance or banking?
No. Shopify’s tools assume product retail: a SKU, a fixed price, a buy-now path, and an order that can be counted. Services are sold through a conversation that generates a personalized quote, handles disclosures, and captures a lead. None of those map onto a product feed, so insurers, banks, and lenders have to build or buy their own conversational app and analytics layer rather than switch on a dashboard.
How do you measure AI sales for a service with no checkout?
You measure the funnel, not the order. Because there is no single transaction to count, you track the conversation, the quote generated, the lead captured, and the eventual sale, and attribute them back to the AI channel. This usually means combining server-side logs from your AI app with declarative attribution (asking customers how they found you), since default Google Analytics and UTM tracking capture only a fraction of AI-sourced traffic.
Should services track AI bots, not just AI sales?
Yes, and bot traffic is often the first signal available. AI assistants reach your business through crawlers and agents that identify themselves in server logs, including GPTBot, ChatGPT-User, PerplexityBot, ClaudeBot, and Google-Extended. Tracking those hits shows whether AI can find and read your content before any human arrives. If you run your own AI app, you also see the agent calls directly: which questions get asked and where conversations drop off. Bot tracking doubles as security, since some agents are competitors pulling your quotes rather than customers.
What is the services equivalent of an agentic storefront?
A conversational app, built on the MCP protocol, that runs the full discovery-to-lead journey inside the AI conversation. Instead of a product feed and a buy-now button, it understands the customer’s situation, generates a personalized estimate, explains coverage, handles regulatory disclosures, and captures a qualified lead. The merchant owns the app, the data, and the customer relationship across ChatGPT, Claude, and Gemini.
What should a services company do now?
Treat AI as a channel you own rather than one you wait on. Build a conversational experience for your quote-based products, measure the full path from conversation to lead before judging the channel, and build compliance controls in from day one, since regulated disclosures travel with services in a way they do not with retail. Early presence compounds, because assistants learn to route repeat questions to the tools that reliably answer them.
Sources:
- Modern Retail. “Marketplace Briefing: Shopify launches tools to help merchants track sales and traffic from AI platforms.” June 2026. https://www.modernretail.co/technology/marketplace-briefing-shopify-launches-tools-to-help-merchants-track-sales-and-traffic-from-ai-platforms/
- Adobe Analytics. “Online holiday shopping and the rise of AI-referred traffic.” 2025. https://business.adobe.com/blog/the-latest/adobe-analytics-record-online-holiday-shopping
- Morningstar / PR Newswire. “New research finds AI-referred traffic converts higher.” 2026. https://www.morningstar.com/news/pr-newswire/20260218ph89495/new-research-finds-ai-referred-traffic-converts-3-6-higher
- Shopify. “Shopify Editions.” 2026. https://www.shopify.com/editions