How Can I Measure DIY AEO Results on Shopify?
Measure four layers, mostly with tools you already have. For visibility, keep a fixed list of 20 to 40 buyer questions and check them monthly in ChatGPT, Perplexity, Copilot and Google, recording whether your products are named. For accuracy, check whether the prices and details those answers show match your catalogue. For traffic, use AI referral sessions in Shopify's sessions-by-referrer report and in GA4's AI Assistant channel. For revenue, count orders from AI referrers and from Shopify's agentic storefront channels, and add a post-purchase "How did you hear about us?" question. Compare the products you changed with ones you did not. And expect undercounting, because many buyers who met you in an AI answer arrive later as direct visitors.
Why measure in layers?
Because AI search works like a funnel with leaks. A product can be named in an answer without anyone clicking, clicked without the referrer being recorded, and bought days later through a branded search. A single number, such as AI referral revenue, captures only the part that survives every step. Measuring each layer tells you where the work is paying off:
| Layer | Question | Free or low-cost source |
|---|---|---|
| Visibility | Are my products named for the questions buyers ask? | Monthly manual checks; entry-level trackers |
| Accuracy | Is what AI says about them right? | The same checks, compared with your catalogue |
| Traffic | Are AI answers sending visitors? | Shopify sessions by referrer; GA4 AI Assistant channel |
| Revenue | Do those visitors buy? | Shopify sales by referrer and channel; post-purchase survey |
How do you measure visibility for free?
With a spreadsheet and discipline:
- Write 20 to 40 questions buyers ask before purchasing, grouped by need, such as "best waterproof dog coat for a large breed" or "sustainable yoga mat under £60". Include some with your town or country if you sell locally.
- Freeze the list for at least three months, so results are comparable.
- Run each question monthly in ChatGPT with search on, Perplexity, Copilot and Google's AI Mode, in a logged-out or private session where possible.
- Record three things: whether any of your products are named, their position in the answer, and which sources are cited.
- Report the share of questions where you appear, by group, not individual wins.
Answers vary from run to run, so small month-to-month changes mean little. Look for movement across a whole group of questions. If the manual work becomes too much, low-cost trackers can automate it; options that plug into Shopify are compared in which Shopify apps track AI visibility, traffic and conversions.
How do you check accuracy?
Use the same monthly run. For every answer that names your product, compare the price, variants, key specifications and availability with your catalogue. Note errors such as an old price, a discontinued colour or the wrong material. Accuracy problems usually trace back to outdated feeds, marketplace listings or third-party pages, and fixing them can matter as much as being named, because a wrong price loses the sale.
How do you measure AI traffic in Shopify?
Shopify's reports show sessions by referrer. Look for referring domains such as chatgpt.com, perplexity.ai, copilot.microsoft.com, gemini.google.com and claude.ai. If you prefer a query, Shopify's ShopifyQL sessions schema has a referrer_domain dimension and a filter for human sessions:
FROM sessions
SHOW sessions, conversion_rate
WHERE human_or_bot_session = 'human'
GROUP BY referrer_domain
SINCE last_year UNTIL today
ORDER BY sessions DESC
In GA4, the AI Assistant channel groups visits whose referrer matches Google's list of assistants, and a custom channel group can add others and apply retroactively. ChatGPT adds utm_source=chatgpt.com to many of its links, which helps attribution even when the referrer is lost. A full setup is described in how to track ChatGPT and other AI referrals in GA4.
Google's own AI features are different: clicks from AI Overviews and AI Mode count as organic search. Search Console's generative AI report shows which pages earn impressions in them, which is a useful proxy for Google-side visibility.
How do you measure revenue?
Three sources, reported separately:
- Sales by referrer in Shopify, for orders that started with an AI referral.
- Agentic storefront channels. Shopify says orders from its AI channels, such as ChatGPT, Copilot and Google's AI features, show their channel or referrer in the order's attribution.
- Self-reported source. A post-purchase survey asking "How did you hear about us?", with "an AI assistant such as ChatGPT" as an option. This catches buyers who arrived by typing your name after reading an answer.
Report the first two as measured AI revenue, and the survey share as an estimate of the wider influence. Do not add them together, since they overlap.
How do you know your changes caused the results?
Use a simple control. When you improve product pages, such as rewriting descriptions, adding specification tables or fixing feed attributes, change one group of products and leave a similar group alone. After eight to twelve weeks, compare visibility, AI sessions and sales between the two groups. Also note outside events on the same timeline: seasonal demand, promotions, price changes and AI platform updates. A change that lifts both groups equally is probably seasonal or platform-wide, not your work. Separating AI gains from ordinary SEO gains is discussed in how to tell whether AI visibility gains came from SEO or AEO.
Which free checks help before you measure?
- Shopify's agentic readiness scanner, which audits the structured data AI agents read on a product page, is covered in Shopify's free AI shopping readiness tool.
- Search Console and Bing Webmaster Tools, for indexing, AI impressions and Copilot citations, as described in whether Search Console and Bing Webmaster Tools can track AEO for free.
- A crawler check that your robots.txt and theme do not block AI crawlers.
What results are realistic?
For most stores, AI referrals are a small share of sessions: Conductor's benchmark put AI referrals at about 1% of website traffic across industries. The early signal is usually visibility: your products starting to appear for a handful of questions, and fewer wrong prices in answers. Traffic and revenue follow more slowly and are undercounted. Judge progress over a quarter, not a week.
What might this look like in practice?
Consider an illustrative pet accessories store, Corvstead, on Shopify. It set up a sheet of 30 buyer questions and found its products named in 4 of them in ChatGPT and 2 in Perplexity, with two answers quoting an old price. It rewrote 40 product pages with sizing tables and materials, left 40 similar products unchanged, and fixed its Google Merchant Center feed. After three months, its products appeared in 11 of the 30 questions, the changed products had three times the AI referral sessions of the unchanged group, and 6% of post-purchase survey answers named an AI assistant.
What do people ask about measuring AEO on Shopify?
Can Shopify show traffic from ChatGPT?
Yes. Shopify's sessions-by-referrer data shows referring domains such as chatgpt.com.
Do I need a paid tool to measure AEO?
No. A fixed question list checked monthly, plus Shopify and GA4 referral data, covers the basics.
Why do AI referrals look so small?
Many buyers read an AI answer and arrive later by typing your name, so the visit counts as direct.
How long before results show?
Visibility changes can appear within weeks; traffic and revenue usually take a quarter to judge.
Published by The Kovmere Team · updated Sep 29, 2026 · More resources