Solutions / Ecommerce and DTC
AI citation tracking for product brands: who the engines cite when a shopper asks
When someone asks an assistant for the best product in your category, the answer is assembled from review sites, forum threads and round-ups. Answer Radar records which ones, flags which are yours, and tells you when the page that decides the answer is not on your site.
Updated 2026-09-04
What is the same, and what we have not measured
The instrument is identical to the one we run for software companies. The sample is not.
Answer Radar asks ChatGPT, Gemini, Perplexity and Claude the buying questions you approve, records whether your brand was named or recommended, and stores every cited domain with a flag for yours, a tracked competitor's or neither. That works for a grinder as well as for a CRM. What we do not have is an ecommerce run to show you: 19 of the 310 people who named an industry during onboarding said ecommerce, and every number we publish comes from B2B software projects. This page tells you what changes for a product brand and prints our own B2B citation data as the example, labelled as such.
Citation share, from our own B2B run on 4 September 2026
The metric a product brand reads first. Each cited domain counted once per answer that cites it, flagged as the customer's own or a tracked competitor's. Ours is a software category; yours will name different domains, which is the point.
answer_metrics_get({
project_id: "<your project id>",
run_id: "<latest run id>"
})
// read: segments[].metrics.citation_share.domainsThrough the Linkeddit MCP connector, or the project's Overview. Definition version visibility-metrics-v1.
citation_share, top domains per engine
- Gemini
- reddit.com 5 of 105 citing samples (4.8%), then alpha-sense.com, autobound.ai, linkeddit.com (is_customer true) and siftly.ai at 3 each. 73 further domains omitted from the summary.
- ChatGPT
- cdn.featuredcustomers.com, g2.com, keeptabz.ai and outmano.com at 2 of 33 each. linkeddit.com absent. 24 omitted.
- Claude
- alhena.ai, g2.com and linkeddit.com (is_customer true) at 2 of 35 each. 27 omitted.
- Perplexity
- linkeddit.com 2 of 29 (is_customer true), then aiclicks.io, aimultiple.com, airops.com, blog.hubspot.com at 1 each. Three admissible samples only; ten excluded as thin.
- definition
- Each domain's share of all cited domains in the segment, counting a domain once per sample that cites it. Flags whether a domain is the buyer's own or a tracked competitor's.
Read the flags before the shares. A review site at the top of your list is a page you cannot edit, and that changes what the fix is.
Three things that change when the product is physical
Same measurement, different reading. Each comes from what our B2B runs showed about how engines assemble a recommendation.
The question is a product, not a category
A B2B buyer asks for the best tool for a job. A shopper asks for the best espresso grinder under a price, or whether a named brand is worth it. Write the prompt set the way a shopper types, with the price band and the use, and approve the ones that carry impressions in Search Console.
The cited sources are review sites and forums
On our own run Gemini cited reddit.com more than any other domain and ChatGPT cited g2.com. For a product brand the equivalents are Trustpilot, Reddit threads and editorial round-ups. The citation share metric flags each cited domain as yours, a tracked competitor's or neither, so you can see which review page is speaking for you.
The fix is often not on your site
When the engine cites a review profile or a forum thread, the page that decides the answer is not yours to edit. The content gap view still names your closest page; the citation list tells you which third-party page needs a real review or a real answer from you, through the review site's own process.
The demand we can see for this page
From our own Search Console, 6 June to 3 September 2026. Small, specific, and unserved by a product page until now.
- “ai citation tracking for ecommerce and reviews”: 53 impressions at an average position of 41, landing on a blog listicle.
- “best ai brand monitoring tools for ecommerce”: position 7 on one impression.
- “best profound alternative for shopify / dtc brands”: position 9 on one impression.
We print the demand rather than a market size because the market size would be a guess and the impressions are not.
What this will not do for a product brand
So you can decide with the gaps in view.
- It will not report revenue or traffic from AI answers. The engines pass little or no referrer; attribution stays in your analytics.
- It will not show you an ecommerce benchmark, because we do not have one yet. The first public one will be a customer's, with permission, or ours.
- It will not edit a review site. When the cited page is Trustpilot or a Reddit thread, the fix is a real review or a real answer through that site's own process, and the content gap view names the page on your site that should carry the rest.
- It will not measure Amazon or marketplace answers. The four engines are ChatGPT, Gemini, Perplexity and Claude, with Google AI Overviews arriving with its engine.
Frequently asked questions
Does AI citation tracking work for ecommerce and product reviews?+
Yes, with one honest caveat: the measurement is the same one we run for software companies, and every number we publish today comes from B2B projects. Answer Radar asks ChatGPT, Gemini, Perplexity and Claude the questions a shopper types, records whether your brand was named or recommended, and stores every cited domain with a flag for yours, a competitor's or neither. For a product brand the cited domains are usually review sites and forum threads, and the citation share metric is built to show exactly that.
Why do you not show an ecommerce example?+
Because we do not have one. Of the 310 people who named an industry during onboarding, 19 said ecommerce, and none has run a public project we can quote. The example on this page is the citation share block from our own B2B run, with its definition, and the section above it explains what changes for a product brand. We would rather print that than a coffee-brand result we made up.
Which review sites matter for AI answers about products?+
The ones the engines actually cite on your prompts, which the run tells you. In our B2B data that was g2.com on ChatGPT and reddit.com on Gemini. For consumer products expect Trustpilot, Reddit and editorial round-ups, but measure it rather than assume it: the fan-out view shows the searches the engine ran, and searches that name a review site tell you which one it trusts in your category.
Can it tell me how much revenue AI answers drove?+
No. The engines pass little or no referrer, so any revenue figure attributed to them is inference from landing pages and codes. Answer Radar measures what is observable, named, recommended and cited, and reports the change between runs. Attribution stays in your analytics.
Is this a Profound alternative for a Shopify brand?+
It is a different shape of product. Profound's entry tier covers one engine; Answer Radar measures four from one prompt set and carries the gap through a fix and a re-measurement, and it is included in the Compete plan with the weekly competitor brief that mines review sites for complaints about the brands you track. Whether that is the right shape depends on whether you want a tracker or a loop. Our reviews of the trackers are linked below.
What does it cost?+
Answer Radar is included with the Compete plan at $99 a month. Runs draw on a monthly credit allowance, one credit per question asked of one engine, and the usage tool shows the position before you start.
Read next
Query fan-out
The searches the engines ran, including the ones that name a review site.
Read CompareAI visibility tools compared
Fourteen trackers in one matrix, with pricing fetched on the day and a measured section.
Read Also forPR and brand teams
The same measurement read for accuracy and sources rather than recommendation.
ReadFind out which review page is speaking for your brand
Answer Radar is included with the Compete plan, alongside the weekly competitor brief that mines review sites for the brands you track.