Now Track Where ChatGPT, Perplexity & Google AI Actually Recommend Your Products

By Sean Breeden|June 17, 2026|7 min read

Until now, Krytho answered one question: can AI shopping engines read and trust your products? That is readiness, and it is essential. But it is only half the story. The question that actually keeps store owners up at night is the next one: do they actually recommend me?

Today that half is live. Krytho now tracks where your products get cited and recommended across ChatGPT, Perplexity, and Google AI, surfaced in a new AI Visibility view on your dashboard.

Readiness is the input. Citations are the result.

Think of it as two halves of the same job. The audit tells you whether your product data is structured, complete, and crawlable enough for an AI engine to use it. Citation tracking tells you whether that work paid off: whether, for the searches your shoppers actually run, the AI surfaces your products. Fix readiness, then watch citations climb. That feedback loop is the whole point.

What we track now, and how

Three engines, each checked the most faithful way available:

ChatGPTvia OpenAI's live web search, the closest programmatic read of ChatGPT's web-grounded answers.
Perplexity via its API, including the sources it cites.
Google AI Overviews via the AI Overview block and its cited links.

For every buyer-intent query you track, Krytho checks whether your domain appears in the answer or its citations, and at what position.

What you see

Appearances per engine over the last 30 days, plus an 8-week trend so you can see momentum.
The exact searches you appeared in: the real queries where AI surfaced your products, tagged by engine.
Test any query, live. Type a search and instantly see which engines cite you right now.
Suggested queries. Krytho generates buyer-intent searches from your own products so you can start tracking in one click.
Share of voice. How often you appear versus the competitors you choose to track.

An honest note on measurement

We will always tell you what a number really means. AI answers are personalized and non-deterministic, and these APIs are the closest available proxies for the consumer apps, not a perfect mirror of every shopper's screen. So treat citation tracking as a representative, sampled panel you follow as a trend, not an exact count of every impression. That is the most honest read available, and it answers the question that matters: are you showing up, and is it getting better? (See the glossary for how we define AI citations and share of voice.)

Why this matters now

Shoppers are shifting from searching to asking, and the platforms know it. Our study of 575 stores found more than 1 in 3 Shopify product pages carry no Product JSON-LD at all, which means most catalogs are not even readable by these engines yet. The stores that get readable first, and can prove they are being recommended, will own the shelf as AI shopping grows.

See where AI recommends your store.

Start with a free readiness scan, then track your citations across every engine.Run a free AI readiness scan

How to get started

Citation tracking is on the Growth and Agency plans. Add the buyer-intent queries you care about (or use the suggested ones), run a check, and add a couple of competitors for share of voice. Weekly automatic checks build your trend from there. Not sure if AI can read you yet? Start free at krytho.com/scan.

Is your store ready for AI shopping?

Paste any product URL. Instant, free results showing exactly what AI shopping engines see.Run a free AI readiness scan

Frequently asked questions

How is citation tracking different from the audit?

The audit measures readiness: whether AI engines can read and trust your product data. Citation tracking measures the outcome: whether ChatGPT, Perplexity, and Google AI actually cite and recommend your products for real buyer-intent searches. Readiness is the input; citations are the result. You want both.

Which engines do you track, and how?

Three. ChatGPT via OpenAI's live web search (the closest programmatic read of ChatGPT's web-grounded answers), Perplexity via its API, and Google AI Overviews via the AI Overview block and its cited sources. For each tracked query we check whether your domain appears in the answer or its citations, and at what position.

Is this exactly what every shopper sees?

No, and we are deliberate about that. AI answers are personalized and non-deterministic, and these APIs are the closest available proxies for the consumer apps, not a perfect mirror. So treat it as a representative, sampled panel that you track as a trend over time, not as an absolute count of every impression. It is the most honest read available, and it is directionally what matters: are you showing up, and is it improving.

What is share of voice?

Across your tracked queries, share of voice is how often your store appears versus the competitors you choose to track, found in the same AI answers. It tells you not just whether you appear, but how you stack up against the specific rivals you care about.

Which plans include it, and how do I start?

Citation tracking is on the Growth and Agency plans. To start, add the buyer-intent queries you care about (or let Krytho suggest them from your products), run a check, and add a few competitors for share of voice. Weekly automatic checks then build your trend over time. Run a free readiness scan first at krytho.com/scan.