Jobs to be done · Monitor · Ecom · SEO · Agency

How to track competitor products in AI answers.

Q: Is it ethical to track competitor products in AI answers?

Yes. The tracking queries publicly available AI answers with buyer-style prompts and reads public PDP data — no scraping of private dashboards, no auth bypass. It is the same observable-shelf analysis competitive teams have always done with search results and retail shelves, made systematic.

Q: What triggers a big AI recommendation bump for a competitor?

The recurring patterns in eCommerce Insights data as of early 2026: a PDP rewrite with stronger answer coverage, a review campaign pushing aggregateRating past a threshold, a schema cleanup (gtin, material, color fields), strong new internal links from a collection page, or third-party coverage that AI crawlers cite. Sometimes there is no identifiable driver — engine retrains happen, and the report says so rather than inventing a cause.

Q: How many competitor SKUs can I watch?

Plan-dependent: Starter covers a starter watchlist, Growth expands it substantially across more competitor brands, and Agency & Enterprise customizes limits for agencies running per-client watchlists. Most brands get the signal they need from 20–50 well-chosen SKUs across three to five rivals.

Q: Can I see which engine a competitor gained on?

Yes — the change report breaks movement down per engine. A rival gaining on Perplexity but not ChatGPT usually means a retrieval-specific cause (often third-party citations), which tells you where to focus the response.

Q: What should I do when a competitor takes my slot?

Run the side-by-side: their PDP versus yours on schema, answer coverage, review signal, and crawl access. The gap list is the response plan, and the watchlist tells you within a week whether your fix flipped the slot back.

Your competitor just got a big AI recommendation bump and nobody can say what changed. Without a systematic watch on rival products, the first signal you get is the revenue dip — three months after the slot was lost and long after the fix window closed.

Quick answer

Add competitor product URLs to an eCommerce Insights watchlist. Citation count, position, and engine mix are tracked on every refresh, and the weekly change report flags meaningful movers with the likely driver named. Part of SKU-level tracking; the mechanism behind the bumps is in how AI engines pick which products to cite.

The slow way: bookmarks and vague concern

The manual version is a calendar reminder that says "check ChatGPT about the category." Once a month — or whenever someone remembers — you rerun last time's prompts, eyeball which competitors appear, and feel vaguely concerned or vaguely relieved. That is anecdote, not tracking. The signal you actually need — which rival doubled their Perplexity citations over three weeks, and what they changed — never surfaces, because manual sampling is too noisy to detect deltas.

The systematized manual version is a shared sheet: prompts down the side, competitors across the top, ones and zeros each week. It works for about six weeks. Then the analyst who owned it gets busy, two weeks go unfilled, and the sheet quietly dies. Competitive tracking is a recurring-work problem, and recurring work without automation has a half-life measured in sprints.

The eCommerce Insights way

Pick the SKUs that actually compete. Start from your first share-of-voice scan: the competitor products already appearing in AI answers for your prompt set. Skip the paid-media rivals that never show up in answers — they are not competing for these slots.
Add them to the watchlist. Paste product URLs. eCommerce Insights reads the public product data to identify SKU, category, and price, and starts tracking on the next refresh. Tag by competitor, category, or price band.
Set the cadence. Weekly by default; daily for hero competitor SKUs during launch windows on Growth.
Read the weekly change report. Citation count, average position, and engine mix per watched SKU, with big movers flagged at the top and a likely driver annotated — "PDP rewritten March 10, FAQ block and material metafields added." Where no driver is identifiable, the report says so; engine retrains happen.
Respond with your own SKU. A rival gaining ground is usually a fixable gap on your side; a rival losing ground is a slot opening. Each major mover links the matching SKU in your catalog and the fix that would contest the slot — the side-by-side workflow from find out why ChatGPT recommends a competitor.

For the aggregate version of this question — your share of the whole answer surface versus theirs — see compare my brand's AI share of voice.

What "good" looks like

Watchlist size covering the brands that actually win AI slots20–50 SKUs

Normal week-over-week drift in competitor citation counts<10%

Response playbook ready for big moversyes

Quarterly pattern log of who rises on which enginekept

Drift thresholds from eCommerce Insights tracking as of early 2026, illustrative. The pattern log compounds: after two quarters you know which rivals respond to review campaigns, which win on schema, and which only move when an engine retrains — which is strategy, not trivia.

Ask AI about this job

Have your favorite AI engine apply this walkthrough to your category.

Frequently asked questions

Is it ethical to track competitor products in AI answers?

Yes. The tracking queries publicly available AI answers with buyer-style prompts and reads public PDP data — no scraping of private dashboards, no auth bypass. It is the same observable-shelf analysis competitive teams have always done with search results and retail shelves, made systematic.

What triggers a big AI recommendation bump for a competitor?

The recurring patterns in eCommerce Insights data as of early 2026: a PDP rewrite with stronger answer coverage, a review campaign pushing aggregateRating past a threshold, a schema cleanup (gtin, material, color fields), strong new internal links from a collection page, or third-party coverage that AI crawlers cite. Sometimes there is no identifiable driver — engine retrains happen, and the report says so rather than inventing a cause.

How many competitor SKUs can I watch?

Plan-dependent: Starter covers a starter watchlist, Growth expands it substantially across more competitor brands, and Agency & Enterprise customizes limits for agencies running per-client watchlists. Most brands get the signal they need from 20–50 well-chosen SKUs across three to five rivals. See pricing.

Can I see which engine a competitor gained on?

Yes — the change report breaks movement down per engine. A rival gaining on Perplexity but not ChatGPT usually means a retrieval-specific cause (often third-party citations), which tells you where to focus the response.

What should I do when a competitor takes my slot?

Run the side-by-side: their PDP versus yours on schema, answer coverage, review signal, and crawl access — the workflow in find out why ChatGPT recommends a competitor. The gap list is the response plan, and the watchlist tells you within a week whether your fix flipped the slot back.

Related jobs

Monitor

Know what changed before the revenue tells you.

Competitor product watchlists with the likely driver named, weekly.

Start free trial See SKU-level tracking