See how your products compare
Compare how your products appear, which competitors are named and which sources support the answers. Then test the strongest gaps.
Three ways a catalogue goes invisible
All three are silent. The store works, the pages load, and nothing tells you the crawler saw something different.
Product pages an assistant cannot read
Price, specs, variants and reviews can be rendered by JavaScript. Some crawlers may receive less information than a shopper sees, so we compare the raw response with the rendered page instead of assuming they match.
Missing or invalid Product schema
Valid Product and Offer markup gives an engine another machine-readable source for identity, price and availability. It can improve understanding or eligibility, but it does not determine recommendations by itself.
Crawl budget spent on filters
Faceted navigation generates a near infinite URL space. Left unmanaged, a crawler spends its visit on colour and size combinations and never reaches the four hundred products you actually want indexed.
What we check on a store
Render gap, measured on a real PDP
How many words of your product page exist in the raw HTML before hydration, plus the framework markers that explain why. This is the check that most often changes the conversation.
Structured data, validated on the live response
Whether Product, Offer, price, availability and review markup are present and parse, checked in the deployed HTML rather than a preview that has never been served to anyone.
AI crawler access, per bot
GPTBot, OAI-SearchBot, ClaudeBot, PerplexityBot, Google-Extended, CCBot and Bytespider, evaluated against your live robots.txt with real group precedence rather than a top to bottom read.
Fixes shipped to your store
On a managed plan our team ships the rendering, schema and template work into your stack, and the before and after sits in the monthly report against the exact change.
The six that are specific to catalogues
Take this list to whoever owns your templates. Each one is a template level decision, which is why fixing it once fixes it across every product you sell.
Variant URLs
Every colour and size resolving to its own indexable URL, competing with the parent product for the same query. Canonical to the parent unless the variant is genuinely a different purchase.
Out of stock handling
A 404 on a temporarily unavailable product throws away every signal that URL earned. Keep the page, mark availability in schema, and say when it returns.
Thin category pages
A category page that is only a grid gives an engine little category context. Useful copy, headings and internal links can make the page easier to interpret without getting in the shopper’s way.
Faceted navigation
Filters can produce a crawlable URL for every combination. Decide which facets deserve an indexable page and consolidate the rest so crawlers spend more time on the products and categories that matter.
Duplicate descriptions
Manufacturer copy repeated across retailers gives an engine little unique evidence for choosing one listing over another. Check it alongside price, availability, reviews and the authority of the page.
Pagination markup
Deep listings without a clear paging signal leave products on page seven effectively undiscoverable, however well the first page performs.
Questions ecommerce teams ask
Does this work on Shopify?
Yes, and the audit is stack agnostic because it reads the live response rather than the platform. Shopify themes usually pass the render gap check, and the recurring findings there are duplicate manufacturer descriptions, variant URLs and thin collection pages.
We are headless. Is that a problem?
It is worth checking early. Some headless builds serve less product information in the initial response than a shopper sees after JavaScript runs. Comparing both versions shows whether your site has that gap; it does not predict a recommendation by itself.
How many products can you audit?
The crawl limit is per plan and published: 50 pages on Individual, 100 on Startup, 500 on Enterprise. For a large catalogue the useful approach is auditing one page of each template rather than every product, because the findings are template level almost every time.
Does being cited by AI actually sell anything?
A citation can put a product in a buyer’s consideration set, but visibility is not the same as a sale. The useful test joins prompt appearances and referral traffic to the store’s own conversion data instead of assuming an uplift.
Will you touch our live storefront?
Only with scoped, revocable keys and only after you approve the change. Nothing ships without sign off, every change is individually reversible, and you can revoke access in one action without losing the work already shipped.
What does the audit actually check?
36 checks against the live response, covering crawlability and indexing, on page markup, structured data, AI and GEO signals, content depth, media, performance and security. The full list runs on any URL on a free account.
Start with one product page
Start with a measured competitor comparison. If it exposes a reproducible product-page gap, we will show the evidence and explain how we would test it.