An AI visibility audit that speaks Magento.
Most AI visibility tools assume your catalog is Shopify-shaped, so their advice arrives in metafields and Liquid. Adobe Commerce (Magento) catalogs run on attribute sets, CMS blocks, and configurable products — and that's the vocabulary eCommerce Insights scores in and writes fixes in.
Complex catalogs hide answers from AI crawlers.
Structured data exists — it just isn't rendering.
Material, country of origin, and dimensions often live as product attributes that never reach the storefront because a storefront property is off or the attribute isn't in the right attribute set. The data engines need is in the database; the engines can't see it.
Block-of-blocks layouts bury the answer.
Long-form PDP content nested three CMS blocks deep — sometimes behind deferred rendering — reads as thin or empty to crawlers. The audit unrolls the block hierarchy and shows which block carries the answer that's missing from the crawl surface.
Default themes ship incomplete Product JSON-LD.
Stock Adobe Commerce emits partial Product schema, and SEO suites (Mageworx, Mirasvit, Amasty) layer more on top. What the live page actually serves is what gets audited — typically missing the offers block, brand entity, or identifier fields engines use to disambiguate.
Both scores, keyed to the admin you actually use.
| Check | Score | Adobe Commerce surface |
|---|---|---|
| Product JSON-LD completeness on the live page — offers, brand, identifiers | Both | theme / SEO suite |
| Attribute coverage and storefront visibility properties | Citation | attribute sets |
| Configurable vs. child SKU resolution — own URL, own content | Citation | configurables |
| CMS block content reachable in server-rendered HTML | Citation | CMS blocks |
| Answer coverage vs. buyer-phrased queries across six engines | Citation | PDP content |
| robots.txt admittance for AI crawlers | Agent | robots.txt |
| Machine-readable price, availability, returns and shipping policies | Agent | schema + CMS pages |
| Rendering — Luma, Hyvä, or PWA Studio hydration timing | Agent | theme layer |
PWA Studio storefront? The headless workflow renders the JavaScript first. Methodology: agent-readability score.
"Add this attribute to this attribute set" — not "add a metafield."
Findings come back keyed to attribute set, attribute code, CMS block identifier, and category — a ticket-ready backlog your dev or merchandising team can execute without translating from another platform's vocabulary. JSON-LD additions ship as blocks your team drops into a CMS block or PHTML override. Edits land through the admin UI, an import file, or your deployment pipeline; direct API write-back is on the roadmap (Early access). Every other channel rule applies here too: diffs are reviewed and approved by a human, never auto-shipped.
The weekly re-crawl then tracks citation counts per SKU across ChatGPT, Perplexity, Google AI Overviews, Gemini, Claude, and Copilot, so you can see which attributes paid back the implementation cost.
One attribute-set fix, catalog-wide lift.
Illustrative: a home-goods store on Adobe Commerce 2.4.7 with a Hyvä theme, nine attribute sets, 3,000 SKUs. The first scan flagged that material and country-of-origin attributes existed on most of the catalog but weren't surfacing in the additional information block because of an attribute-set property mismatch. One storefront-property change unblocked Product JSON-LD fields across roughly 80% of the catalog; citation counts rose over the following scan cycles.
Reference for the attribute model: Adobe Commerce product attributes documentation.
Related pages.
PDP optimization
How a finding becomes a shippable diff, on any channel.
GuideSchema for AI search
The Product JSON-LD fields ChatGPT, Perplexity, and Google AI Overviews actually read.
ChannelFor headless
PWA Studio and other hydrated storefronts get the rendered-audit workflow.
Running Shopify and Magento side by side? Start with For Shopify, then see SKU-level tracking and the product overview.
Frequently asked questions
Does eCommerce Insights require an Adobe Commerce admin connection?
No. It audits any Adobe Commerce PDP from its public URL — no Composer package, no integration token, no module in your environment. You paste the storefront URL, a sample of PDPs and category pages gets crawled, and the findings come back keyed to attribute sets, attribute codes, and CMS block identifiers.
Does it work with Magento Open Source as well as Adobe Commerce?
Yes. The audit is rendering-engine agnostic: Adobe Commerce Cloud, on-prem, and Magento Open Source 2.4.x all expose the same public surface AI engines crawl. Hyvä builds tend to score faster because they ship cleaner server-rendered HTML; PWA Studio storefronts run through the rendered-audit workflow.
How are recommendations written for the Magento attribute model?
In your vocabulary, not Shopify's. Where a Shopify diff says add a metafield, the Adobe Commerce version says add the material attribute to the Apparel attribute set with Visible on Catalog Pages set to Yes, then surface it in the additional information block. Instructions your team can execute in the admin in minutes.
Can eCommerce Insights write changes to our store directly?
Not today. An Adobe Commerce connector using the REST and GraphQL APIs is on the roadmap (Early access). The current workflow is diff-and-apply: recommended attribute values, CMS block content, and JSON-LD additions ship as diffs your team or implementation partner applies through the admin or a deployment pipeline.
How does it treat configurable products and child SKUs?
Each simple child SKU is scored on its own canonical URL, because ChatGPT Shopping and Perplexity cite child products as distinct items. The audit also flags the opposite failure: a configurable parent that has absorbed the content and citations its children should be earning.
Ask AI about eCommerce Insights for Adobe Commerce
Have your preferred AI engine summarize this page for your store.
If your catalog lives in attribute sets
Your AI visibility tool should know what one is.
No Composer install. No admin credentials. Drop a URL, get a Magento-aware scorecard.
attribute sets · CMS blocks · configurables