What Is Catalog Readiness? (It's Not the Same as Clean PIM Data)
Catalog readiness is whether each SKU carries enough structured signal for an AI agent to retrieve, filter, compare, and transact on it — without a human catching exceptions. PIM completeness measures something different.
Here's what catalog readiness failures actually look like in practice.
A buyer asks ChatGPT Shopping: "Heavy-duty degreaser safe for aluminum engine components." Your product doesn't appear. It's listed as "industrial cleaner" with material compatibility set to "metal surfaces." That's not specific enough for the filter. The agent excluded it before forming a shortlist.
A buyer asks Gemini: "6-inch round duct fittings for residential HVAC, in-stock." Your fitting isn't in the results. The sub-category field is blank. The agent couldn't place the product correctly.
A buyer asks Perplexity: "Stainless steel hex bolts, grade 5, pack of 50." Your product shows up in retrieval. It disappears at the filter step. The grade field says "high-strength" — not the same as grade 5, and the agent won't guess.
These aren't ranking failures. They're filter failures. The product existed in the catalog, was retrieved, and was dropped before any recommendation formed. No error notice. No suppression flag. Just silence.
Your PIM dashboard shows 94% field completeness. Your channel suppression rate is 22%. Both numbers are accurate. The gap between them is what catalog readiness measures.
Why "clean data" stopped being a useful north star
Three years ago, "clean" meant field fill rate above 90%. That was enough for manual review workflows and basic feed validation. Today it isn't.
Two things changed at the same time.
Channels raised their standards continuously. The fields that passed validation two years ago fail today. Taxonomy structures change quarterly. Compliance requirements shift by jurisdiction and product category. A product that was correctly classified at onboarding may now be misclassified — with no notification.
AI agents introduced a new filter layer. ChatGPT Shopping, Perplexity, Google AI Overview, and Gemini Shopping all work the same way: retrieve a candidate set, filter it against structured constraints, compare what's left, recommend the best match. A product with missing compatibility data fails the filter. A product with an incorrect hazmat flag fails the filter. There is no ranking penalty to diagnose — the product simply isn't in the consideration set.
Visibility = structure. Not copy quality. Not keyword density. Structure.
The definition that matters now
Catalog readiness is the degree to which each SKU can be acted on by an automated system — found, filtered, compared, and transacted — without routing to a human exception queue.
A product that passes retrieval but fails the compatibility filter isn't ranked lower. It disappears. A product with a valid global identifier but wrong taxonomy lands in the wrong browse category regardless of how well its attributes are filled. A product with placeholder values in required fields passes your completeness check and fails channel submission.
These aren't the same problem. They fail at different points in the pipeline and require different fixes.
The three readiness buckets
| Bucket | What it covers | What "ready" looks like |
|---|---|---|
| Discovery | Brand identity, global identifiers, taxonomy, core attributes, compatibility and fitment | AI agents can find, identify, filter, and compare the product against buyer constraints |
| Compliance | Hazmat flags, multipack signals, restricted product status, tax category | Product can be published and sold without triggering channel rejections or post-listing suspensions |
| Content | Description specificity, use cases stated, no placeholder values, consistent specs | AI agents can reason about the product and form a confident recommendation |
PIM completeness scores none of these correctly. It measures field presence — not value correctness, not channel alignment, not compliance signal accuracy.
A field that says "metal surfaces" instead of "aluminum, carbon fiber, steel" is complete. It is not ready.
Readiness is a continuous state, not a project outcome
Enrichment is a project. You run it once, fix the gaps, and the catalog looks right for a few months. Then new suppliers join. Products get updated. Taxonomies change. Compliance rules shift. The drift starts again.
Readiness is different. Maintaining it means running the checks every time something changes — not on a quarterly schedule. When a supplier sends a new file, the identity checks run. When a channel updates its taxonomy, affected products get re-evaluated. When a product description changes, compliance signals get re-assessed.
A catalog that was ready last quarter may not be ready today — not because anything was done wrong, but because the channels it publishes to kept moving.
How to assess where you stand today
The fastest way to understand your current readiness is to score each product across the three buckets. The gaps that matter most are usually in Discovery first — brand normalization failures, wrong taxonomy, missing identifiers. A product with perfect content still gets suppressed if its taxonomy is wrong or its compliance signals are missing.
Fix Discovery and Compliance gaps before investing in Content. The order matters. A buyer asking Gemini for "food-safe industrial degreasers" won't see your product if the hazmat flag is missing — regardless of how specific your description is.