Dataset: the first 11 stores audited · 19–20 October 2026 · Every store scored from its public pages against the same 100-point rubric. Counts below are descriptive — this page is updated as the dataset grows, and it reports only what the underlying audits document.
Four stores scored "Mostly ready" and seven "At risk". These are established, well-regarded independent stores — not abandoned storefronts. The gap between how good these businesses are and how readable their catalogs are is the finding.
One more, narrower pattern: listing each colour as a separate product instead of a variant was confirmed in 4 of the 5 stores where the audit could determine the architecture from public data — in one store, a single knife model lives at five different addresses, its data split five ways.
Average score per area across the 11 stores. Two areas (identifiers, variant naming) can't be scored from public pages at all — in 11 of 11 audits, whether a store's GTINs are populated was simply not verifiable from outside; it takes the export check we run at kickoff.
| Area | Average | Of |
|---|---|---|
| Product titles & descriptions | 11.2 | 15 |
| Structured attributes & specifications | 8.5 | 15 |
| Image alt text | 6.0 | 10 |
| Categorisation & taxonomy | 5.1 | 8 |
| Shipping, returns & warranty information | 7.7 | 12 |
| FAQ & Knowledge Base coverage | 5.5 | 8 |
| Feed readiness signals (publicly verifiable part) | 2.6 | 5 |
None of these failures needs a redesign, a new platform, or a bigger team. They are source-data repairs: facts stated once, consistently, where machines look for them — which is exactly what the Agentic Readiness Fix does, change list first, your written approval before anything is applied. The free audit tells you where your store stands against the same rubric.
Read the method in full: the published worked examples, including a complete audit with its change list.