Alderfield Home — AI-Commerce Readiness Audit
SyntropyData {{ docDomain }}  ·  {{ docDate }}  ·  {{ docRef }}

Demonstration document — the fictional Alderfield Home store. No live store was inspected.

Stage 1 of 4 — the full diagnosis · this is the complete picture of your catalogue. This document diagnoses; it proposes no corrections and asks you to decide nothing. If you decide to have it put right, stage 2 is your Merchant Review Pack inside a Catalogue Fix Sprint — and stage 3, the Re-check, reports what actually went live.

AI-Commerce Readiness Audit

On {{ nWithFinding }} of your {{ nProducts }} products, at least one fact buyers ask for is missing from your catalogue data. {{ ownAnswerVerdict }}

This Audit evaluates the information your products need in order to be understood accurately — by buyers, and by the AI shopping systems that read your catalogue for them. We read every product in scope: your catalogue data and your product pages — the evidence your store publishes today, and nothing else. The result is {{ nFindings }} catalogue issues, grouped into {{ nPatterns }} recurring issues you can inspect below, product by product.

{{ nWithFinding }} products carry at least one issue — {{ ownAnswerHeroNote }} Of the {{ nProducts }} products in scope, {{ nWithFinding }} carry at least one issue, {{ nWithoutFinding }} need nothing at all, and {{ nNotFullyInspected }} we could not fully inspect — so we conclude nothing about those. What matters more than the count is the shape below: what we established, what your published product data actually carries, and which few things genuinely need you. {{ nProducts }} products in scope¹  ·  {{ nFindings }} issues — {{ nProductFindings }} on products, {{ nStoreFindings }} on the store  ·  what we read, and what we could not ↓

What this means for improvement

{{ improvementLead }} Behind {{ nOwnEstablished }} of the {{ nFindings }} issues, your own store already states the answer — the issue is that your catalogue data does not carry it. A further {{ nResearchEligible }} facts your store cannot establish qualify for bounded research into official sources — resolution work the Fix Sprint performs after you buy it, validated against the exact product before anything is proposed. Nothing has been changed and nothing is proposed yet: this document diagnoses, and unresolved evidence stays held with us rather than becoming homework.

A small set of decisions only your business can make — conflicts between your own statements, and commitments only you can give — reaches you once, in your Review Pack, inside the Fix Sprint. Everything the evidence settles is resolved without you.

The Fix Sprint turns these catalogue issues into exact catalogue changes — every one shown to you in full before anything is implemented. What fixing this involves ↓

¹ {{ nExclusions }} non-merchandise items (48 swatches & samples, 11 gift cards, 9 shipping-protection & fee SKUs, 4 placeholder SKUs) were excluded from product counts.

Coverage by catalogue fact

Buyers need certain facts before they can choose with confidence. For each fact this Audit evaluates, here is where the answer stands across your {{ nProducts }} products — held in your catalogue data, found elsewhere on your store, conflicting, missing, or not assessable. Each state resolves differently, so they are never collapsed into one number.

The requests beside each row are illustrative. The Audit evaluates the underlying catalogue facts, not a fixed list of natural-language questions.

{{ cv.label }}“{{ cv.request }}”{{ cv.surface }}
{{ b.label }}

{{ cv.notes }}
Colour and colourways are two separate facts
A product offered in six colourways can already support a request for a specific colour — that is catalogue structure working, not a missing colour. Where a product is sold in exactly one colourway, we read that as the product’s colour and record it — it is never counted as a gap and never becomes a decision for you.

What your published product data carries

{{ repHeadNote }}

Many product pages also publish a structured product record — the machine-readable data an AI shopping system reads first. We measured what those records actually carry against the facts we established.

One product, as an AI shopping system meets it

{{ exAlt }} {{ exTitle }} ↗ {{ exMeta }}
{{ exFactLabel }} found on your store {{ exTruth }}{{ exTruthFrom }}
Published product record {{ exRead }}

An AI shopping system relying on that product record would not receive this product's {{ exFactLower }} from it. The value is already established, so closing the gap is correction work a Fix Sprint carries — no decision from you.

Fact by fact, on the products where the record was measured

{{ r.label }} {{ r.denomLabel }}
{{ lg.t }}

Catalogue-level summary — per-fact detail is turned off

{{ repGapLine }}

{{ repTopGaps }}

Measured absence — the whole record

{{ repWholeLine }}

{{ wholeTitle }} {{ wholeSku }}{{ wholeRead }}{{ wholeNote }}

Not measured — and therefore not a gap

{{ repUnmeasuredLine }} They sit outside every count in this section, and no correction is proposed for them.

One record disagreeing with the established fact

{{ cfTitle }} {{ cfSku }}established: {{ cfTruth }}published record: {{ cfRead }}{{ cfNote }}

Recurring catalogue issues

{{ topPatternsLine }}

The kinds of problems that affect multiple products in your catalogue, grouped by what is actually wrong. Each issue is explained once, with one representative example; every affected product is one click away in the product index.

{{ fam.title }}

{{ fam.countLine }}

{{ fam.blurb }}

{{ pt.count }} {{ pt.problem }} {{ pt.reachLine }}

Two of your own surfaces disagree

{{ pt.reading }}

Both statements are yours, and the evidence we hold does not establish which of them overrides the other. So this Audit does not choose between them and does not write a value here. The exact wording of each, and every product it touches, is below.

{{ pt.reading }}

Why it matters{{ pt.matters }}

{{ ex.title }} {{ ex.sku }} established: {{ ex.established }} established components: {{ ex.partial }} {{ ex.conflict }} {{ ex.note }}
More examples
{{ ex.title }} {{ ex.sku }} established: {{ ex.established }} established components: {{ ex.partial }} {{ ex.conflict }} {{ ex.note }}

Where the evidence did not establish a complete answer

{{ tailPatternsLine }}

These {{ nNothingEstablishes }} facts remain incomplete: some establish no value, others establish components but not the complete fact. Nothing is guessed, so nothing is proposed for them yet. Across the catalogue, {{ nResearchEligible }} facts your store cannot establish qualify for bounded research into official sources — the maker's or brand's own published material, an official supplier, an authoritative registry — and that research is work the Fix Sprint performs, validating any result as this exact product before proposing a change.

{{ ue.count }}{{ ue.label }}

{{ ue.note }} E.g. {{ ue.example }}.

Evidence and search detail

Established components: {{ ue.components }}

{{ quote.where }}
“{{ quote.text }}”

{{ ue.searched }}

{{ nStoreTruths }} store-level records

store-wide, not repeated per product
{{ sr.title }}scope: {{ sr.completeness }}

Your own published terms state “{{ sr.quote }}” — {{ sr.why }}

Products covered and excluded

Covered by this record ({{ sr.scopeCount }}): {{ sr.scope }}

Excluded by its own wording ({{ sr.excludedCount }} · {{ sr.exclusionTerms }}): {{ sr.excluded }}

What already works

{{ nWithoutFinding }}products answer every fact we evaluate {{ nProducts }}prices & availability in your catalogue data — every product {{ nHealthyColourways }}healthy colourway structures — {{ nColourwayVariants }} colourway variants map cleanly {{ nCareInCatalogue }}products already carry care in your catalogue data

Rugs and lighting are your strongest areas — most answer everything already. These {{ nWithoutFinding }} products require no action, and none of them generated an issue.

Product index — every product, every issue

{{ apxTotal }} products

The complete catalogue, product by product. Every row carries authoritative identity, option axes and a live-page link. Open it to see exact evidence and consequence where issues exist, the useful completed-inspection outcome where none exists, or the explicit inspection limit where no conclusion was possible.

View

{{ viewNote }}

Filtered: {{ filterSummary }}

Showing {{ apxShowing }} of {{ apxCount }} matching · {{ apxTotal }} in the catalogue. No product is excluded from this index.

ProductSKUOptionsIssue statusLinks
{{ row.artAlt }} {{ row.title }}{{ row.type }} · {{ row.price }} SKU{{ row.sku }} Options{{ row.optionSummary }} Issues {{ row.statusCount }} {{ ch.t }}
{{ row.artAlt }} {{ row.title }} ↗ {{ row.type }} · {{ row.sku }} · {{ row.price }} · {{ row.optionSummary }}

{{ row.stateLabel }} {{ row.summaryText }}

{{ e.request }} {{ e.famLabel }} {{ e.provenance }}
{{ s.n }} {{ s.v }} {{ s.note }}

{{ e.reading }}

{{ e.askedLabel }}{{ e.asked }}
{{ pageLabel }} Search and filters always run over the complete catalogue — paging only limits what is drawn at once.

Where we inspected

We inspected all {{ nProducts }} catalogue records, usable product pages and published product data — your store's current evidence, and nothing else. {{ nPageExceptions }} pages could not support page-dependent conclusions — they are flagged in the index and carry no issues.

Exactly what was read, and the exceptions

Your catalogue data — read for all {{ nProducts }} products in scope. It is where a fact can be stated once and read the same way every time, and where a correction lands.

Product pages — {{ nPagesAcquired }} of {{ nProducts }} usably fetched. A fact that lives only here is real — it is simply not stated in your structured data.

The published product record, where a page publishes one — {{ repMethodNote }}. Read to completion or reported as not measured, never inferred from page prose.

Nothing outside your store. This Audit diagnoses from your current evidence only. Where a missing fact qualifies for bounded research into official sources, that research is Fix Sprint work — performed after purchase and validated as this exact product before anything is proposed.

Exceptions. {{ nPageNoResponse }} pages did not return a usable response, {{ nPageUnavailable }} were unavailable, and {{ nPageNotBound }} could not be established as the product's own page. No absence was concluded from any of them. {{ nExclusions }} non-merchandise items (swatches, samples, gift cards, fees, placeholders) were set aside from product counts.

What fixing this involves

This audit is the diagnosis, and it stands on its own — everything above is yours to act on however you like. If you'd like us to put it right with you, that work is a Catalogue Fix Sprint.

Your part stays small The Fix Sprint resolves what evidence can settle without you: it establishes each fact from your own store, researches the eligible missing facts against qualifying official sources, and prepares every exact catalogue change for your review. What does need you: the few decisions only your business can make — asked once, in your Review Pack.
Then the changes get made — two complete routes, both included Once you've decided, you choose who makes the changes, once, inside the Review Pack — managed by us under permission you grant separately, or by your own team from a complete package. Both are part of the Sprint price.
How each route works Once you've decided, you choose who makes the changes, once, inside the Review Pack. Managed implementation — recommended: we make the changes we can make safely, under permission you give separately and can refuse, and read each one back afterwards. Manual implementation: your own team makes them from a package that states every change in full, and we never ask for access. Both are part of the Sprint price, neither re-opens a decision, and either way the Sprint isn't finished until every approved change has an outcome.
A re-check confirms what actually landed We inspect your store again afterwards and report what is now answerable, what only partly landed, and anything that didn't — in the same words and under the same products as this report.
Two of these issues, as the Sprint resolves them preview only — nothing here is live