AEO audit
A scored diagnostic of whether answer engines can use your site
Not a crawl report with 400 warnings. Six modules, each scored, each tied to a specific reason your content is or is not being extracted — and a backlog ordered by how much of the catalogue each fix touches.
- Modules
- 6 scored areas
- Output
- Prioritised backlog
- Level
- Template, not URL
Overall readiness
61/100
- Retrieval accessBlocked in part
- Rendered contentGaps
- Entity clarityPartial
- Passage structureWeak
- Structured dataStrong
- External consistencyGood
Why audit first
Publishing more content cannot fix a retrieval problem
When a site is not being cited, the cause is usually structural: content that never renders, specifications trapped in images, passages that cannot stand alone, or facts that contradict each other. Writing more of the same adds volume to a broken pipeline.
- Diagnose the gate that content is failing at.
- Score at template level so fixes scale.
- Order work by blast radius, not by ease.
- Re-score after each phase to prove causation.
- Crawler directives and AI user agents
- Server-rendered content coverage
- Entity and attribute definitions
- Passage self-containment
- Structured data accuracy
- Cross-source fact consistency
- Question ownership across templates
- Duplicate and thin answer content
The six modules
What we score, and what each score means
Each module has a defined pass condition. Indicators below are illustrative interface values.
- 01
01 — Retrieval access
Whether the user agents behind answer surfaces are permitted, how fast pages respond under load, and whether any protection layer silently blocks fetches.
Score54/100 - 02
02 — Rendered content
How much of the substance — specifications, guidance, reviews, availability — exists in server HTML rather than appearing only after interaction.
Score48/100 - 03
03 — Entity clarity
Whether brand, products, categories and attributes are described consistently enough for a model to restate them without inference.
Score63/100 - 04
04 — Passage structure
Whether individual passages name their subject, lead with a claim, give a measurable reason and state their condition.
Score37/100 - 05
05 — Structured data
Whether markup validates, matches visible content, and covers Product, Offer, Breadcrumb, FAQ and Organisation where relevant.
Score86/100 - 06
06 — External consistency
Whether the facts third-party sources publish about your brand agree with your own — naming, positioning, availability and specifications.
Score78/100
How the audit runs
Five stages over two to three weeks
- Read-only access is enough to begin.
- Findings are reviewed live, not emailed as a PDF.
- Every finding names the template it affects.
- 01
Prompt and question set
Define the questions your category actually generates.
- 02
Access and render testing
Fetch as answer-surface agents; compare rendered against source HTML.
- 03
Template sampling
Score representative category, product, comparison and guidance templates.
- 04
Consistency review
Compare on-site facts against structured data and external sources.
- 05
Backlog and scoring
Prioritised fixes with expected coverage impact.
Typical findings
The failures that repeat across ecommerce sites
Illustrative pattern library. Most audits surface four to six of these, and they explain most missing citations.
| Finding | Gate it fails | Scope | Fix type |
|---|---|---|---|
| Specifications only inside images | Extract | All product URLs | Template |
| Reviews injected client-side | Parse | Product templates | Rendering |
| Category intros with no definition | Extract | All categories | Content |
| Inconsistent product naming | Compare | Catalogue-wide | Data |
| Schema contradicting visible price | Compare | Offer markup | Technical |
| Comparison pages without verdicts | Cite | Comparison set | Content |
Deliverables
What you receive
- Scored readiness report across six modules
- Prompt and question set for your category
- Template-level passage assessment
- Access and rendering test results
- Structured data validation and mismatch list
- External consistency comparison
- Prioritised backlog with expected coverage impact
- Re-score plan to verify each phase
After the audit
The backlog becomes a sequenced programme
Findings map directly onto delivery work. Most clients move from audit into a phased roadmap rather than a fixed retainer of undifferentiated tasks.
- AEOthe passage-level delivery work findings feed into.
- GEO Strategysequencing the fixes across quarters.
- Ecommerce SEO Auditthe classic-search counterpart audit.
- 01
Audit
Score and diagnose
- 02
Unblock
Access and rendering
- 03
Restructure
Templates and passages
- 04
Expand
Unowned questions
- 05
Re-score
Prove the change
Measurement
How audit progress is verified
Illustrative demo data. Re-scoring uses the same modules and prompt set as the initial audit so movement is attributable.
- Readiness score84Composite across six modules after phase two
- Blocking findings closed18Findings marked as retrieval-blocking
- Templates remediated9Category, product, comparison and guidance templates
- Questions newly owned+142Mapped questions with an assigned answer passage
Figures shown are illustrative interface data, not client results.
FAQ
AEO audit questions
A scored diagnostic of whether answer engines can retrieve, extract and trust your content. It covers retrieval access, rendered content coverage, entity clarity, passage structure, structured data accuracy and external fact consistency, and produces a backlog ordered by how much of the catalogue each fix affects.
A classic audit asks whether pages can rank. An AEO audit asks whether a passage can be lifted out of a page and attributed. That adds passage-level assessment, fetch testing as answer-surface user agents, and cross-source fact consistency checks that ranking audits do not include.
Two to three weeks for most ecommerce sites. Larger catalogues take longer only in the template sampling stage, because more distinct templates need scoring.
Read-only analytics and search console access, plus a crawlable staging or production environment. Access to the CMS or theme is helpful for assessing how content renders but is not required to complete the audit.
The audit itself is diagnostic. Most clients then move into delivery, where the backlog is executed in phases with re-scoring at the end of each phase to confirm the change was the cause.
Often yes, because strong technical SEO does not guarantee extractable passages or consistent external facts. Sites with excellent crawl health frequently score lowest on passage structure, which is where citations are actually won.
Book a diagnostic
Get your answer readiness scored
Six modules, template-level scoring and a prioritised backlog — so the next quarter of work is aimed at the gate your content is actually failing at.
No obligation. A senior strategist reviews your store before the call.