Score, access findings, identity gaps, evidence gaps, and next steps — a report template that does not promise rankings.
Client reports fail in two opposite ways. One version promises ChatGPT rankings the agency cannot control. The other dumps raw crawl logs nobody reads. A good AEO report sits in the middle: score, access findings, identity gaps, evidence gaps, and next steps — without pretending a website checklist is a citation forecast.
Use this template monthly. It matches the ten-client agency workflow and BrandKnown's product reality: ~60-second scans, free first scan for prospects, Agency plan at $49/mo, readiness score as website checks only.
Report principles
- Lead with what changed on the website.
- Separate proven fixes from hypotheses.
- Show the re-scan diff.
- Keep AI naming / referral metrics in a clearly labeled "directional" section.
- End with a short backlog, not a novel.
If a sentence could not survive a skeptical CFO, cut it.
Suggested outline (copy this)
- Executive summary (five lines)
- Readiness score then vs now
- Access findings
- Identity findings
- Evidence findings
- Directional visibility notes (optional)
- Shipped work this period
- Next steps and owners
- Appendix: method and ceilings
Target length: two to four pages. Attach JSON-LD snippets and robots diffs in an appendix if needed.
Section-by-section
1. Executive summary
Answer four questions:
- What did we fix?
- What is still blocked?
- Did the re-scan improve?
- What happens next month?
Example: "We allowed PerplexityBot and ChatGPT-User through Cloudflare, aligned Organization name with the homepage H1, and published a definition-led product page. Readiness score moved from 61 to 79 on re-scan. We do not claim more Perplexity citations yet; GA4 shows 14 AI referral sessions (directional)."
2. Readiness score then vs now
| Metric | Start | Re-scan | Notes |
|---|---|---|---|
| Readiness score | 61 | 79 | Website checks only |
| Critical access issues | 3 | 0 | WAF allowlist shipped |
| Identity mismatches | 4 | 1 | Legal name still differs on footer |
| Open evidence tasks | 6 | 4 | Docs freshness remains |
State explicitly: the score is not citation likelihood and not an AI ranking.
3. Access findings
Table the crawler/WAF/robots issues.
| Issue | Status | Proof |
|---|---|---|
| PerplexityBot 403 via Bot Fight Mode | Fixed | 200 with bot UA after allowlist |
Pricing page Disallow | Fixed | robots.txt PR #842 |
| JS-only About body | In progress | SSR ticket DEV-219 |
Link to deeper primers only when useful — for example Cloudflare Bot Fight Mode vs AI crawlers.
4. Identity findings
Cover name consistency across title, H1, schema, and key profiles.
| Surface | Expected name | Found | Action |
|---|---|---|---|
| Title tag | Northstar Analytics | Northstar — Insights Platform | Updated |
| H1 | Northstar Analytics | Welcome | Updated |
Organization name | Northstar Analytics | North Star Analytics LLC | Pending legal OK |
| Crunchbase | Northstar Analytics | Match | Linked in sameAs |
Identity work is dull and high impact. Keep it visible.
5. Evidence findings
List pages that should be citable and what is missing: definition lead, steps, tables, freshness, schema where appropriate. Do not invent twenty FAQs. If you recommend FAQ schema, only for questions customers actually ask (FAQ schema guidance).
6. Directional visibility notes (optional)
Put prompt checks, Overview citation spot checks, AI referrals, and branded query trends here — labeled directional.
| Signal | This month | Interpretation |
|---|---|---|
| Prompt naming (20 prompts) | 5 / 20 | Up from 3 / 20; still sample-sized |
| AI referrals (GA4) | 14 sessions | Too small for CRO claims |
| Branded clicks (GSC) | +8% WoW | Could be seasonality; watch 6 weeks |
Cite external benchmarks only when they help context — for example Similarweb / TechCrunch (June 2025) on AI referral growth versus Google's still-larger referral volume — never as the client's implied results.
7. Shipped work
Bullets with dates and PRs. Clients renew for shipped work.
8. Next steps
Three to five items max. Each needs an owner and a verification method (re-scan, curl, Search Console).
9. Appendix: method and ceilings
Paste a short ceiling block every month so new stakeholders see it:
- BrandKnown readiness ≠ AI rankings or citation probability.
- Allowing a bot ≠ guaranteed fetch on every prompt.
- Re-scan proves website checks changed.
- Zero-click behavior (SparkToro / Similarweb ~68% of Google searches zero-click in early 2026; Pew on AI Overview click patterns) means referrals undercount exposure.
What to leave out
| Leave out | Why |
|---|---|
| "Guaranteed citations" | You cannot sell model behavior |
| Fifty unchecked Lighthouse nits | Dilutes AEO |
| Competitor fear screenshots without method | Panic without action |
| Unsourced "studies show 73%" stats | Fabrication risk; stick to STYLE-approved figures or hypothetic |
| Score theater without open issues list | Number without diagnosis |
Delivery checklist
- Re-scan production the same day you finalize numbers.
- Paste only live URLs.
- Separate Fixed / In progress / Won't do.
- Attach copy-ready snippets for anything still open.
- Schedule the next scan window on the calendar.
One-page mini template (for Slack-friendly updates)
AEO update — {client} — {date}
Score: {old} → {new} (website checks only)
Fixed: {1–3 bullets}
Open: {1–3 bullets}
Directional: {AI referrals / prompts / branded} — sample-sized
Next: {owner} ships {thing} by {date}; we re-scan after prod
Closing
A client AEO report should make the website more answer-ready and make the agency harder to fire. Score, access, identity, evidence, next steps. Leave the ranking prophecy off the PDF.