The deliverable, shown
A real audit report, anonymized.
This is a real audit from our files — run on a brand in our founder's own portfolio, before any fixes were made. Every number below is real; the company and competitor names are redacted. What you see here is exactly the structure you get for your own brand.
Honesty notes: this report was measured on Google AI Overviews (the engine we measured live at the time). Audits today run the same method across ChatGPT, Perplexity, Gemini and Google AI Overviews. "Company A" is a German commerce brand from our founder's own portfolio — not a paying client — which is why we can publish it. AI answers are non-deterministic; a report is a dated snapshot, and we never promise rankings.
1 · The scorecard
Composite GEO score = Citation 30% · Share of Voice 25% · Consistency 15% · Source Diversity 15% · Authority 10% · Query Coverage 5%.
2 · The evidence matrix — every number traced to a query
Each row is a real buyer query. The full report contains all rows plus an appendix with the raw answers; here is an excerpt:
| Buyer query | Type | AI answer? | Brand named? | Cited instead |
|---|---|---|---|---|
| buy [product] with customization | transactional | — (shopping results) | — | — |
| which [product] for a [occasion] | informational | yes ✓ | no ✗ | competitor-1.de · competitor-2.de · competitor-3.de |
| how to choose the right [product] | informational | yes ✓ | no ✗ | competitor-4.de · competitor-5 · competitor-6 |
| which [product] types exist | informational | yes ✓ | no ✗ | competitor-1.de · competitor-2.de · competitor-6 |
| [product] meaning and tradition | informational | yes ✓ | no ✗ | competitor-7.de · competitor-3.de · wikipedia.org |
Result across the full set: 5 of 12 buyer queries triggered an AI answer — the brand was named in 0 of 5, and appeared on 0 of the 15 cited sources. Transactional "buy…" queries returned shopping results instead of AI answers (a separate channel, handled in the roadmap).
3 · Share of voice — who gets the visibility instead
The report then explains why the winners win — guide content on exactly the queries where AI answers appear, presence in the cited source ecosystem, established review footprint — each with the evidence behind it.
4 · The prioritized roadmap (excerpt)
- Guide content for the AI-answer queries. Become citable exactly where AI answers appear — structured, extractable pages for the informational queries the engines actually answer. Lifts: citation, share of voice · effort M.
- Enter the cited source ecosystem. Build presence on the platforms the engines cite (industry portals, review platforms, an encyclopedic brand entity) — where the competitors sit and the brand is absent. Directional impact per Princeton GEO research: ~30–40% for citation-type changes · effort L.
- Make product pages machine-extractable. Structured data (Product/Offer/FAQ), comparison tables, llms.txt — so the brand is cleanly citable the moment an answer appears. Effort S–M.
- Win the transactional queries via their own channel. Where engines show shopping instead of AI answers, the lever is product feeds and reviews — a separate, parallel channel. Effort M.
Every roadmap item is written to be handed to your team — or we implement it for you.
And the free snapshot?
The snapshot is page one of this, for one buyer query: which engines mention you, who they name instead, and the 2–3 fixes that move it. For this brand it read: "Named in 0 engines · cited instead: Competitor 1, 2, 3 · fixes: guide content, source ecosystem, structured data." That's what lands in your inbox within 48 hours — free, no call.
Want this for your brand?
Start free with the 1-page snapshot, or book the full audit — the fee is credited to month 1 if you continue to monitoring.