The fields, and why each one is there
An AI visibility report for a client has three blocks. Fill them in order.
Metrics block
- Prompt set and version
- Engines, and the date of the run
- Number of runs
- Answers checked
- Answers naming the client, and the share
- Answers citing the client's pages, and the share
- The denominator, written out
A share without its denominator is not a metric. See AI share of voice: define the denominator.
Findings block, one row per prompt
- Prompt and buyer intent
- Engine and date
- Mention status
- Citation status
- Cited domains: who got the citation instead
- Accuracy issues in what the engine said about the client
- Competitor named
Next actions block
- Recommended action
- Owner: the agency, the client, or an Adviora Action Pack
- Follow-up date
- What to re-check
If a client asks for "client-ready" AI visibility reports, this is the shape that gets a yes: every number has its denominator and every action has an owner.
The filled example: Adviora as the client, 15 October 2026
The client here is Adviora itself, from a run on 15 October 2026. It is one run on API samples, so treat it as directional.
Metrics
| Field | Entry |
|---|---|
| Prompt set and version | Panel v1, 22 prompts |
| Engines and date of run | Google AI Overviews (stored answers); ChatGPT, Perplexity, Gemini and Claude (API samples); 15 October 2026 |
| Number of runs | One |
| Answers checked | 15 prompt-by-engine cells, from 7 unbranded category prompts |
| Answers naming the client | 0 of 15 (0%) |
| Answers citing the client's pages | About 128 citation links observed; 0 pointed to adviora.ai |
| Denominator | 15 answers (prompt-by-engine cells) |
Findings
| Prompt | Mention | Citation | Cited domains instead | Accuracy issue | Competitors named |
|---|---|---|---|---|---|
| P01: "best AI visibility tools" | Not named | No | zapier.com, frase.io, business.com and others | None recorded | Profound, Peec, Otterly, Semrush |
| P08: the Generative Engine Optimization (GEO) tools-for-agencies prompt | Not named | No | Not recorded here | ChatGPT read "GEO" as local SEO | Not recorded here |
| P20: "What is Adviora?" | Named in 4 of 4 answers | Claude cited demo.adviora.ai | demo.adviora.ai | Claude listed three different entities | None |
Next actions
| Recommended action | Owner | Follow-up date | What to re-check |
|---|---|---|---|
| Pitch the independent list-article editors | Agency (marketing) | 11 November 2026 | Whether the cited domains for P01 change on the next run |
| Write "Generative Engine Optimization (GEO)" in full on every title | Content lead | Week 1 | Whether P08 still reads GEO as local SEO |
| 301 the demo host and add Organization markup to /about | Developer | Weeks 1 and 2 | Whether Claude still lists three entities for P20 |
The blank template
Download the sheet and copy it into the spreadsheet tool you already use. It is free to copy and adapt, and there is no sign-up. Keep the agency's own prompts fixed between runs. If you pool prompts across clients, you can describe them that way, but split the numbers per client in each report.
Download the blank template (CSV) · Download the filled example (CSV)
The blank sheet has these columns: Block, Field or prompt, Engine and date, Entry or mention and citation status, Cited domains, Accuracy issue and competitor named, Action, owner, follow-up date and what to re-check.
How to present it to a client
- Lead with the three metrics and their denominator.
- Show two findings rows the client will recognize: a competitor named in their place, and a wrong fact an engine stated.
- End with three owned actions and their dates.
- Say what one run cannot prove. A single run is a snapshot, and it does not show how many visitors an answer sends. For the referral side, see track AI referrals in GA4.
A 20-minute agenda that fits: 5 minutes on the metrics, 5 on the findings, 5 on the actions, and 5 for questions. If you want to package this as a client service, see AI search reporting for agencies, and agency plan prices for the plans.
What the agency plan adds
- One login for every client, with each client in its own workspace and its own prompt panel.
- Agency covers up to 10 clients and Agency Pro up to 25, with prompts pooled across them.
- White-label reporting is not available yet.
Plan prices in USD and INR are on the pricing page. Before you commit, run the free AI visibility checker on one client's domain.