# AI visibility audits for agencies: a diagnosis you can ship to a client

> Your clients ask why ChatGPT names a competitor and not them. Answer with a measurement: the buyer questions read from the client's own site, where each model names them, who fills the slots instead, and a 14-day plan. Free first run per client domain, no login; full report USD 9 one time; lifetime pass USD 29.99 for every report after that.

Canonical page: https://askedthrice.com/for/agencies

## An AI visibility audit per client, measured, with the raw answers as evidence

## A deliverable, not a dashboard login

The agency version of "are we visible in AI search" is an audit per client that you can put in front of them: what OpenAI and Claude answer when their buyers ask, counted, with the raw answers attached so nobody has to take your word for it.

Monitoring platforms sell this as a recurring dashboard with an agency tier, and for some retainers that is right. For the pitch, the onboarding, the quarterly review and the client who asked once, a one-time diagnosis with the evidence attached is the deliverable, and it costs a fraction of a month of a platform.

## What exists now, stated exactly

There is no separate agency tier in the offer. Here is what you can do with the product as it ships.

1. **Run the free audit per client domain.** Enter the client's site, category and known competitors; the first run is free, takes about a minute and needs no signup. Up to 12 buyer questions are derived from the client's own pages and asked to OpenAI and Claude.
2. **The measurement language follows the client's site.** A client whose site is in Spanish gets Spanish questions and Spanish answers, with the competitors that market names; the language is read from the site, not from the page you started on.
3. **Share without a login.** The measured map has a public link, and the full report has its own share link and print to PDF. Send the link to the client; nobody has to create an account to read it.
4. **Re-runs are self-serve.** After the client publishes, run the same audit again and keep the history. No schedule, no ticket: measure when something changed.
5. **One lifetime pass covers every report.** The full report is USD 9 one time per audit. The lifetime pass is USD 29.99 one time and unlocks every report and every re-measurement on your account, across every client domain you run. Cheaper than single reports from the fourth one.

## What to put in front of the client

The pieces the audit gives you, in the order a client reads them.

1. **The map.** Question by question, model by model: named, not named, or nobody named. Counts with their N, no score. It is the one picture that answers "are we visible" without adjectives.
2. **The competitor counts.** Who fills the slots the client is missing, counted per model. This is where the conversation about positioning starts, with names the client recognizes.
3. **The missing pages.** Which buyer questions have no page on the client's site that could answer them. This turns the audit into scope: it is the page backlog.
4. **The 14-day plan.** From the full report: which pages to ship first, with blueprints, so the next two weeks of work are written down before the meeting ends. Re-run after shipping to show movement as counts, not as a story.

## One-time diagnosis vs the agency tiers of monitoring platforms

Read from each vendor's pricing page on 2026-08-23. Peec AI's pricing page, same day: Starter USD 95/mo (50 prompts, 3 AI models, 1 project), Pro USD 245/mo (150 prompts), Advanced USD 495/mo (350 prompts, multi-country), Enterprise custom; annual billing minus 15 percent; extra AI model add-on USD 35, 85 or 165 per month by tier; agency plans from USD 245/mo.

| Tool | Entry price | Engines | Free tier or trial | Best for |
| --- | --- | --- | --- | --- |
| Asked Thrice (the one-time audit option) | Free first run, no signup; USD 9 one-time report; USD 29.99 lifetime pass for every report and re-run after that | OpenAI and Claude model APIs (budget-tier models, named on the methodology page), no web browsing; the paid report also measures Google Gemini | Free first run, no signup, no card | A diagnosis once, before or instead of a monitoring subscription |
| Otterly.AI | Lite USD 29/mo (15 prompts); Standard USD 189/mo (100 prompts); Premium USD 489/mo (400 prompts); Enterprise from USD 1,000/mo | ChatGPT, Google AI Overviews, Perplexity, Copilot; add-ons per month by tier: Google AI Mode USD 9/59/149, Gemini USD 9/59/149, Claude USD 29/109/439 | 7-day trial with 50 prompts, no card; no permanent free plan; 12 free GEO tools on its site | The lowest-priced entry into recurring tracking, paying per engine |
| Peec AI | Starter USD 95/mo (50 prompts, 3 AI models, 1 project); Pro USD 245/mo; Advanced USD 495/mo; Enterprise custom; annual minus 15 percent; extra AI model USD 35 to 165/mo by tier | 3 AI models on Starter, more as paid add-ons; Claude and GPT-5 Search reported as enterprise-only | 7-day trial, no card; no free tier | Teams that want ongoing dashboards with unlimited seats |
| Scrunch | Starter USD 250/mo annual or USD 300 month-to-month (350 prompts, 3 users); Growth USD 417/mo annual or USD 500 monthly (700 prompts); Enterprise custom | ChatGPT, Claude, Gemini, Perplexity, Google AI Mode, Google AI Overviews, Meta | 7-day free trial of Starter, no card | Brands and agencies that want 7 engines from the entry plan |

The comparison is not "cheaper is better". A platform's agency plan buys weekly tracking, seats and a reporting cadence; the audit buys a measured diagnosis per client with the evidence attached. Many agencies do both: the audit to win and onboard the client, the platform once the retainer includes weekly action. Checked 2026-08-23.

Sources: [Otterly.AI](https://otterly.ai/pricing), [Peec AI](https://peec.ai/pricing), [Scrunch](https://scrunch.com/pricing)

## Buy the agency plan when these are true

If your retainer already includes someone acting on AI visibility every week, the subscription is the right tool and the audit is the baseline you run before it starts.

- You report to the client weekly or monthly on the same prompt set and need the trend line, not a point.
- You run 100 or more prompts across several clients and need them in one place with seats for the team.
- Clients sell in several countries and you need per-market tracking.
- The client pays for monitoring as a line item and expects a dashboard they can open.

## One run is a sample, so the audit shows counts with their N

In our repeatability study (4,500 calls, 30 categories, three model APIs) the same buyer question asked 5 times to the same model returned the same set of brands in 15 percent of cases. The first-named brand held in 75 percent of runs; the rest of the list moved. So a single run is a sample, not the answer: read every count with its N, re-run before you call a change real, and treat any tool that reports a 0-100 score from one run as printing noise.

That is also why there is no score on any screen: the verdict is three bands plus counts ("named in X of N answers"), and the methodology page lists the exact models, temperatures and run counts behind every number.

- [How repeatable are AI recommendations (the study, raw answers downloadable)](https://askedthrice.com/research/how-repeatable-are-ai-recommendations)
- [Methodology: models, parameters, what we do not claim](https://askedthrice.com/methodology)
- [How many runs does a measurement need](https://askedthrice.com/learn/how-many-runs)

## Measurement in Spanish, and public boards by rubro

Clients in Argentina and the rest of Latin America are measured in Spanish when their site is in Spanish, with the competitors the models name for that market. The public index by rubro shows who the models name in each category measured so far, and the Spanish landing explains the audit in the client's language.

- [Índice por rubro: a quién nombra la IA](https://askedthrice.com/rubros)
- [Auditoría de visibilidad en IA (en español)](https://askedthrice.com/es)

## Questions agencies ask before running it

### Is there an agency tier or white-label?

Not in the offer today. What exists: free first run per client domain, the full report at USD 9 one time, and the lifetime pass at USD 29.99 that unlocks every report and re-run on your account. If you run many client brands and need something else, write to jorgenahuelsoria@gmail.com; we would rather say what exists than promise a tier.

### Can the client see the report without an account?

Yes. The measured map has a public link, and the full report has a share link and print to PDF. The link is the deliverable; no login on their side.

### How does this compare with the agency plans of Peec AI, Otterly or Scrunch?

Different job. As of 2026-08-23: Peec AI lists agency plans from USD 245/mo, Otterly starts at USD 29/mo for 15 prompts and Scrunch Starter is USD 250/mo annual or USD 300 month-to-month. Those buy ongoing dashboards; the audit buys a measured diagnosis per client, once, with the evidence attached, for USD 9 per report or USD 29.99 for all of them.

### How reliable is one audit?

It is a sample, and the report says so. In our 4,500-call study the same question to the same model returned the same set of brands in 15 percent of cases, so every count is shown with its N and re-runs are self-serve. Show the client movement as counts across two measurements, not as one number.

### Does it work for clients whose site is not in English?

Yes for Spanish: the questions, answers and competitors follow the language of the client's site. The public index by rubro and the Spanish landing are linked above.

## Next step

Audit your first client for free: https://askedthrice.com/

Free first run per client domain, no signup. Full report USD 9 one time; lifetime pass USD 29.99 covers every report.

---

Machine-readable index: https://askedthrice.com/llms.txt · Full context: https://askedthrice.com/llms-full.txt · All URLs: https://askedthrice.com/sitemap.xml · Developer and agent resources: https://askedthrice.com/developers.md