How Does White-Label AI Visibility Reporting Work for Agencies?

Short answer: A tool asks the same set of buyer questions in ChatGPT, Gemini, Perplexity and Google AI Overviews every day. Once a month you send the client the results under your agency's name and logo: which questions name them, who AI names instead, what your team fixed, and how it all moved since last month.

It is the reporting half of AEO (Answer Engine Optimization). AEO is the work of getting a brand named, and described correctly, in AI answers. If you are still deciding whether to offer AEO at all, start with how SEO agencies offer AEO services. This guide is about the report itself.

Your client will ask why ChatGPT recommends a competitor. The monthly report is where you answer before they ask.

What goes in the report

Keep the same sections in the same order every month. The client learns where to look, and month-to-month changes stand out.

• Your cover. Your agency's name, logo and colours. The client sees your work, not a tool.
• The headline number. The client's share of brand mentions across the fixed question set. Show this month, last month and the first month.
• The question table. Every buyer question, on every engine, with a score from "Recommended + Cited" down to "Not Visible". SiteSignal uses eight levels.
• Competitors. Which brands AI names instead, and their share of voice on each engine.
• Sources. The websites and pages AI cites when it answers. This tells you where to earn mentions next.
• Brand facts. Each answer marked accurate, partial, wrong or not enough data. Most months this is one line. When it isn't, it goes near the top.
• AI traffic. With the client's GA4 connected, visits from AI assistants, which assistant sent them and where they landed.
• Work done and next. The tasks your team finished from the weekly action plan, and the next three.

For what each metric means, see the complete guide to AI visibility tracking.

How to run it every month

1. Lock the question set in month one. Write the questions a buyer asks before they know any brand names. Keep a stable core so you compare like with like. If you swap questions, say so in the report.
2. Let the checks run daily. One run is an anecdote. Thirty runs show a pattern.
3. Watch for paused questions. In SiteSignal, a question that gets the same answer on every engine for 14 days pauses itself. Resume it or replace it, and note the change.
4. Generate the report. SiteSignal builds it on the 3rd of every month. You can also pick any date range.
5. Read before you send. Open every answer marked "Mentioned". If the brand name is also an everyday word, AI may mean the word, not the business. We hit this in a real report run from Claude.
6. Write three lines on top. What moved. Why you think it moved. What you will do next.

How to explain a trend to a client

AI answers change from one run to the next. So report the trend across the month, never a single screenshot.

The hard month is the one that goes down. The Uniform Consultants, a SiteSignal client (read the case study), had 7.3% of brand mentions in May 2026. Then it rebuilt its website. In June its average position slipped from 2.4 to 2.9. In July, Gemini stopped naming it. By September it had 13.6% of mentions, and all four engines named it first.

The July report is the one that keeps a client. Name the dip, name the likely cause, and show the plan. For a client in that spot, a line like this works: "Gemini stopped naming you after the site launch. We have resubmitted the new pages and added the missing service details. We will report the effect next month."

Two more rules help. Keep AI answer data apart from visits, leads and sales, which come from analytics. And don't count every mention as a win. An answer can cite your client's guide while it recommends a competitor.

How to price it

Price by scope: how many questions, how many engines, how many competitors and how many countries. That ties your fee to real work and real tool costs.

Know your cost first. SiteSignal's Agency plan is $399 a month for 3 client domains, plus $89 for each extra domain. Each domain gets its own 60 buyer questions. You use your own AI provider keys, at about $1 per question, per engine, per month. Checking a client in another country costs $49 a month.

A worked example from our agency page: 5 clients, 20 questions each, on ChatGPT, Gemini and Perplexity. That is $577 for SiteSignal plus about $300 in AI usage, or about $175 per client. Add your team's time on top. The AEO pricing models section covers retainers, tiers and sprints.

What not to promise

• Not "number one in ChatGPT". No one controls AI answers. Promise a process and a measured trend on a fixed question set.
• Not results in a fortnight. Engines that search the web live can pick up on-site fixes within weeks. Off-site authority takes months. Agree a 90-day review on day one.
• Not exact AI traffic. Some AI visits show up as direct traffic. Treat the GA4 numbers as a floor.
• Not "just one more question" for free. Write the engines and the question count into the contract, so extra checks become extra scope.

A test you can run this week

Pick one client. Write five buyer questions without their brand name in them. Ask each one in ChatGPT, Gemini, Perplexity and Google, on three different days. For each answer, note four things: are they named, in what position, who else is named, and which page is cited.

You will see answers change between days. That is why the report needs daily checks. If your client is missing from most answers, that is the first page of their first report.

Want to show the same view to a prospect first? Build an AI search proposal with free tools.

Try it on your own clients

White-label monthly reports come with the SiteSignal Agency plan. Start with a 7-day free trial. No card needed.

Start your 7-day free trial