Skip to content

For agencies

AI visibility reporting for agencies.

Show each client how often AI engines name them when buyers ask about their category, who gets named instead, and what to publish next, with a range on every number so the report holds up when a client pushes back.

How it works

A client report, step by step.

Engines running today: ChatGPT, Claude, Gemini and Google AI Mode. Each plan asks the running engines it includes, shown in the table below.

  1. Step 1

    One brand per client, with its own questions

    Add each client's brand and the competitors you want measured against it. On a paid plan AmpliRank drafts buyer questions from the client's site, and you review each one before it is asked; you can also write your own.

  2. Step 2

    A run per client, weekly or monthly

    Every question goes to each engine the plan includes, with web search on, the same way every time. The headline is the share of answers, to questions that don't name the client, that name the client, with its 95% range and the number of answers behind it.

    Named, questions that don't mention you

    1
    this run

    31%

    25–38% range · 192 answers · previous 25% (19–32%)

    0255075100

    Up 6 points since 4 Nov · Not yet a clear change: the range of the change, −1 to +13 points, includes zero.

    By engine

    95% range
    ChatGPT17 of 4835%23–50%
    Claude16 of 4833%22–47%
    Gemini14 of 4829%18–43%
    Google AI Mode12 of 4825%15–39%

    Each engine is measured on its own answers; ranges this wide can't rank the engines.

  3. Step 3

    Competitors on the same answers

    Each competitor is measured on exactly the answers the client is. A gap, or a change since the last run, is called only when its range excludes zero; otherwise the report says there is no clear change.

    On the same answers

    192 answers
    Workloft84 of 19244%Ahead by 13
    Taskmoor59 of 19231%You
    Planhaven42 of 19222%Behind you by 9
    Stackwell17 of 1929%Behind you by 22

    A gap is called only when its 95% range, paired over the same questions and engines, excludes zero.

  4. Step 4

    Why, and what to publish

    Each answer keeps the searches the engine reported and the sources it cited, so you can show a client which sites decide their category. On Growth and Agency, action plans turn the gaps into page-by-page work.

    Question the engines searchedSearched inYouMost namedStatus
    best project management software for agencies64%123 of 1923 of 12Workloft · 9Opportunity
    SOC 2 compliant project management tool27%52 of 1921 of 5Stackwell · 3Opportunity
    how to migrate projects from spreadsheets22%42 of 1920 of 4Workloft · 2No page answers it
    kanban vs gantt for client projects16%31 of 1920 of 3Planhaven · 2Opportunity
    resource planning software for agencies41%79 of 1926 of 8Taskmoor · 6You lead

    Sources that leave you out

    cited in
    saas-reviews.example53%
    workloft.example49%
    pm-compare.example36%

    What the engines searched

    this run

    38 questions the engines searched on their own while answering yours. The ones searched most, where competitors are named and you aren't, are marked as opportunities.

  5. Step 5

    A brief per client, shared by link

    Each week's brief is written from the run's own figures: where the client stands, what changed, and what to do next. Send the link; it opens that brief only. On Agency, the brief carries your agency's name, logo and colour.

    Brief · week of 10 Nov

    shared by link

    Taskmoor is named in 31% of answers to questions that don't mention it (range 25–38%, 192 answers).

    Up 6 points since 4 Nov. Not yet a clear change: the range of the change, −1 to +13 points, includes zero. Workloft still leads on agency buying questions.

    1. 01Answer “best project management software for agencies”
    2. 02Get reviewed on saas-reviews.example
    3. 03Ask “how to migrate from spreadsheets” directly

    Changes are observed between runs, never claimed as caused.

Sample dataTaskmoor is a fictional brand, its rivals and sources too

What it costs

Two plans fit an agency.

One answer is one question asked to one engine once. Answers a month are shared across every client in the workspace.

Growth and Agency plans compared
Plan limitGrowthAgency
Price a month$149$399
Answers a month4,0005,000
Engines asked todayChatGPT, Gemini and Google AI ModeChatGPT, Claude, Gemini and Google AI Mode (Claude up to 1,000 answers a month)
Brands (one per client)1050
Questions100500
Seats1025
White-label briefs: your name, logo and colourNoYes
Action plansYesYes
Alert webhooksNoYes
CSV exportYesYes

Billed monthly in US dollars, plus VAT where it applies. Cancel any time. Every plan, compared.

Worked example

Eight clients, weekly

8 clients with 12 questions each is 96 questions. Asked weekly (up to 5 runs in a month) to ChatGPT, Gemini and Google AI Mode, that is 8 × 12 × 3 × 5 = 1,440 answers a month. It fits Growth: 4,000 answers, 100 questions and 10 brands.

Thirty clients, monthly

30 clients with 15 questions each is 450 questions, more than Growth's 100 and more brands than its 10. On Agency, asked monthly to ChatGPT, Claude, Gemini and Google AI Mode, that is 1,800 answers a month, 450 of them from Claude (up to 1,000 a month). It fits Agency. Weekly, the same setup would use 9,000 answers, 2,250 from Claude, more than Agency's 5,000 and its Claude allowance: run the larger clients weekly and the rest monthly, or talk to us about Enterprise.

The run-size calculator does this arithmetic for your own numbers.

Limits

What it doesn't do

  • It doesn't promise rankings. It shows what changed after you acted, as observed, never that your work caused it.
  • It measures fresh buyer sessions, not a signed-in person's history, which can move that person's own answers. The method explains why that is the neutral baseline.
  • Briefs open on app.amplirank.com. On Agency they carry your agency's name, logo and colour; serving them from your own domain isn't available yet.
  • It asks only the engines switched on. AmpliRank supports ChatGPT, Claude, Gemini, Perplexity and Google AI Mode; a plan never asks an engine that isn't running.

Before you choose any tool for client reporting, the 8 questions to ask a vendor are a useful check, AmpliRank included.

Questions

Questions agencies ask.

How many clients fit a plan?
One brand per client: Growth holds 10 brands and 100 questions, Agency 50 brands and 500 questions. Answers a month are shared across all clients in the workspace. Enterprise sets custom limits.
Can each client see only their own report?
Yes. Each brief is shared by its own link, and the link opens that brief only. Seats belong to your agency's workspace, not to clients.
Can the brief carry our agency's branding?
Yes, on Agency. Shared briefs carry your company name, logo, colour and footer, and can leave out AmpliRank's name. They open on app.amplirank.com; briefs on your own domain aren't available yet.
How do I explain a range to a client?
Say the rate, the range and the count together: "Named in 30.2% of answers, very likely between 24.2% and 37.0%, from 192 answers." The free AI visibility calculator shows how more answers narrow it.
Can we use our own questions?
Yes. Drafting from the client's site is a starting point on paid plans, and you review every drafted question before it is asked. You can write your own questions on any plan and keep them in sets.

■ Ask · Measure · Compare · Trace · Fix · Share

See what AI says about your clients.

Cookies