Agencies

GEO Agency Guide: How to Package and Sell AEO Services

How a GEO or AEO agency can package AI visibility work: the audit as the entry offer, what goes in a proposal, how to price it, and how to report results.

The simplest way for an agency to sell GEO or AEO services is to start with a fixed-price AI visibility audit, then sell the fixes it uncovers as a project or retainer. The audit shows a client, in their own category’s questions, how often ChatGPT, Gemini and Google’s AI Overviews name them against competitors, and that evidence does most of the selling. Price the follow-on work like the SEO and PR work it largely is, and report mention rate across many questions rather than promising positions.

Is there real demand for a GEO or AEO agency?

Yes, and it is growing from a small base. In DataForSEO’s US data for November 2026, “geo agency” and “aeo agency” each averaged 720 searches a month. In the latest month “geo agency” reached 880 (from 320 a year earlier) and “aeo agency” reached 1,600 (from 260). “ai visibility audit” went from 70 to 1,600 over the same year.

The client side of that demand is easy to explain. A Pew Research Center study of 900 US adults found Google users clicked a regular result in 8% of visits when an AI summary appeared, against 15% without one. Marketing leads see traffic flatten and want to know whether they are in the answers that replaced the clicks. Most do not know how to check.

You do not need to rebrand as a “GEO agency” to sell this. Existing SEO and content agencies already have most of the skills. What they usually lack is a way to measure AI answers and a deliverable that makes the problem visible.

What are clients actually buying?

Strip away the labels and a client wants three things:

  1. To know where they stand. Do assistants mention them for the questions their buyers ask? Who gets named instead?
  2. To know what to do about it. A short list of changes with reasons, not a 60-page deck.
  3. To see whether it worked. The same measurement, repeated after the work is done.

GEO, AEO and LLM SEO are different names for overlapping work. Sell the outcome, and use whichever term your client already uses.

Why lead with an audit?

An audit is the natural entry offer for AI visibility work, for four reasons.

  • It answers a question the client already has. “Do AI assistants mention us?” is the question behind most inbound interest.
  • The evidence persuades. A table showing a competitor named in 70% of answers and the client in 15% does more than any pitch.
  • It scopes the next project. The gaps it finds (missing questions, sites cited instead of the client, crawler blocks, negative descriptions) become the work plan.
  • It is low risk to buy. A fixed fee for a defined deliverable is easier to approve than an open-ended retainer for a new discipline.

Use it as a paid diagnostic, or run a shorter version free as part of a pitch. Either way, the full audit should be a real deliverable, not a teaser.

What the audit should cover

A useful GEO audit or AEO audit measures:

  • Mention rate per engine across 15 to 30 buyer questions that do not contain the brand name, each asked more than once on engines whose answers vary.
  • Position in the answer: named first, in the middle, or as an afterthought.
  • Competitors and share of voice: who is named instead, and the client’s share of all brand mentions.
  • Citations: which websites the assistants cite, and how often the client’s own site is one of them.
  • Sentiment and themes: what the answers praise or criticise.
  • Technical access: whether robots.txt or a firewall blocks the crawlers that feed AI search.
  • Branded questions, kept separate: how the assistant describes the brand when asked directly.

Our step-by-step audit guide covers the method in detail.

How should you price GEO and AEO services?

There is no reliable industry benchmark yet. Most published figures come from agencies pricing their own services, so treat every range below as a rough indication.

Source What it says Caveat
First Page Sage Median GEO/AEO retainer of $6,850 a month; traditional SEO median $4,740 Published by an agency that sells GEO
WebFX Agency retainers $1,500 to $50,000+ a month; projects $5,000 to $50,000; AI visibility tracking software $50 to $1,000+ a month Very wide; agency-published
Digital Agency Network One-time GEO audits $2,000 to $8,000; add-on to an SEO retainer +$1,500 to $3,000 for small and mid-sized clients No methodology given for the figures

A few patterns hold across these sources and are worth borrowing:

  • Audits are priced as fixed-fee projects. A few thousand dollars is a common quote for small and mid-sized clients; smaller agencies and local clients will sit below that.
  • Ongoing work is priced as a retainer or an add-on. Many SEO agencies add AI visibility to an existing SEO retainer rather than selling it separately, which is easier for existing clients to accept.
  • Implementation costs more than advice. Content production, digital PR and technical work are where the hours go.

A simple three-step package

One structure that maps onto the work:

Offer What it includes How to price
AI visibility audit Question set, measurement on each engine, competitor and citation analysis, branded report, 30-minute walkthrough, prioritised fix list Fixed fee
Fix sprint Crawler and indexing fixes, new or rewritten pages for missing questions, outreach to the sites that get cited Fixed fee or monthly, 2 to 3 months
Ongoing Content and PR against the fix list, with a fresh audit each quarter to compare Monthly retainer or add-on to SEO

Work out your cost per audit before you set a price: your tool subscription divided by audits per month, plus the analyst time to choose questions, check matches, write the summary and present it. The analyst time is usually the larger part, and it is what the client is paying for.

What should go in a GEO or AEO proposal?

Keep it short and specific to the client. A proposal that works usually has:

  1. The problem in their terms. Three or four of their buyers’ questions, and what the assistants answered. If you have run a free sample, include it.
  2. What you will measure. The engines, the number of questions, how many times each is asked, and what counts as a mention.
  3. What you will deliver. The report, the fix list, the walkthrough, and any implementation.
  4. What you will not promise. Positions in ChatGPT, guaranteed citations, or results on engines you do not measure.
  5. How success is judged. Mention rate and share of voice on the same question set, measured again after a stated period.
  6. Price and terms. Fixed fee for the audit; monthly figure for anything ongoing.

The “will not promise” section protects you. A SparkToro study with Gumshoe.ai found AI tools returned the same brand list less than 1% of the time across repeated runs. Promising “#1 in ChatGPT” sets up an argument you cannot win. Promising a measured improvement in how often the brand appears across a fixed set of questions is defensible.

How should you report AI visibility results?

Clients read the first page and skim the rest, so put the answer up top.

  • Lead with one number per engine. “ChatGPT named you in 18% of answers; your closest competitor was named in 64%.”
  • Show the competitor comparison next. Share of voice in a single chart or table.
  • Show the sources. Which sites the assistants cited. These are your PR and content targets.
  • Quote the answers. A short passage from a real answer convinces more than a percentage.
  • Be clear about what is a reading. Sentiment is a model’s interpretation of the text, not a hard measurement. Say so.
  • Say what is missing. If some checks failed or an engine did not answer, report how many.
  • End with the fix list, ranked by likely impact and effort.

Where do white-label reports fit?

Clients buy from the agency, not from the agency’s tools. A white-label report lets you hand over a finished deliverable under your own name, logo and colours, without explaining which software produced it or building reports by hand.

That matters most at the audit stage, where the report is the product. It also saves time. Assembling screenshots of AI answers into a slide deck for every client is slow and hard to repeat consistently; a generated report you check and annotate is faster and the same every time.

What to check in any white-label setup: what the client actually sees (branding, links, any mention of the vendor), whether you can edit the summary and recommendations before publishing, whether you can correct wrong brand matches, and whether you can export the raw answers if the client asks how a number was reached.

Mistakes to avoid

  • Overclaiming coverage. If you measure three engines, say three. Do not imply every AI assistant is covered.
  • Measuring once. A single run per question is noise. Ask more than once and report rates.
  • Using branded questions as the score. “What is [brand]?” will always name the brand. Keep those separate.
  • Selling tactics without evidence. llms.txt files and “AI schema” are cheap to add but have no evidence of impact, and Google says its Search ignores llms.txt (Google’s AI optimization guide). See LLM SEO: evidence vs guesswork.
  • Skipping the re-audit. Without a second measurement on the same questions, you cannot show the client what changed.

Running audits under your brand with AI Visibility Desk

AI Visibility Desk was built for this entry offer. You add your agency’s name, logo, colour and contact line once, enter a client and its competitors, and choose up to 30 buyer questions (or have it read the client’s site and suggest them). It asks each question on ChatGPT, Gemini, Google AI Overviews, Google AI Mode, Perplexity and Claude, then gives you mention rates per engine, position, share of voice, cited sites, sentiment, crawler checks and every answer to review. You can draft and edit the summary, then publish the report under your brand at a private link, print it to PDF, or download the answers as CSV.

Paid plans add weekly or monthly monitoring for the retainer, pitch audits for prospects that don’t use the allowance, and a white-label SEO audit. Plans run from one free audit to Studio at $349 a month for 100 audits. See pricing or how it works for agencies.

Frequently asked questions

What does a GEO agency do?

A GEO agency helps brands appear in answers from AI assistants such as ChatGPT, Gemini and Google AI Overviews. The work usually combines measurement (which questions name the brand and which competitors win), technical fixes such as crawler access, content that answers buyer questions directly, and PR to earn mentions on sites the assistants cite.

How much do agencies charge for GEO and AEO services?

Published ranges are wide and mostly come from agencies themselves. First Page Sage reports a median GEO and AEO retainer of $6,850 a month. WebFX quotes $1,500 to $50,000 or more a month. One-time audits are often quoted in the low thousands. Treat all of these as rough guides, not benchmarks.

What is the best entry offer for selling AI visibility services?

A fixed-price AI visibility audit. It answers the question clients already have (do assistants mention us?), shows competitors winning in plain view, and produces a list of fixes that becomes the scope for ongoing work. It is quick to deliver and easy for a client to say yes to.

Can an agency guarantee results in ChatGPT or AI Overviews?

No, and it should not try. AI answers change between runs, and research by SparkToro found that AI tools repeated the same brand list less than 1% of the time. Promise the work and the measurement method, report mention rate across many questions, and avoid promising positions or specific citations.

Written by the AI Visibility Desk team

We build AI visibility audits for agencies and run them across ChatGPT, Gemini, Google AI Overviews, AI Mode, Perplexity and Claude every day. Guides are checked against what the audits show and updated when the engines change. About us.

Run your first audit free

Enter a client’s website, let the AI fill in competitors and buyer questions, and get a branded report across six AI engines. No card needed.

Start free