---
title: "From White-Label SEO to AEO: The Agency Retainer’s Next Pricing Shift"
description: "A timeline of how search retainers moved from rank reports to AI-answer monitoring—and why agencies selling AEO should buy execution per domain, not prompts by the hundred."
url: "https://groas.com/post/from-white-label-seo-to-white-label-aeo"
image: "https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/026115b6-89f1-461e-a4b5-79d2e8a48e0a.png"
published: "2026-10-08T05:24:53.453Z"
modified: "2026-10-08T05:24:53.524Z"
---

October 8, 2026 · 9 min read

# From White-Label SEO to AEO: The Agency Retainer’s Next Pricing Shift

[Alexander PerelmanHead Of Product @ groas](https://groas.com/author/alexander-perelman)

![A man feeds coins into a seaside viewer labelled "1 PROMPT" to watch a run-down house whose scaffolding stands empty—paying to look, not to fix.](https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/026115b6-89f1-461e-a4b5-79d2e8a48e0a.png)

In this article

1. [2010s: White-label SEO makes the report the product](#2010s-white-label-seo-makes-the-report-the-product)
2. [May 2023 to October 2024: Search answers outgrow the rank table](#may-2023-to-october-2024-search-answers-outgrow-the-rank-table)
3. [Late 2024 to early 2025: Prompt monitors inherit the keyword meter](#late-2024-to-early-2025-prompt-monitors-inherit-the-keyword-meter)
4. [May 2025 to early 2026: AI Mode stretches the query, not the retainer](#may-2025-to-early-2026-ai-mode-stretches-the-query-not-the-retainer)
5. [October 2026: Execution per domain becomes the entry point](#october-2026-execution-per-domain-becomes-the-entry-point)

Agencies are buying prompt allowances for AEO just as they once bought keyword allowances for SEO. I would not build a retainer around either one: the useful unit is a client domain where someone does the work, not a question someone counts.

The pattern is familiar from PPC. Bid tools gave way to scripts, then to systems that took on more of the account work; the value moved away from watching and adjusting individual bids. Search retainers are following a similar line from rank reports to AI-visibility monitors to execution platforms. **The agency that sells a fixed monthly result cannot afford to buy its delivery by the counted prompt.** Here is where that pricing logic came from, and where I think it goes next.

## 2010s: White-label SEO makes the report the product

The original white-label SEO arrangement was easy to understand. An agency licensed a reporting platform, put it behind [its own domain and branding](https://www.vendasta.com/blog/white-label-seo-dashboard/), and gave clients rank trackers, site audits and PDFs bearing the agency logo. The client could see whether a keyword moved. The report did not necessarily show who changed a page, built a link or fixed a technical issue between check-ins.

That distinction mattered less while rankings were the main thing a client expected to see. A table of positions looked like a reasonable account of progress, even when the labor that might move those positions sat somewhere else in the retainer. **The report was easy to white-label; the work still had to happen.**

The software bill was also fairly legible. One comparison lists [AgencyAnalytics at around $20 per client per month, Swydo with a base charge plus per-source pricing, and Whatagraph starting around $812 per month](https://www.vendasta.com/blog/white-label-seo-dashboard/). If an agency charged $1,500 a month for SEO, it could estimate the reporting cost per client and budget separately for content, links and technical work. Whether that was a good deal depended on the work, but the unit was clear.

PPC had its own version of this bargain. Match type structures, bid tools and scripts all gave operators ways to manage more of an account. As more routine decisions moved into systems, billing as though every adjustment still took the same human effort became harder to defend. I am not saying PPC became a per-domain product; I am saying its tooling exposed the weakness in charging for an old unit after the work behind that unit changed.

For SEO agencies, the rank report faced a sharper problem. It measured a position on a results page. Once the results page began answering the question itself, position alone could no longer explain what the searcher saw. **Keep the report if it helps, but do not mistake it for the whole deliverable.**

## May 2023 to October 2024: Search answers outgrow the rank table

Google [announced Search Generative Experience at I/O in May 2023 as an experiment through Search Labs](https://www.engadget.com/google-search-generative-experience-preview-a-familiar-yet-different-approach-160018022.html). For an agency, the immediate lesson was not to rename an SEO package. It was that the space its rank tracker measured would increasingly share attention with a generated answer.

Then the experiment moved into the main product. [AI Overviews began rolling out in the U.S. on May 14, 2024](https://www.techrepublic.com/article/google-io-2024-highlights/). [ChatGPT Search launched in October 2024 with web information and links to sources](http://techcrunch.com/2024/12/16/openai-brings-its-ai-powered-web-search-tool-to-more-chatgpt-users/). An agency now had two related questions to answer: where does a client rank, and does the client appear in an answer that may satisfy the searcher before a click?

Those questions are not interchangeable. A page can have a conventional ranking worth reporting while an answer above or elsewhere draws on different sources. A client looking at the old PDF could reasonably ask why its No. 4 ranking was not reflected in the answer it saw. The PDF had not become false. It had become incomplete.

This is where the retainer starts to lose its old pricing logic. If the agency promises visibility, it needs to see more than a list of ranked links. If it promises to improve visibility, it also needs a way to change the pages and signals an answer might use. **Measuring the new surface is necessary; selling measurement as the finished work repeats the old mistake.**

## Late 2024 to early 2025: Prompt monitors inherit the keyword meter

The first practical response was to watch the answers. A monitor lets an agency choose questions and models, then track whether a client appears. That is useful. It tells the team what a rank table cannot: which answers mention the client and where gaps appear.

The pricing should look familiar to anyone who has bought a rank tracker. One comparison lists [Otterly Lite at $29 for 15 prompts, Peec Starter at $80 for 50 prompts across three models, Scrunch Core at $250 for 125 unique prompts, and Profound Starter at $99 for ChatGPT-only tracking](https://www.sona.com/blog/how-to-choose-an-aeo-checker-that-measures-more-than-mentions/). The allowances are not identical, so I would not reduce them to a single neat cost-per-prompt league table. The shared feature matters more: coverage is tied, in one way or another, to what you count.

That gets awkward inside a flat agency retainer. [Some tools count a prompt separately on each platform](https://www.mywebaudit.com/blog/aeo-prompt-strategy-for-agencies). Under that method, checking 150 platform-and-prompt combinations across seven platforms gives you roughly 21 distinct questions, not 150. Say ten clients each need 30 questions checked on Google, ChatGPT and Perplexity. That is 300 distinct client questions but 900 counted combinations.

The point is not that every client needs that coverage. It is that a proposal can say “30 questions per client” while the invoice counts something else. Add an answer surface and the questions the strategist cares about may stay the same, but the metered volume rises. The agency has already promised its client a monthly price. It now has to ration coverage, absorb the added cost or reopen the retainer.

A monitor can still be the right purchase for an audit. It can show a client what appears today and help identify the next job. But a monthly retainer normally sells continued progress. After the monitor finds a missing citation or an unhelpful page, somebody must decide what to change and do it. **A prompt allowance is a measurement budget, not a delivery plan.**

## May 2025 to early 2026: AI Mode stretches the query, not the retainer

Google [tested AI Mode in Labs before rolling it out in the U.S. on May 20, 2025, without a Labs sign-up](https://blog.google/products-and-platforms/products/search/google-search-ai-mode-update/). Its query fan-out matters more to agency operations than the product name: one question can lead the system through multiple sub-queries before it assembles an answer. A short keyword list is a poor stand-in for that process.

The difference in input length makes the problem easier to picture. One agency prompt-strategy discussion puts a typical [Google search at about four words, an AI Mode query at about seven, and a ChatGPT query at about 23](https://www.mywebaudit.com/blog/aeo-prompt-strategy-for-agencies). These are useful approximations, not instructions to write 23-word prompts. They show why extending an old keyword list with a few extra words will not, by itself, tell an agency how its client appears in answers.

Longer questions and fan-out also make a single ranking line less informative. A client may ask about a service in several ways, and an answer may draw on several sources. The agency needs to choose representative questions, inspect the answers and act on what it finds. Merely buying a larger prompt cap makes the inspection broader; it does not make the response better.

That is the bridge from monitoring to execution. **When each new question creates another item to watch, the retainer needs a fixed unit of work—not an ever-expanding meter for watching it.**

## October 2026: Execution per domain becomes the entry point

My reading of the line now is that tracking does not disappear. It gets absorbed into a service that can also act: publish content, address technical issues and work on citations, then show what happened to visibility. The record here is clearer on the available agency model than on precisely when a whole market changed course, so I would not pretend every platform has made that turn.

Groas offers agencies [SEO and AI Search from $199 a month per client domain, including content, technical fixes and citations](https://groas.com/for-agencies). Google and AI visibility tracking sits alongside that work. Reports carry the agency’s branding, and a strategist supports its team rather than stepping in front of the client. The purchased unit is a domain with delivery attached, not a bundle of 15 or 125 prompts that the agency has to spread across clients and engines.

**Monitoring tells you what the answer says; execution gives you a way to change what it may say next.** That is the difference I would put at the center of an AEO retainer, rather than hiding it below a dashboard screenshot.

![White-label SEO reports beside screenshots of AI-generated answers](https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/60978a89-d361-43e6-8ef7-096a8ae512d0.png)

The margin comparison needs honest arithmetic. Say you manage ten clients on $1,500 monthly retainers. On a monitoring-led path, a $250 plan with 125 counted prompts does not cover the 900 combinations in the earlier example. You would need more capacity, and the price of that capacity depends on the tool and plan. You would also need the people who turn findings into page changes, technical fixes and citation work. Four hours per client at $75 an hour fully loaded is $300 in labor before software and overhead.

On an execution-led path, the quoted groas starting price is $199 per client domain. Subtract that from a $1,500 retainer and $1,301 remains before strategy time and overhead. That is not a promise of $1,301 profit, and I would not compare it with a monitor’s sticker price as though both purchases did the same job. One is primarily a way to observe; the other includes specified work. **Compare what must still be done after you pay the bill.**

![Chart contrasting costs tied to tracked prompts with a flat per-domain fee](https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/47087c24-4f73-4779-80bc-aaa187fdf4e4.png)

That distinction should shape the package an agency sells. I would put one client domain on the proposal as the unit, with AI visibility tracking across Google and AI answers alongside the content, technical and citation work intended to improve it. Keep the strategic decisions and the client relationship with the agency. Put the work that can be delivered under its name into the delivery model. [Groas describes that white-label arrangement for agencies](https://groas.com/for-agencies).

Before signing for any white-label service, I would get four answers in writing:

1. **How is one prompt counted** when the same question is tracked on four engines?
2. What gets done for each domain each month besides producing a chart?
3. What happens to the agency’s cost if another model is added?
4. Can the agency test the service on a live client account before moving ten accounts onto it?

The last question cuts through a lot of polished demos. [Groas offers agencies seven days of the full service on one real client account before payment](https://groas.com/for-agencies). Use that week to look at what gets done and what still falls to your team. A live account is a better test of a delivery promise than a tidy sample report.

I would not tell every agency to abandon monitors. If you sell one-off audits and the client implements your recommendations in-house, a monitor may be exactly what you need. You are selling diagnosis. If you sell a monthly retainer and own the ongoing result, buying diagnosis alone leaves the expensive part of the job sitting with you.

That is where this timeline points. Rank reports could not account for generated answers. Prompt monitors made those answers visible, then brought back the per-unit meter. Per-domain execution puts measurement beside the work it is meant to inform. For a retainer agency, I would buy for that last step: test one real domain, price the package around the work, and keep the strategy and client relationship for yourself. Most agencies will keep buying prompts by the hundred and billing hours by the four. Let them.

## Frequently Asked Questions

### What was the original white-label SEO arrangement?

An agency licensed a reporting platform, put it behind its own domain and branding, and gave clients rank trackers, site audits and branded PDFs. The report showed whether a keyword moved, but not necessarily who changed a page, built a link or fixed a technical issue between check-ins.

### Why can a client rank well on Google but not appear in AI-generated answers?

A conventional ranking and an AI answer are measured on different surfaces. An answer can draw on different sources than the ranked results page, so a No. 4 ranking may not be reflected in the answer the searcher sees. The old rank report has not become false, but it has become incomplete.

### How are AEO prompt monitoring tools priced?

Most price by counted prompt allowances, similar to rank trackers. Examples include Otterly Lite at $29 for 15 prompts, Peec Starter at $80 for 50 prompts across three models, Scrunch Core at $250 for 125 unique prompts, and Profound Starter at $99 for ChatGPT-only tracking. Coverage is tied to what you count.

### Why does checking the same question on multiple AI platforms cost more prompts?

Some AEO tools count a prompt separately on each platform. So 30 distinct client questions checked on Google, ChatGPT and Perplexity become 90 counted combinations. Ten clients needing that coverage would hit 900 counted combinations, which can make an invoice balloon beyond the '30 questions per client' a proposal promises.

### How long are AI Mode and ChatGPT queries compared to Google searches?

One agency prompt-strategy discussion puts a typical Google search at about four words, an AI Mode query at about seven, and a ChatGPT query at about 23. These are approximations, but they show why extending an old keyword list with a few extra words will not reveal how a client appears in answers.

### What unit should an agency use to price an AEO retainer?

The article argues the useful unit is a client domain with delivery attached, not a bundle of counted prompts. Groas offers SEO and AI Search from $199 a month per client domain, including content, technical fixes and citations, with Google and AI visibility tracking alongside the work and reports carrying the agency's branding.

### When is an AEO prompt monitor the right purchase?

A monitor can be the right purchase for a one-off audit, showing a client what appears today and helping identify the next job. If the client implements your recommendations in-house, you are selling diagnosis. But if you sell a monthly retainer and own the ongoing result, buying diagnosis alone leaves the delivery work sitting with you.

## Pay For Results, Not For Hours

Businesses buy the outcome, agencies resell it, and groas answers for it either way.

[See If You Qualify](https://groas.typeform.com/to/xC1bQNUT)

## Structured data

```json
{"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://graph.groas.com/entity/groas","name":"groas","url":"https://groas.com/","logo":{"@type":"ImageObject","url":"https://groas.com/icon-512.png","width":512,"height":512},"sameAs":["https://www.linkedin.com/company/groas/"]},{"@type":"WebSite","@id":"https://groas.com/#website","name":"groas","url":"https://groas.com/","inLanguage":"en-US","publisher":{"@id":"https://graph.groas.com/entity/groas"}},{"@type":"WebPage","@id":"https://groas.com/post/from-white-label-seo-to-white-label-aeo#webpage","url":"https://groas.com/post/from-white-label-seo-to-white-label-aeo","name":"From White-Label SEO to AEO: The Agency Retainer’s Next Pricing Shift","description":"A timeline of how search retainers moved from rank reports to AI-answer monitoring—and why agencies selling AEO should buy execution per domain, not prompts by the hundred.","inLanguage":"en-US","isPartOf":{"@id":"https://groas.com/#website"},"about":{"@id":"https://graph.groas.com/entity/groas"},"primaryImageOfPage":{"@type":"ImageObject","url":"https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/026115b6-89f1-461e-a4b5-79d2e8a48e0a.png"},"breadcrumb":{"@id":"https://groas.com/post/from-white-label-seo-to-white-label-aeo#breadcrumb"},"dateModified":"2026-10-08T05:24:53.524Z","mainEntity":{"@id":"https://groas.com/post/from-white-label-seo-to-white-label-aeo#article"}},{"@type":"BreadcrumbList","@id":"https://groas.com/post/from-white-label-seo-to-white-label-aeo#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://groas.com/"},{"@type":"ListItem","position":2,"name":"Blog","item":"https://groas.com/blog"},{"@type":"ListItem","position":3,"name":"From White-Label SEO to AEO: The Agency Retainer’s Next Pricing Shift","item":"https://groas.com/post/from-white-label-seo-to-white-label-aeo"}]},{"@type":"BlogPosting","@id":"https://groas.com/post/from-white-label-seo-to-white-label-aeo#article","url":"https://groas.com/post/from-white-label-seo-to-white-label-aeo","mainEntityOfPage":{"@id":"https://groas.com/post/from-white-label-seo-to-white-label-aeo#webpage"},"isPartOf":{"@id":"https://groas.com/blog#blog"},"headline":"From White-Label SEO to AEO: The Agency Retainer’s Next Pricing Shift","description":"A timeline of how search retainers moved from rank reports to AI-answer monitoring—and why agencies selling AEO should buy execution per domain, not prompts by the hundred.","datePublished":"2026-10-08T05:24:53.453Z","dateModified":"2026-10-08T05:24:53.524Z","image":{"@type":"ImageObject","url":"https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/026115b6-89f1-461e-a4b5-79d2e8a48e0a.png"},"author":{"@id":"https://groas.com/author/alexander-perelman#person"},"publisher":{"@id":"https://graph.groas.com/entity/groas"},"keywords":"agency, client, seo, work, white-label, aeo, per, white-label seo","wordCount":1976,"inLanguage":"en-US"},{"@type":"Person","@id":"https://groas.com/author/alexander-perelman#person","name":"Alexander Perelman","jobTitle":"Head Of Product @ groas","description":"Ex Goldman Sachs and Ex Stanford Computer Science","image":"https://groas.com/media/blog/4f17cac3a81acc224cc1b2cfab8f94bbc295a3089d13e8421125771ef8d4064f.jpg","worksFor":{"@id":"https://graph.groas.com/entity/groas"},"sameAs":["https://www.linkedin.com/in/alexander-433793253/"]},{"@type":"FAQPage","@id":"https://groas.com/post/from-white-label-seo-to-white-label-aeo#faq","mainEntity":[{"@type":"Question","name":"What was the original white-label SEO arrangement?","acceptedAnswer":{"@type":"Answer","text":"An agency licensed a reporting platform, put it behind its own domain and branding, and gave clients rank trackers, site audits and branded PDFs. The report showed whether a keyword moved, but not necessarily who changed a page, built a link or fixed a technical issue between check-ins."}},{"@type":"Question","name":"Why can a client rank well on Google but not appear in AI-generated answers?","acceptedAnswer":{"@type":"Answer","text":"A conventional ranking and an AI answer are measured on different surfaces. An answer can draw on different sources than the ranked results page, so a No. 4 ranking may not be reflected in the answer the searcher sees. The old rank report has not become false, but it has become incomplete."}},{"@type":"Question","name":"How are AEO prompt monitoring tools priced?","acceptedAnswer":{"@type":"Answer","text":"Most price by counted prompt allowances, similar to rank trackers. Examples include Otterly Lite at $29 for 15 prompts, Peec Starter at $80 for 50 prompts across three models, Scrunch Core at $250 for 125 unique prompts, and Profound Starter at $99 for ChatGPT-only tracking. Coverage is tied to what you count."}},{"@type":"Question","name":"Why does checking the same question on multiple AI platforms cost more prompts?","acceptedAnswer":{"@type":"Answer","text":"Some AEO tools count a prompt separately on each platform. So 30 distinct client questions checked on Google, ChatGPT and Perplexity become 90 counted combinations. Ten clients needing that coverage would hit 900 counted combinations, which can make an invoice balloon beyond the '30 questions per client' a proposal promises."}},{"@type":"Question","name":"How long are AI Mode and ChatGPT queries compared to Google searches?","acceptedAnswer":{"@type":"Answer","text":"One agency prompt-strategy discussion puts a typical Google search at about four words, an AI Mode query at about seven, and a ChatGPT query at about 23. These are approximations, but they show why extending an old keyword list with a few extra words will not reveal how a client appears in answers."}},{"@type":"Question","name":"What unit should an agency use to price an AEO retainer?","acceptedAnswer":{"@type":"Answer","text":"The article argues the useful unit is a client domain with delivery attached, not a bundle of counted prompts. Groas offers SEO and AI Search from $199 a month per client domain, including content, technical fixes and citations, with Google and AI visibility tracking alongside the work and reports carrying the agency's branding."}},{"@type":"Question","name":"When is an AEO prompt monitor the right purchase?","acceptedAnswer":{"@type":"Answer","text":"A monitor can be the right purchase for a one-off audit, showing a client what appears today and helping identify the next job. If the client implements your recommendations in-house, you are selling diagnosis. But if you sell a monthly retainer and own the ongoing result, buying diagnosis alone leaves the delivery work sitting with you."}}]}]}
```