---
title: "The AEO Retainer That Dies in Month Four Was Broken at Purchase"
description: "A month-by-month postmortem of an agency AEO retainer built on a monitoring tool: the prompt caps, unpriced fixes, and purchase decision behind the cancellation."
url: "https://groas.com/post/anatomy-of-an-aeo-retainer-that-dies-in"
image: "https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/9c594efc-de61-4151-bb21-fdd22ca28429.png"
published: "2026-10-08T05:43:40.265Z"
modified: "2026-10-08T05:43:40.329Z"
---

October 8, 2026 · 9 min read

# The AEO Retainer That Dies in Month Four Was Broken at Purchase

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

![A pristine gauge reads red beside a leaking pipe while the toolbox below holds only more gauges: measuring the problem instead of fixing it.](https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/9c594efc-de61-4151-bb21-fdd22ca28429.png)

In this article

1. [Before month one: the service is priced from a tool tier](#before-month-one-the-service-is-priced-from-a-tool-tier)
2. [Month one: the baseline earns trust it cannot yet repay](#month-one-the-baseline-earns-trust-it-cannot-yet-repay)
3. [Month two: the client asks what is being fixed](#month-two-the-client-asks-what-is-being-fixed)
4. [Month three: the agency buys more monitoring instead of fulfillment](#month-three-the-agency-buys-more-monitoring-instead-of-fulfillment)
5. [Month four: the same red cells arrive again](#month-four-the-same-red-cells-arrive-again)
6. [The fix belongs in the vendor decision](#the-fix-belongs-in-the-vendor-decision)

The AEO retainer did not die in month four. That is just when the cancellation email arrived. It died three months earlier, when the agency bought a monitoring tool and decided to resell it as a service.

This is not one agency’s account. It is a typical composite: clients ask why competitors appear in AI answers, an owner buys a tracker, the team turns its reports into a monthly deliverable, and everyone assumes they can work out the fixes later. I recognise the pattern from PPC dashboards. With AEO, the distance between a red cell on a report and the work needed to clear it can be wider. **The purchase decision sets the retainer up to fail.** Month four merely makes it visible.

## Before month one: the service is priced from a tool tier

First, the boundary around this story: nobody sent me a churned account, and no client here is being quoted. The composite draws on category pricing pages and [agency roundups that rank tools by starting price and prompt counts](https://www.rankability.com/blog/best-aeo-tools-for-agencies/). Otterly starts at $29, Peec around $95, Profound at $99 and $399, Semrush as a $99 add-on, and HubSpot at $50. Those tables help answer which tool to buy. They do not answer what happens when an agency puts that tool inside a retainer and promises a result.

Say you run a 12-person shop. Three clients ask why competitors show up in ChatGPT answers and they do not. You buy a credible monitor, put it on two $1,200-a-month retainers, and plan to work out fulfillment later. The purchase feels sensible. The tool gives your team something concrete to send on Friday, and $1,200 looks generous beside a software bill of $189.

That Friday deliverable is the trap. **You have priced the service from the software tier, so the tier now owns your scope.** The proposal says AEO. The operating plan says export a dashboard and hope the work behind it fits into the hours left over.

### The prompt cap becomes the service boundary

Scope gets set before anyone discusses the client’s actual questions. [Otterly Lite allows 15 prompts, Standard allows 100 prompts for $189, and Premium allows 400 for $489; Profound Starter allows 50 prompts on ChatGPT only, while Growth allows 100 prompts across three engines for $399](https://subscribepr.com/blog/best-aeo-tools-for-law-firms-2026/). Peec Starter starts at 50 prompts. Scrunch Core runs from 125 to 350.

One example makes the constraint plain: 4 practice areas times 3 cities times 8 phrasings comes to 96 prompts. That nearly fills a 100-prompt tier for one client before the agency decides whether those are even the best questions to follow. Add clients, engines, or new questions and the comfortable starter plan stops looking comfortable.

The costs also vary with what is tracked and how often. [A comparison of daily tracked prompts puts cheaper tools at $0.45 to $0.53 per prompt and Profound, Scrunch, and AthenaHQ at $1.50 to $2.46](https://www.sona.com/blog/how-to-choose-an-aeo-checker-that-measures-more-than-mentions/). That matters less as a neat price comparison than as a warning: the client thinks they bought coverage, while the agency bought a defined amount of observation. **A prompt cap is not an execution plan.**

## Month one: the baseline earns trust it cannot yet repay

The first report lands well. It shows share of voice by engine, prompts that mention the client, prompts that cite competitors, and a few red cells that look urgent. The client forwards it internally. Someone finally has a picture of the problem. A clean baseline can buy an agency time; I remember the same reaction to PPC audits.

But the baseline answers only the first question: where do we stand? It does not answer the one that will decide the renewal: who is going to change this?

That distinction is easy to miss when everyone is pleased to have a report. [Early agency check-ins matter because silence gets read as inactivity, and reports need to show issues fixed before rankings move](https://www.searchenginejournal.com/why-agencies-lose-clients-in-the-first-90-days-and-how-to-stop-it/582895/). The baseline gives the team something to discuss at the first check-in. It does not supply the work for the next one.

At this point the agency should be turning each useful finding into an owned task. Instead, the retainer is still organised around what the monitor can export. The client cannot see the difference yet. The agency can, if it looks past the PDF. **Month one measures the gap; it does not close it.**

## Month two: the client asks what is being fixed

Around week six, the client has seen two reports. The chart has moved a little, or not at all. They ask the reasonable question: what are you actually fixing?

This is the first sign that the retainer is in trouble. Calling it a reporting problem is the wrong diagnosis. A clearer chart will not give anyone CMS access or put a directory correction through. The tool was bought to observe; the proposal let the client believe observation came with execution.

[Monitoring tools report mentions, citations, and share of voice, but do not themselves improve visibility, guarantee citations, or verify accuracy](https://subscribepr.com/blog/best-aeo-tools-for-law-firms-2026/). Closing a gap can take better pages, profiles, reviews, and earned coverage. It takes somebody with access to the CMS, the content calendar, and outreach. I used to tell PPC clients that a new dashboard would sharpen decisions. It can sharpen the question. It still cannot do the work.

Look at the scope an agency might actually need to deliver. [One AEO engagement outline calls for six items: a baseline citation report, Organization schema with sameAs, a canonical entity page, a directory plan, comparison and listicle production, and monthly citation rate by engine](https://doodleweb.io/blog/best-ai-search-optimization-agencies-2026/). That is not a prettier export. It is a queue of content, technical, and distribution work, with permissions and people attached.

If the proposal sold blog posts plus FAQs under an AEO label, the mismatch is worse. The client is asking about missing citations while the agency is delivering the familiar SEO calendar with new vocabulary. In this composite, month two is when the team discovers the real scope one unbudgeted task at a time. **The question to answer is not what the dashboard found. It is who owns the fix.**

## Month three: the agency buys more monitoring instead of fulfillment

The wrong call comes next. The client wants more prompts tracked because the initial list missed service lines. The agency upgrades the tier to keep them happy. The report covers more ground. The underlying work queue does not move.

Then a third client signs on the same retainer template, and the vendor’s account or domain pricing starts to matter. [Semrush charges $99 a month per domain, Profound uses one workspace per account, and per-domain and per-prompt pricing can become expensive across 15 to 50 clients](https://www.rankability.com/blog/best-aeo-tools-for-agencies/). The $189 line in the original margin calculation was never a reliable estimate of what serving more clients would cost.

More important, it was never an estimate of fulfillment. Ten domains can mean ten sets of CMS permissions, content calendars, and people to chase for approvals. Dev hours per client CMS were not in the SOW because the SOW came from the tool’s feature list, not the work the findings would trigger. I have made the PPC version of that pricing mistake: tool plus report, with a prayer that delivery fits into whatever hours remain. It does not become a better plan because the dashboard has AI in its name.

The scale of the mismatch is visible in [ongoing AEO benchmarks of $3K to $10K a month, or $8K to $25K at content scale](https://doodleweb.io/blog/best-ai-search-optimization-agencies-2026/). Those are not instructions to charge every client that amount. They show why a $1,200 retainer built around a monitoring subscription cannot quietly absorb a substantial content and technical scope. An upgrade narrows the margin; unpriced work can erase it.

So the team makes smaller, less visible cuts. It tracks fewer questions than the client wants, trims engines, or lets findings wait because nobody has hours to act on them. The next report still arrives. **More measurement is not the fix for a fulfillment gap.**

## Month four: the same red cells arrive again

Three reports in, the dashboard shows the gaps it showed in month one. Missing citations. Competitor mentions. Red cells that now look less like opportunities and more like standing agenda items. The client has paid $3,600 and still cannot point to much that was fixed.

The email is polite: we are going to pause this for now. Nobody needs to dispute the data. The data did its job. The service did not deliver the work the data called for. **A report that repeats itself is not a retainer; it is a subscription the client could have bought directly.**

![Ink cartoon of an agency owner facing three identical monthly dashboards and a client email asking what was fixed](https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/e1d9b302-742a-45cc-ac58-7bec8ec0e4fc.png)

The damage extends beyond one cancelled agreement. This client is unlikely to refer another business for the same service, and the owner may walk away convinced AEO itself does not work. That is the wrong conclusion. Observation was sold as execution. The monitor could identify a gap, but no one had a funded, owned process to close it.

I recognise the old PPC mistake in that sequence. Agencies could wrap rank trackers and bid alerts in a management fee, send automated PDFs, and still leave clients asking what had changed in the account. A bid alert might at least point to a change someone could make in ten minutes. An AI visibility gap may call for a published page, a schema change, a directory cleanup, or a review profile the agency does not control. The meter is useful. It is not the workbench.

That is why the ending was set at purchase, not at the cancellation call. The only uncertainty was how many nearly identical reports it would take the client to notice.

## The fix belongs in the vendor decision

The decision that prevents this retainer happens before month one, in the vendor demo. When the tracker finds a missing citation or a weak entity page, ask who ships the fix. Then wait for the answer.

If the answer is your team, you may have bought a useful internal QA tool. You have not bought a service you can resell unchanged as an execution retainer. Your team still needs time, access, and a priced scope to publish pages, make technical changes, and work through citation gaps. Calling that labor ‘follow-up’ does not make it free.

![Miniature model of two AEO offers: a dashboard stack beside a workbench shipping fixes](https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/d0bf776d-f13f-4ba4-9c3e-5c9e07e7bc41.png)

This is why I start the vendor calculation with the work, then examine the meter. I use a [comparison of what AEO tools do, including caps and per-domain pricing](https://groas.com/post/the-best-aeo-and-ai-search-visibility-to) to see where monitoring costs could grow as clients ask more questions. A flat fee per client domain, without a prompt meter and with execution included, is a cleaner basis for the retainer described here. That is how groas prices its white-label SEO and AI search offering: from $199 a month per client domain for execution plus visibility tracking. Adding prompts does not create another prompt-based charge in that structure.

Pricing alone is not proof that fixes ship. Before putting any vendor into a proposal, I would still ask for the work behind the report: show a gap the system found and the corresponding fix with a timestamp. Ask whether publishing an entity page, making a schema change, or correcting a listing requires your own dev queue. Ask whether white-label reporting goes to the client under your brand or sends them to another login. The [five questions I would ask before buying another dashboard](https://groas.com/post/aeo-for-agencies-the-questions-i-keep-ge) come down to that separation between finding work and doing it.

If your team is responsible for the fixes, price and staff that responsibility before selling AEO. If you cannot do that yet, keep the tracker as an internal check and do not sell the retainer. My rule fits above the vendor spreadsheet: **if the tool cannot ship the fix, it does not go in the retainer by itself.**

## Frequently Asked Questions

### Why do AEO retainers built around a monitoring tool fail?

They fail because the agency prices the service from the software tier instead of from the work required. The proposal promises AEO results, but the operating plan is really just exporting a dashboard, and the prompt cap ends up defining the scope of the service.

### How many prompts do AEO monitoring tools let you track?

Otterly Lite allows 15 prompts, Standard 100 for $189, and Premium 400 for $489. Profound Starter allows 50 prompts on ChatGPT only, while Growth allows 100 across three engines for $399. Peec Starter starts at 50 prompts, and Scrunch Core runs from 125 to 350. A prompt cap is not an execution plan.

### Does a baseline AEO report close the gaps it identifies?

No. A baseline shows share of voice, prompts that mention the client, and prompts that cite competitors, so it answers where you stand. It does not answer who will change anything, and month one measures the gap rather than closing it.

### Do AEO monitoring tools actually improve AI visibility?

No. Monitoring tools report mentions, citations, and share of voice, but they do not improve visibility, guarantee citations, or verify accuracy. Closing a gap takes better pages, profiles, reviews, and earned coverage, plus someone with access to the CMS, content calendar, and outreach.

### Is buying a bigger monitoring tier a good way to fix a client's AEO results?

No. More prompts tracked can make the report cover more ground, but the underlying work queue does not move. More measurement is not the fix for a fulfillment gap, and upgrades can also narrow margins when per-domain and per-prompt pricing scales across clients.

### Why do clients cancel AEO retainers after a few months?

In the pattern described, the client cancels after three reports show the same red cells as month one. Having paid $3,600 with nothing visibly fixed, the client pauses the retainer. The data did its job; the service never delivered the work the data called for.

### What should an agency ask before buying an AEO monitoring tool?

Ask who ships the fix when the tracker finds a missing citation or a weak entity page, then wait for the answer. If the answer is your own team, you have bought an internal QA tool, not a resellable service, so price and staff that labor before selling the retainer.

### Should agencies sell an AEO retainer built only on a monitoring tool?

No. If the tool cannot ship the fix, it does not go in the retainer by itself. If the agency cannot price and staff the execution work, the tracker should stay an internal check rather than become the basis of a client-facing retainer.

## 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/anatomy-of-an-aeo-retainer-that-dies-in#webpage","url":"https://groas.com/post/anatomy-of-an-aeo-retainer-that-dies-in","name":"The AEO Retainer That Dies in Month Four Was Broken at Purchase","description":"A month-by-month postmortem of an agency AEO retainer built on a monitoring tool: the prompt caps, unpriced fixes, and purchase decision behind the cancellation.","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/9c594efc-de61-4151-bb21-fdd22ca28429.png"},"breadcrumb":{"@id":"https://groas.com/post/anatomy-of-an-aeo-retainer-that-dies-in#breadcrumb"},"dateModified":"2026-10-08T05:43:40.329Z","mainEntity":{"@id":"https://groas.com/post/anatomy-of-an-aeo-retainer-that-dies-in#article"}},{"@type":"BreadcrumbList","@id":"https://groas.com/post/anatomy-of-an-aeo-retainer-that-dies-in#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":"The AEO Retainer That Dies in Month Four Was Broken at Purchase","item":"https://groas.com/post/anatomy-of-an-aeo-retainer-that-dies-in"}]},{"@type":"BlogPosting","@id":"https://groas.com/post/anatomy-of-an-aeo-retainer-that-dies-in#article","url":"https://groas.com/post/anatomy-of-an-aeo-retainer-that-dies-in","mainEntityOfPage":{"@id":"https://groas.com/post/anatomy-of-an-aeo-retainer-that-dies-in#webpage"},"isPartOf":{"@id":"https://groas.com/blog#blog"},"headline":"The AEO Retainer That Dies in Month Four Was Broken at Purchase","description":"A month-by-month postmortem of an agency AEO retainer built on a monitoring tool: the prompt caps, unpriced fixes, and purchase decision behind the cancellation.","datePublished":"2026-10-08T05:43:40.265Z","dateModified":"2026-10-08T05:43:40.329Z","image":{"@type":"ImageObject","url":"https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/9c594efc-de61-4151-bb21-fdd22ca28429.png"},"author":{"@id":"https://groas.com/author/alexander-perelman#person"},"publisher":{"@id":"https://graph.groas.com/entity/groas"},"keywords":"retainer, month, client, aeo, agency, work, aeo retainer, month four","wordCount":1972,"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/anatomy-of-an-aeo-retainer-that-dies-in#faq","mainEntity":[{"@type":"Question","name":"Why do AEO retainers built around a monitoring tool fail?","acceptedAnswer":{"@type":"Answer","text":"They fail because the agency prices the service from the software tier instead of from the work required. The proposal promises AEO results, but the operating plan is really just exporting a dashboard, and the prompt cap ends up defining the scope of the service."}},{"@type":"Question","name":"How many prompts do AEO monitoring tools let you track?","acceptedAnswer":{"@type":"Answer","text":"Otterly Lite allows 15 prompts, Standard 100 for $189, and Premium 400 for $489. Profound Starter allows 50 prompts on ChatGPT only, while Growth allows 100 across three engines for $399. Peec Starter starts at 50 prompts, and Scrunch Core runs from 125 to 350. A prompt cap is not an execution plan."}},{"@type":"Question","name":"Does a baseline AEO report close the gaps it identifies?","acceptedAnswer":{"@type":"Answer","text":"No. A baseline shows share of voice, prompts that mention the client, and prompts that cite competitors, so it answers where you stand. It does not answer who will change anything, and month one measures the gap rather than closing it."}},{"@type":"Question","name":"Do AEO monitoring tools actually improve AI visibility?","acceptedAnswer":{"@type":"Answer","text":"No. Monitoring tools report mentions, citations, and share of voice, but they do not improve visibility, guarantee citations, or verify accuracy. Closing a gap takes better pages, profiles, reviews, and earned coverage, plus someone with access to the CMS, content calendar, and outreach."}},{"@type":"Question","name":"Is buying a bigger monitoring tier a good way to fix a client's AEO results?","acceptedAnswer":{"@type":"Answer","text":"No. More prompts tracked can make the report cover more ground, but the underlying work queue does not move. More measurement is not the fix for a fulfillment gap, and upgrades can also narrow margins when per-domain and per-prompt pricing scales across clients."}},{"@type":"Question","name":"Why do clients cancel AEO retainers after a few months?","acceptedAnswer":{"@type":"Answer","text":"In the pattern described, the client cancels after three reports show the same red cells as month one. Having paid $3,600 with nothing visibly fixed, the client pauses the retainer. The data did its job; the service never delivered the work the data called for."}},{"@type":"Question","name":"What should an agency ask before buying an AEO monitoring tool?","acceptedAnswer":{"@type":"Answer","text":"Ask who ships the fix when the tracker finds a missing citation or a weak entity page, then wait for the answer. If the answer is your own team, you have bought an internal QA tool, not a resellable service, so price and staff that labor before selling the retainer."}},{"@type":"Question","name":"Should agencies sell an AEO retainer built only on a monitoring tool?","acceptedAnswer":{"@type":"Answer","text":"No. If the tool cannot ship the fix, it does not go in the retainer by itself. If the agency cannot price and staff the execution work, the tracker should stay an internal check rather than become the basis of a client-facing retainer."}}]}]}
```