AEO Retainers: Resell the Dashboard or White-Label the Work?

An agency owner pitched AEO for $750 a month. Behind the offer sat a $99 tracker and a PDF. The client signed, then cancelled three months later because nothing on their site had changed. I sat in on that sales call last year, and it left me with the question agencies now have to answer: should you resell AI-visibility tracking or white-label the work that changes what it tracks?

Smart operators disagree. The dashboard model offers a fast launch, control of the client relationship and an attractive markup. The execution model costs more and puts another team under your brand, but it gives the client something to renew for: pages published, technical fixes shipped and citations gained. Both can look profitable in a spreadsheet. The argument is about which one still looks profitable at the second quarterly review.

Position A: Resell the dashboard and keep control

The dashboard case starts with math that looks hard to beat. Agencies typically resell basic AEO monitoring for $500–800 a month, while the tool behind it costs $79–250 a month. Buy a seat, connect clients, mark up the service and keep strategy in house. You do not need to build a content operation before you learn whether anyone wants the offer. I used to tell clients this was the smart way to productize SEO. From a sales and launch perspective, I can still see the appeal.

The agency remains the strategist. You choose which prompts to track, decide what the results mean and own the client conversation. White-label can mean branded PDFs and free read-only logins on Geoptie, or a full platform on your own domain with SE Ranking. You can launch without inviting a fulfillment partner into Slack or giving anyone else access to the client site. For an owner who has already built a capable delivery team, that independence matters. The dashboard supplies a map; your people decide where to go.

There is a risk argument too. AEO software priced from $29 to $699 or more a month shows where a brand stands but does not rewrite content, build schema or land PR. If you sell it honestly as monitoring, that limited scope is a feature. No outside partner changes a page, introduces a technical problem or makes a citation you would rather not explain. You can show a client their visibility in ChatGPT, Perplexity and AI Overviews, then let their own team act on it. Not every buyer wants an agency touching the site. Some want a clear list of what to fix and the freedom to assign it themselves.

The strongest version of Position A, then, is not “charge for a login and hope nobody notices.” It is sell a bounded intelligence service to a client with the capacity to execute. Define the prompts, explain the gaps, agree on who owns each fix and keep the monitoring contract separate from implementation. That is a defensible offer. It is also a narrower offer than a lot of agencies want to sell.

Even its tidy margin needs a closer look. Agency AEO cost scales on prompts times engines, not just on the number of clients. For 10 clients tracking 40 prompts, listed monthly costs include $399 on Geoptie Enterprise, $489 on Otterly Premium, $795 on Peec Agency Scale, $990 on Semrush with a 25-prompt-per-client cap, and $4,089 for Profound per-client workspaces. At 10 prompts per client, the markup can look generous. At 40 prompts across eight engines with history turned on, that margin thins. Cap prompts in the contract, or the tracker reprices your retainer for you.

Position B: White-label execution and sell movement

The opposing case begins where the dashboard pitch usually ends: the client has seen the chart. Now what? Clients fire SEO firms when results cannot be tied to revenue, even if the reports are full of rankings, traffic and authority. A tracker can tell a plumber the brand appeared in four of 40 prompts. It cannot, by itself, make the service page work as a standalone answer, fix missing schema or secure a third-party citation. Those are delivery jobs. That is why programs covering content, technical fixes and schema sell at $1,500–3,000 a month, while basic tracking sits at $500–800. One offer identifies the gap. The other is accountable for working on it.

The work is unglamorous. Good. A monthly AEO report should show citation frequency, competitor share of voice, an engine breakdown, cited pages, month-over-month trends tied to changes and a three-to-five-item fix list. The useful question is who acts on that last list. A dashboard can populate much of the report; it cannot rewrite thin content or repair the site. Agency retainers from $1,500 to $50,000-plus cover structural work. Under a white-label execution model, a partner ships that work under your brand: pages published, technical tickets closed, citations placed. The client can inspect changes on the site instead of admiring a new login.

That distinction also changes the sales conversation. When a tracked prompt stays red, a monitoring-only agency has to explain the observation and then scope somebody else to solve it. An execution-led agency can point to what it changed, what the next report shows and what it will work on next. The chart still matters. It tells you where the work should go and whether the work appears to be moving anything. Measurement is part of delivery, not a substitute for it.

Position B has a unit-economics case, although it is less pretty as a percentage. Per-prompt tracking costs can climb as you add clients and coverage. Fulfillment priced per domain is easier to plan around: sell a $2,000 retainer, pay a fixed partner cost per site and keep the spread. Agencies already pay close attention to reporting costs: AgencyAnalytics charges around $20 per client a month, Swydo has a $62 base plus $2–4.50 per source, and Whatagraph starts from an $812 floor. Once you manage 10-plus domains, small changes to the unit cost stop being small.

Take 12 clients at $1,800 each: $21,600 in monthly retainer revenue. Suppose tracking costs $795 on a mid-tier agency plan. That leaves plenty of apparent margin, but your team still owns 12 sets of fixes. Suppose instead that execution costs $800 per domain, all in. The delivery cost is $9,600, leaving $12,000 before your other costs, while work ships each month. The first model has the better-looking tool bill. The second has a defined answer to “Who is doing this?” Do not confuse the two when you price a retainer.

There is a staffing argument as well. One white-label SEO account describes two account managers servicing 20-plus retainers through a partner, versus six to eight hires doing the work in house, with typical resale margins of 30–50%. I know the pressure the in-house version creates. I once owned 14 Google Ads accounts solo; the fourth week of every month turned into copy-paste reports until midnight. What the deck calls “capacity planning,” I call deciding which client gets ignored. A partner can turn some of that fixed staffing pressure into a variable delivery cost.

It can also create a mess if you pick badly. Hidden onboarding fees, dashboard seats, revision charges and link shortcuts that trigger manual actions all belong in the cost calculation. A cheap partner whose citations you later have to disavow is not cheap. Vet the work and the scope, not just the resale margin.

Where the arguments break under pressure

Position A has one question it cannot dodge: who does the fixes the dashboard surfaces? The monthly report calls for citation and engine data, trends tied to changes, and a fix list. Without a writer, a technical owner and time to audit the findings, that last section becomes a growing inventory of work nobody bought. I have sent reports like that. The client gets to page 12, asks what we are doing about the red prompts, and the choices are ugly: scope work we did not staff for, or explain why red is actually fine.

Reporting can help retain a client. In one survey, 45% of clients said proof of results increases trust. But proof that you watched a problem is not proof that you addressed it. And when 66% of agencies name AI search as a top new request while 48% cannot reliably track it, it is easy to mistake a monitoring subscription for a complete service. The client may not make that distinction on the sales call. They will when the second report resembles the first.

Position B has a different weak spot: you own work you did not personally do. A partner operates under your name, on client domains and sometimes inside client accounts. If onboarding drags, revisions cost extra or citations come from places you would not show a client, the apology is yours. That is why I would read the scope and ask who does the work in a white-label AEO partnership before getting excited about pricing. Which pages will the partner change? Who approves them? What happens when a fix needs revising? Those are not procurement details. They determine whether “execution” describes the service or merely decorates the pitch.

The risks do not cancel each other out. Position A can leave you collecting fees while an untouched fix list grows. Position B can leave you accountable for sloppy delegation. I would rather manage the second risk by choosing and overseeing the partner than build a retainer around work nobody is assigned to ship.

The verdict: Choose the model that survives renewal

I take Position B for agencies whose clients need the work done, not just the gaps identified. The deciding consideration is lifetime margin after churn, not the percentage on the first invoice.

Here is the deliberately simple math. Sell tracking at $650 a month with $100 in tool cost, and you keep $550 a month before other expenses. If the client cancels after three months because nothing changed, that is about $1,650 kept. Sell execution at $2,000 a month with $800 in partner cost, and you keep $1,200 a month before other expenses. If the client stays for twelve months while pages ship, technical tickets close and citations move, that is $14,400. Those are scenarios, not a promise that every execution client stays a year or every dashboard client leaves in month three. They show what the higher percentage margin leaves out: renewal.

Tracking without fixing is the 2026 version of the monthly PDF report. The agency can make it handsome, brand it and deliver it on time. None of that changes the site.

When I would still sell the dashboard

There are three cases where I would buy the tracker and mark it up:

1. The client has writers and developers in house. They need a prompt list and gap analysis, not another team making changes.
2. The retainer is under $800. It cannot carry content and citation work, so sell monitoring with a tight fix list and be clear about who will act on it.
3. The agency is testing demand. Two or three months of tracked prompts can help you sell a larger program before you commit to a fulfillment partner.

That shift from monitoring to execution-based search retainers does not make the tracker useless. It makes the scope of the tracker important. Outside those three cases, selling visibility alone leaves your team writing checks the software cannot cash.

The vendor test: What changes in the first 30 days?

Before you buy either model, ask for a dated answer to one question: what changes on the client site in the first 30 days? A dashboard vendor can tell you about logins, prompt counts and refresh cadence. An execution partner should be able to name the pages, technical tickets, schema and citation work it will take on, and show you how that work will be reported under your brand. If your team still has to make every fix after you pay for the platform, you bought homework, not fulfillment. That is the distinction behind the warning that a white-label program leaving all the fixes to your team is just a dashboard.

That test is why I white-label execution through groas for agencies instead of reselling a tracker. Its model is per client domain, so adding sites does not mean repricing every prompt. It covers content, technical fixes and citations, tracks Google and AI visibility weekly, and reports under the agency’s brand. Execution runs around the clock with actions and reasoning logged. Your team stays client-facing, and you can trial it on one real client before paying, then continue month to month without a lock-in. For a shop running five to 15 domains, that is a more defensible retainer: clients pay for citations gained and work shipped, not charts viewed.

Buy the tracker if the client already has builders and just needs the map. For everyone else, sell movement. Pick the model that leaves the site measurably different after month one, price the work at $1,500-plus, and let someone else chase the $650 accounts that may be gone by quarter two. I learned this selling PPC reports that looked great and changed nothing. I do not sell that anymore.