LinkedIn
A $99-per-domain AEO tool looks easy to fit inside a $1,500 client retainer. Then you add the prompts the client wants, the seats your team needs and the work required to improve anything. I built a composite pricing page from the eleven vendor pages agency owners keep sending me because no single page shows that whole bill in one place.
The artefact has the usual order: headline price, tier grid, prompt and domain limits, seats, then the add-ons drawer. I am going to read it in that order and give each line a verdict before moving on. My argument is simple: caps turn a per-domain price into a tax on agency growth. Most of these pages sell a dashboard first. The work an agency needs to fulfill a retainer comes later, if it comes at all.
The artefact: a familiar price with unfamiliar limits
At the top of my composite sits $99 per domain per month, billed annually, with a row of engine logos underneath. Next come three tiers named Starter, Growth and Scale. Their grids count tracked prompts, projects or domains, AI engines, seats and reports. Footnotes add overage credits, extra workspaces, white-label fees and content. A comparison of agency pricing across ten-client setups shows how differently that structure can price the same workload: $399, $489, $795 or $4,089.
I read pricing pages the way I used to read insertion orders: in the order a buyer meets the terms. The headline establishes what feels affordable. The grid narrows what that price buys. The add-ons determine whether an agency can resell it. Do not stop reading when you find a number that fits the retainer. That is usually where the useful reading starts.
Line 1: $99 per domain buys entry, not coverage
The first line is clean: $99 per domain per month, billed annually. One domain, one price. I can put that in a proposal without opening a spreadsheet.
But the Semrush toolkit at that $99 level tracks 25 prompts daily on a single domain; 50 daily prompts costs $199 per month. The Profound Starter tier at $99, as documented in a 2026 pricing comparison, covers ChatGPT only and includes no articles. The headline may be accurate. It still does not describe the coverage or output an agency expects to sell.
Annual billing matters, too: it is the condition attached to that $99 headline, not a detail to discover after you have quoted the client. Verdict: use the entry price as a starting bid for one small client, not your cost basis for a standard retainer. If that client needs 40 prompts across four engines, price that workload instead.
Line 2: prompt caps charge for broader coverage
The next grid lists 25 prompts daily, 50 prompts, five custom prompts, one project. Notice how the unit changes. It makes quick comparisons harder than the tidy columns suggest.
I used to tell clients keyword counts did not matter much. I was wrong to carry that instinct into AEO. A tracked prompt is inventory: each buyer question, checked on each engine at the required frequency, uses some portion of a plan. When a client asks you to cover more questions, the meter moves.
HubSpot’s standalone tier at $50 per month covers 25 prompts daily across three engines and up to 2,500 answers per month; expansion comes in packs of 10 prompts. Peec Starter at $95 per month covers 50 prompts across three models for one project, with a choice of three out of seven engines. Ahrefs Lite at $129 per month includes five daily custom prompts, with each check using a credit per prompt, platform and update.
Those are different allowances, not three interchangeable versions of “50 prompts.” Verdict: ask what one check consumes before you put the tool in a fixed-price retainer. A client who likes the first report and asks to track 60 questions instead of 20 may increase your tool bill without increasing your fee.
Line 3: seats and white-label fees tax the handoff
After the tracking grid comes the part an agency cannot ignore: who gets to see the work, and under whose name?
Extra Semrush users start at $45 per month. A report that hides the vendor’s logo costs $20 per month per report on Semrush Pro. SE Ranking’s Agency Pack add-on costs $69 per month on annual billing and includes an own-domain platform and 30 client seats. Each term answers a different question, so “white-label included” is not enough detail for a proposal.
Nor does a row of engine logos settle what coverage costs. Otterly white-labels through Looker Studio only, while some engine coverage is sold as tier-dependent add-ons: Claude at $29, $109 or $439, and Gemini and AI Mode at $9, $59 or $149. You can see the engines near the top of a page and find their prices much farther down.
The mechanism is plain. Your strategist, account manager and client all need access to the result; seat and report charges turn that ordinary handoff into another cost. Verdict: if the plan does not let you invite the people who need the report and present it under your brand at a price you have modeled, you have not finished pricing white-label. You have priced the logo.
Line 4: credits separate finding work from doing it
Now open the add-ons drawer. This is where a page that looked like a delivery platform can become a monitoring bill plus another bill for action.
Peec’s agency tiers show the tracking meter clearly: Essential costs $245 per month for 10,000 credits, about 111 prompts across three projects; Growth costs $495 for 25,000 credits, about 277 prompts across 10 projects; Scale costs $795 for 65,000 credits, about 722 prompts across 25 projects. One credit is defined as one prompt times one model times one day, with a 900-credit minimum per project. Add a client question or another model, and daily use rises.
Content sits on a separate meter. The comparison that documents Profound’s entry tier also says Profound Growth at $399 per month covers three engines and includes three articles monthly. Peec surfaces opportunities without writing content, while AirOps multi-engine work requires roughly $2,000 per month on Pro. Those are materially different answers to the agency’s question: who makes the fix?
I price that gap in staff hours, not in a prettier credit unit. If a dashboard finds 40 gaps but does not produce the pages or other fixes, someone still has to brief, write, review and publish the work. Verdict: credits can buy useful monitoring, but do not call them fulfillment. If the fix is outside the tier, put the person who will do it in your retainer math.
Line 5: a higher tier may still be a monitoring tier
At the bottom, the tier names start to do more work than the features. Starter sounds like a small version of the complete service. Growth sounds like the service an agency can sell. Scale sounds like more of it. Read the included output instead.
On the composite, the entry tier watches one engine and ships nothing. Moving up expands engine coverage but may provide only a small amount of content. What the deck calls a growth plan, I may still have to treat as a monitoring plan with a larger allowance. Monitoring consumes checks; producing content and other fixes takes additional work. The pricing page splits those jobs, even when the tier name makes them sound joined.
An agency retainer promises more than a weekly PDF showing where the client did not appear. Somebody has to write the page, address the technical block or work on the citation. Verdict: if the tier that fits inside your retainer does not include that work, the tool has given you research for a job you still have to staff.
The retainer math: five, ten and fifteen domains
Say the AEO add-on sells for $1,500 per client per month, within the documented $1,300-to-$2,000 monthly agency band. The useful question is not whether the first domain is affordable. It is what happens as you add clients and expand their coverage.
- At five domains, a $99-per-domain headline implies $495 monthly if that price applies to every domain. That is only a baseline: it does not turn 25 prompts on one engine into 40 prompts across four, or include separate seats, reports and fixes. Ask for the five-client price at the coverage you actually intend to sell.
- At ten domains, there is a direct comparison for 40 prompts tracked across four engines: $399 on Geoptie Enterprise, $489 on Otterly Premium, $795 on Peec Scale, $990 on Semrush and $4,089 on Profound Agency Growth once workspaces are added, as priced for that same workload. Those are tool costs before you staff any work the tier does not perform. The $795 and $990 options alone take a substantial share of a $1,500 retainer.
- At fifteen domains, do not multiply or halve a ten-client plan and assume the tier, workspace and overage rules stay put. Get a quote for fifteen. The margin risk is clear even without that shortcut: a $1,200 tool cost against a $1,300 retainer leaves $100 before you answer a client email.
That is the percentage-of-growth tax I care about. New clients add domains and workspaces. Better reports prompt requests for more questions and engines. Your retainer does not automatically rise with each meter. A white-label resale model with a 40-to-60-percent markup leaves a 29-to-38-percent gross margin before account-management time; expansions come out of that margin first. Model fifteen domains on day one, even if you only have five now.
What I would keep from the page
Not every line is a trap. Daily prompt tracking across ChatGPT, Gemini, Perplexity and AI Overviews is useful because you cannot respond to a change you never see. Weekly share-of-voice and citation counts give a client something concrete to review when rankings barely move. A clean projects view that keeps 15 domains and their prompts separate saves account-management time.
I would pay for those three things. They let a dashboard do its proper job: show the change, per client, without manual pulls.
What I would resist is a new charge every time the monitoring becomes more useful. Another 20 buyer questions should not come as a surprise to an agency selling AEO. Neither should a strategist and account manager both needing to see the same report. Keep the tracking layer. Price the meters around it before you commit to a fixed client fee.
Five questions before you put it in a retainer
I use five questions before adding another dashboard to client delivery. They build on the agency-focused buying questions here, but follow the lines of this page:
- What counts as one tracked prompt, and what happens on prompt 61? Get the unit and overage price in writing. Is it one prompt times one model times one day, or something else?
- How many domains, projects and workspaces are included? Ask what domain 11 costs, not just what the first domain costs.
- What does white-label include? Check the domain, report branding and client seats separately. A logo-free PDF is not an own-domain platform.
- What work ships in my tier each month? Count pages, technical fixes and citations, not opportunities flagged.
- What is the price at 15 domains with 60 prompts each across four engines? If the answer requires a custom quote, get it before you set your retainer price.
Do not treat a vague answer as room in your margin. The page is easy to judge during a five-domain trial. Your business has to survive the fifteenth domain.
Where an execution-included model fits
The alternative I would take from this teardown is one flat price per client domain, no prompt meter and the fix included. That is the point of comparison in the flat-rate per-domain pricing model: adding 20 prompts should not, by itself, rewrite your delivery margin.
groas for agencies uses that structure on the delivery side. SEO and AI search are priced per client domain with content, technical fixes and citations included; Google and AI visibility are tracked weekly; reports go out under your brand. Paid ads sits alongside it as unlimited clients for one flat fee, with seven days on one real client before you pay and month-to-month terms after. The distinction is the work between reports: tracking shows what moved, while execution produces the page and the fix. An agency can put that work in a retainer instead of selling a login and staffing around it.
Overall verdict: keep the visibility, reject the growth tax
The composite page looked cheap at $99. In the order an agency reads it, the prompt cap narrowed coverage, seats and branding changed the resale cost, and credits exposed the gap between finding work and doing it. The full ten-client retainer cost model is worth running before any of those terms reach a proposal.
Buy the work that goes inside the retainer, not just the dashboard that describes it. I would keep daily tracking, citation counts and clean per-client separation. I would not sign a price I had only tested at five domains. Ask the five questions, price fifteen, and keep the parts that still earn their place when the agency grows.