LinkedIn
The typical AEO pricing page sells an agency the cheap part of the job: watching whether clients appear in AI answers. The expensive part—fixing pages, publishing content and earning citations—goes back to the agency. I chose a composite agency pricing page for this teardown because its familiar sequence of prompt caps, domain fees, monitoring tiers and white-label charges makes that split easy to see.
It reminds me of a 2015 PPC retainer. Back then, we metered keywords and reports. Now vendors meter prompts and dashboards. Neither meter tells you who does the work when a client is absent from an answer.
I’ll read the page in the order an agency buyer encounters it, give each line a verdict, and check what happens as a client book grows from six domains to ten to fifteen.
The hero price: a safe-looking entry point
The page opens with a number designed to feel manageable. Lite at $29 a month for 15 prompts here, Starter at $49 for 25 prompts across three engines there. Underneath sits a slider or grid: more prompts, more engines, more domains. Each step up adds $50 or $150, while Unlimited appears only on a plan you cannot price without talking to sales.
I used to argue against percentage-of-spend pricing in PPC because the bill could rise when spend rose, whether or not the work did. This page makes a similar move with different meters. The entry price tells you what it costs to observe a small slice of the market, not what it costs to improve a client’s visibility. Keep that distinction in mind as you scroll.
Line 1: “Up to X tracked prompts”
The first allowance reads like generosity. Up to 100 prompts. Up to 400 prompts. Then you do the agency math. Otterly lists Lite at $29 a month for 15 prompts, Standard at $189 for 100 prompts, and Premium at $489 for 400 prompts, all on four base engines. Another 100 prompts on top of the allowance costs $99 a month. HubSpot’s standalone AEO is $50 a month for 25 prompts, with extra tracking at $20 a month for each 10 prompts. That meter runs on questions, not on answers.
A client’s buyer questions do not conveniently stop at 25. A proper audit alone can run 40 to 80 prompts before ongoing tracking begins; standalone audits are priced at $1,500 to $5,000 for a 40- to 80-prompt universe, versus 25-prompt tool caps. I used to tell clients to expand keyword coverage. I was wrong about how far exact match would carry them, but I was right that you never grew an account by tracking less of the market. A prompt cap can make the next useful question feel like an expense.
Verdict on Line 1: it taxes curiosity. Before you buy the allowance, map the questions you actually need to watch.
Run the domain math at client six, ten and fifteen
Say you have ten local service clients, $20k a month in retainers across the book, and a $300 tracking charge per client. A per-domain meter starts eating that charge before anyone has fixed a page. Semrush AI Visibility Base is $99 a month per domain for 25 daily prompts, so ten clients cost $990 a month before add-ons. Profound’s Agency Growth is $99 base plus $399 per client workspace, or $4,089 a month for ten clients at 100 prompts each.
At client six, ask whether the tracking allowance still fits every account. At client ten, put the total vendor bill next to what you charge for tracking. At client fifteen, ask whether the same terms still apply or whether you have reached custom Enterprise pricing. The problem is not that more clients cost more to serve. It is that watching more prompts does not, by itself, fix a citation.
Line 2: Domains and seats enter the bill
Next, the page starts charging for who and what gets access. Add a teammate in Semrush and that means another license; add domains or prompts and the bill can push toward a few hundred dollars, sometimes past a thousand. Scrunch charges $500 a month for Agency Core with 250 prompts and three brands, so client four puts you on custom pricing. Peec AI uses credits, with one credit equal to one prompt on one model for a day; its agency plans run from $245 for three projects to $795 for 25 projects, and each client needs a minimum of 900 credits.
Read the units slowly. Another client, another engine or another strategist can change the invoice without changing what the client receives. I have seen this play in PPC with extra ad accounts. The extra line never made an ad better; it made the invoice longer.
Verdict on Line 2: you are renting chairs, not buying outcomes. If you want a closer look at per-domain math without those meters, read Flat-Rate AEO Pricing Per Domain.
Line 3: The monitoring tier shows you the loss
Scroll down and the language changes. The page moves from what you track to what you see: share-of-voice charts, sentiment graphs, competitor mentions. That can be useful coverage. It can also become the old monthly PDF report in a better jacket. What the deck calls visibility intelligence, I call watching the scoreboard while nobody is on the field.
The engine add-ons expose another limit. Claude costs $29 to $439 a month by tier, while Gemini and AI Mode cost $9 to $149 extra. Profound starts at $99 a month for ChatGPT-only monitoring; multi-engine monitoring starts at $399 for three engines, with Claude gated to custom Enterprise. Even the view of the scoreboard can depend on another purchase.
After the alert, who makes the fix?
Here is the handoff the pricing page rarely diagrams. An alert says a client went uncited for a money prompt. Someone still has to diagnose why, rewrite the page, fix a technical block where needed, publish in the CMS and pursue the third-party mention. Those are different jobs from recording the alert.
Profound caps content at zero articles on Starter and three a month on $399 Growth, while Peec surfaces opportunities but ships no content writing, only a roadmap. Peec is described as monitoring-first without content-optimization recommendations, and monitoring-only tools leave separate tools and manual work between finding a gap and closing it. A useful warning can still leave the whole response on your team’s desk.
Verdict on Line 3: you bought the smoke detector, not the fire extinguisher. Price the fix, not just the flag.
Line 4: White-label makes the report yours
Near the bottom sits white-label, priced as if putting your logo on the report completes the work. Semrush charges $20 a month extra for a Pro Report with white-label, plus $45 a month for each extra user. Profound’s Growth tier includes no white-label. Otterly handles it through Looker Studio rather than natively in the product. Scrunch limits Agency Core to API-only white-label.
An Agency tier at $399 bundles white-labeling, a pitch workspace and unlimited client sites, but still leaves writing and publishing to you. That distinction matters in a client meeting. The client sees your logo on a report showing where they are losing. Your team then has to explain what happens next and do the work to change it.
Verdict on Line 4: the surcharge buys the appearance of delivery. Check who owns the response before you put your name on the warning.
Line 5: The work that is not on the page
Flip the page over. There is no line for the work that moves a citation: technical fixes, new pages written, CMS publishing or third-party mentions earned. That absence is the tell. For unbranded B2B questions, 89% of AI citations come from third-party sources rather than the client’s site. The cited agency guidance says to budget 40 to 50% of program resources for that work. Another tracked prompt will not put a client on G2 or Capterra, into a Reddit discussion or in earned press.
A tracking product does not have to perform every one of those jobs. It does have to be priced and sold honestly enough that an agency can tell what it is buying. On this composite page, the observation gets a prominent price; the labor required to act on it remains your problem.
Verdict on Line 5: if the page has no line for execution, you are the execution. Put those hours into the comparison before calling a tier cheap.
What holds up: a meter for a metering job
Metered tracking is fair in one case. You run a single site, need 25 prompts checked weekly and already have someone in-house to rewrite the pages. Then HubSpot at $50 a month for 25 prompts is inexpensive coverage. Otterly Lite can make sense as an alarm bell, too.
That is a narrower purchase than an agency growth service. The moment a client pays you to be cited, rather than merely watched, a cap can become a limit on the job you promised to do. I would not buy execution by buying more monitoring and hoping the missing work takes care of itself.
Verdict: buy the meter when you only need the meter. If you resell the result, budget for the hands that change it.
Five questions to put to an AEO vendor
I keep a short script for these calls because pricing pages rarely answer the questions that decide margin. Ask for the answers in writing. The fuller version is in AEO for Agencies: Five Questions to Ask Before You Buy Another Dashboard; this is the version for a demo that starts glossing over the meters.
- What is capped? Prompts, domains, seats, engines, or all four? What triggers the next charge?
- What happens at 2x usage? Do we pay more, track less or move to custom pricing? Prompt Caps Are the Wrong Way to Buy AEO for Clients explains why that answer matters.
- Who acts after an alert? Who writes the fix, publishes it in the CMS and earns the third-party mention?
- What does white-label include? Is it native and included, or a report export plus a fee?
- What is the ten-domain total? Show the wholesale cost for ten clients with full engines, team seats and white-label included.
If the vendor cannot show that last calculation, do not build your resale price around the hero number.
Overall verdict: keep the diagnostic, replace the model
This artefact is a well-built alarm system presented like a growth service. I would keep prompt-level visibility. Knowing which buyer questions surface your client and which surface a competitor is useful diagnostic work; a 25-prompt check has a place, just as a search term report did.
I would not treat the rest of the page as a complete agency offer. Prompt caps can shrink useful coverage as the client book grows. Domain and seat meters raise the cost of resale. Engine add-ons can make full monitoring another purchase, and a white-label fee can leave you paying for the privilege of forwarding bad news. Read line by line, the agency plan charges you to watch clients go uncited, then hands the rewrite, the publishing and the citation work back to your team.
That is why I prefer a per-domain, uncapped, execution-included model to a cheaper-looking monitoring tier. When the wholesale price per client stays flat, the tenth client does not force a new kind of calculation. More important, the work belongs in the offer: content written, technical fixes shipped, citations earned and weekly visibility tracked under your brand, rather than left for your team’s nights and weekends. groas runs that way for agencies: connect clients once, turn on paid or organic per domain, see every action logged with its reasoning, and get branded weekly reports ready to forward. You can try it for seven days on one real client before you pay. If you resell outcomes instead of observations, apply for a free trial and run the ten-domain comparison yourself.