AEO pricing in the US currently runs roughly $2,500 to $15,000 per month for agency retainers, and a large share of that scope is SEO work with a new label. The genuinely new deliverables are narrow: entity consolidation, third-party mention footprint building, and citation share measurement, and they are worth paying for. Everything else on a typical AEO proposal is technical SEO, content, and digital PR you were probably already buying.
I sell this work. That is exactly why this teardown is written the way it is: if you can't tell which line items are new, you can't tell whether anyone, including me, is charging you twice for the same thing.
Key Takeaways
- Published AEO retainers cluster around $2,500-$15,000/month in the US; the spread reflects packaging, not capability.
- Price AEO by deliverable, entity coverage, third-party mention footprint, citation share, not by hours or by "AI strategy."
- Roughly 60-70% of a typical AEO scope is renamed SEO. That's not fraud, but you should not pay a premium for it.
- The three genuinely new work items: entity disambiguation across the open web, systematic third-party placement, and prompt-set citation measurement.
- Nobody can guarantee placement in an AI answer. Any proposal that does is disqualifying.
- Roughly 85% of LLM brand mentions come from third-party pages (AirOps/Kevin Indig): so the budget should skew off-domain, and most proposals don't.
Read the proposal by deliverable, not by package name. The package name is marketing; the deliverable list is the product.
Why AEO pricing is so incoherent right now
Three things are happening at once. AI Overviews now trigger on roughly a quarter of US Google searches, up sharply from about half that a year earlier per Conductor. Zero-click behaviour is dramatically higher in AI Mode than standard SERPs per Semrush. And 6sense reports the large majority of B2B buyers used a generative AI tool in their most recent purchase.
Real change, real urgency, and no established pricing norms. That combination produces exactly what you'd expect: every SEO agency in the US rewrote its homepage in about six months, and the SERP for "AEO pricing" is now essentially 100% agency pages quoting their own rate cards.
The SERP problem
If you search this term, you will find agency pages. Each quotes a range that happens to bracket its own pricing. None of them is neutral, and neither, strictly, am I, the difference is that I'm telling you what's inside the box.
What's genuinely new vs repackaged SEO
Here is the honest split.
| Deliverable | Genuinely new? | What it actually is | Fair standalone price (US market) |
|---|---|---|---|
| Technical crawl and index health | No | Core SEO since 2010 | $1,500-4,000 one-time audit |
| Schema markup (Organization, Product, FAQ) | No | Structured data, long-standing practice | $1,000-3,000 setup |
| Content briefs and production | No | Content marketing | $0.30-1.00+/word (US writers) |
| Site speed / Core Web Vitals | No | Technical SEO | $2,000-6,000 project |
| Entity consolidation and disambiguation | Partly | Making your brand legible as one unambiguous entity across the open web | $3,000-8,000 project |
| Third-party mention footprint building | Partly | Digital PR + review platforms + community, aimed at citation corpora | $3,000-12,000/mo ongoing |
| Citation share measurement | Yes | Prompt-set testing across assistants, tracked over time | $500-2,500/mo |
| Answer-shaped content restructuring | Partly | Direct-answer formatting, extractable definitions, disqualifying language | $150-400 per page retrofit |
| "AI strategy workshop" | No | A meeting | Should be free in a sales process |
| "LLM ranking guarantee" | Not a thing | Unsellable | $0, walk away |
The pattern: the new work is concentrated off your domain and in measurement. Most proposals concentrate spend on your domain and in production.
Pricing by deliverable, not by hour
Hourly pricing for AEO is a trap for both sides. The work is lumpy: a month of entity cleanup, then three months of slow third-party accumulation. Hours reward the wrong thing.
Deliverable 1: Entity coverage
What it is: one canonical description of your organisation, consistently expressed across your site, schema, LinkedIn, Crunchbase, Wikipedia (if eligible), review platforms, and directories. Consistent naming, consistent category, consistent geography, consistent founding date.
How to price it: as a fixed-scope project, typically $3,000-8,000, with a defined checklist of properties updated. Ask for the checklist before signing.
How to verify delivery: run your brand name through three assistants and check whether the description returned matches your canonical one. Before and after.
Deliverable 2: Third-party mention footprint
What it is: the count and quality of distinct external domains that mention your brand in a buying context. This is where the AirOps/Kevin Indig 85% figure bites: if most LLM brand mentions come from third-party pages, and your budget is 90% on-domain, your budget is misallocated.
How to price it: monthly retainer tied to a target number of qualified new mentioning domains, with a definition of "qualified" agreed up front. $3,000-12,000/month depending on target volume and vertical difficulty.
Warning: this is where paid link vendors will try to substitute. A paid placement on a low-quality site inflates the count and does nothing for citation share. Insist on a domain quality floor and disclosure of any paid placements.
Deliverable 3: Citation share
What it is: you define 25-50 buyer prompts. Each month, they're run across two or three assistants and logged: mentioned or not, sentiment, sources cited. Citation share is the percentage of the prompt set where you appear.
How to price it: $500-2,500/month. It is largely tooling plus a few hours of analysis. If someone quotes $6,000/month for measurement alone, you're paying for their software margin.
Why it matters most: it is the only deliverable that tells you whether the other two worked.
Citation share is the honest scoreboard. Without it, an AEO retainer is unfalsifiable, which is convenient for the seller.
What a fair proposal looks like at three budget levels
Under $3,000/month
Realistically: measurement plus one focused workstream. Pick entity consolidation first (it's finite), then move budget to third-party work. Do not buy "full AEO" at this level; you'll get a thin version of six things instead of a complete version of one.
$3,000-8,000/month
Entity work front-loaded in months one and two, third-party footprint building from month two onward, citation measurement throughout, and a modest content retrofit programme. This is the band where AEO retainers make sense for most US mid-market B2B and edtech brands.
$8,000-20,000+/month
Adds original research production, a real digital PR function, and multi-market prompt sets. At this level you should expect named senior people on the account and monthly citation share reporting with competitor benchmarks.
Red flags in AEO proposals
- Guaranteed AI Overview or ChatGPT placement. Not deliverable. Retrieval is not a ranking you can buy.
- "llms.txt implementation" as a headline deliverable. It's a file. It takes an hour. It is not a strategy and adoption remains inconsistent.
- No measurement line item. If the proposal can't be falsified, it can't be evaluated.
- Budget skewed 80%+ on-domain. Contradicts where LLM mentions actually come from.
- Per-link pricing hidden inside "authority building." US in-house teams increasingly treat paid link buying as a liability. Ask directly whether placements are paid.
- Deliverables denominated in hours. You'll get hours.
And a red flag about people like me
Offshore strategists, I'm in India, working with US brands, have a specific temptation here: sell AEO at a US-adjacent price on an offshore cost base, and fill the scope with the cheap half (on-domain content) rather than the expensive half (third-party placement, which requires real relationships and often a US partner). Ask any offshore vendor, me included, exactly how the third-party mentions will be earned and by whom. If the answer is vague, the answer is guest posts.
How to structure the contract
- 90-day initial term, then monthly. AEO results take two to three quarters; a 12-month lock-in before you've seen citation share move is unnecessary risk.
- Baseline measurement in week one, before any work ships, agreed in writing.
- Named team members with a clause that changes require your approval.
- A disclosed policy on paid placements, and an AI-content disclosure policy for anything published under your name.
- Deliverable-based invoicing for the entity project; retainer for the ongoing footprint work.
FAQ
What is a normal AEO pricing range in the US? Published agency ranges cluster around $2,500-$15,000 per month, with project-based entity work at $3,000-8,000. Treat those as published ranges from vendors with an incentive in the number, not neutral benchmarks.
Is AEO just SEO with a new name? Mostly, but not entirely. The overlap is roughly 60-70%. The genuinely distinct work is entity disambiguation across the open web, systematic third-party mention building, and citation share measurement.
Can anyone guarantee my brand appears in ChatGPT answers? No. You can influence the corpus models retrieve from; you cannot control the generated answer. Guarantees are a disqualifying signal.
Should I pay separately for AEO if I already have an SEO retainer? Usually not. Ask your current agency to reallocate, measurement plus third-party emphasis, before adding a second vendor. Two vendors on overlapping scope is the most common way US teams overpay here.
How much of my budget should go off-domain? Given that roughly 85% of LLM brand mentions come from third-party pages per AirOps/Kevin Indig, a majority of incremental AEO budget should go off-domain. Most proposals I see are the inverse.
What does citation share measurement actually cost to run? $500-2,500/month is fair. It is tooling plus a few analyst hours. Anything much higher is software margin.
How long until AEO work shows results? Entity fixes can move brand-level answers in six to ten weeks. Third-party footprint work is a two-to-three-quarter horizon. Content retrofits are somewhere in between.
Is llms.txt worth implementing? Implement it, it costs almost nothing. Do not pay a premium for it or accept it as a headline deliverable. Adoption across major assistants remains inconsistent.
How do I compare two AEO proposals fairly? Rewrite both into the deliverable table above, ignoring the package names. Then compare on entity scope, target mentioning domains per month, and measurement cadence. Differences usually collapse to those three.
What if I only have $1,500/month? Buy measurement plus entity consolidation as a one-time project, run it yourself for two quarters, and revisit. That is a genuinely reasonable plan and most agencies won't tell you so.
If you want your AEO proposal read line by line by someone who sells this work and will still tell you which parts you don't need, send it over. I work with US startup and edtech teams from India, and I'd rather scope you honestly than sell you a package. More at younusfardeen.com.