Service
AI SEO services measured in tracked prompts
Twenty, sixty or one hundred and fifty prompts tracked each month from $400/mo, across ChatGPT and Perplexity, with schema and structure work aimed at getting your pages cited rather than just ranked.
In one line
AI search optimization, also called AEO or GEO, is work aimed at getting a site cited inside AI answers rather than only ranked in blue links. Here it is measured by tracked prompts: a set number of real questions checked each month to see whether your pages are cited, and by whom you are being displaced.
AI SEO services and what each tier costs
Starter
20 tracked prompts
$333/mo billed yearly
- AI visibility baseline
- 20 prompts tracked across ChatGPT + Perplexity
- Schema + structure fixes
- Monthly citation report
- Not included: Content writing
- Not included: Guaranteed citations, which nobody can promise
Growth
60 tracked prompts
$750/mo billed yearly
- 60 prompts tracked
- Everything in Starter
- Answer-format content briefs
- Competitor citation gap
- Not included: Guaranteed citations
Scale
150 tracked prompts
$1,667/mo billed yearly
- 150 prompts tracked
- Everything in Growth
- Entity + knowledge-graph work
- Quarterly strategy review
- Not included: Guaranteed citations
These ai seo services are billed at the price shown, with no scoping call in the way. Baseline visibility audit within 10 business days. Every tier is billed monthly, and the scope is fixed by a brief you agree before anything is invoiced.
This search is new enough that page one includes a Reddit thread ranking agencies. Every commercial page promises AI visibility and none of them publish how it is measured, which is the only part a buyer can hold them to.
Everything each tier includes and excludes
The same list that drives the price cards, laid out so the tiers can be read against each other. Exclusions are printed because an unstated exclusion is what turns a good month into an argument.
Starter
20 tracked prompts
Included
- AI visibility baseline
- 20 prompts tracked across ChatGPT + Perplexity
- Schema + structure fixes
- Monthly citation report
Not included
- Content writing
- Guaranteed citations, which nobody can promise
Growth
60 tracked prompts
Included
- 60 prompts tracked
- Everything in Starter
- Answer-format content briefs
- Competitor citation gap
Not included
- Guaranteed citations
Scale
150 tracked prompts
Included
- 150 prompts tracked
- Everything in Growth
- Entity + knowledge-graph work
- Quarterly strategy review
Not included
- Guaranteed citations
Which tier you actually need
Tiers are set by how many prompts are tracked each month. More prompts means more coverage of how buyers actually ask, not more work per prompt.
All three ai seo services run the same process and the same reporting. What changes between them is volume.
Starter
A site with published content that wants to find out, factually, whether it is being cited anywhere.
You are on this tier when: You suspect buyers are asking models about your category and you have no measurement at all.
Growth
A site acting on the findings, where answer-format briefs are needed to close the gaps the baseline exposed.
You are on this tier when: Your baseline showed competitors cited on prompts you should own.
Scale
Broad coverage with entity and knowledge-graph work across the whole site rather than page by page.
You are on this tier when: You are competing on many prompts at once and your entity data is inconsistent across the web.
The unit of measurement is published, and so is the limit
Short answer
A tracked prompt is one real question checked monthly across ChatGPT and Perplexity, recording whether you are cited and which sources are cited instead. Guaranteed citations are excluded at every tier, because no provider controls what a model returns.
Every competitor page on this search sells the outcome. This one sells the measurement, because the outcome is not in anyone’s gift.
Tracking makes the work falsifiable month to month, which is the only honest way to sell into a channel this young.
What a month produces
The deliverable is a repeatable measurement plus the structural work that changes it.
- A citation report: your prompts, whether you were cited, and who was cited instead
- Schema and page-structure fixes applied to the pages being tested
- From Growth upward, answer-format content briefs for the gaps
- From Scale, entity and knowledge-graph work across the whole site
What an AI search month actually contains
AI answer engines cite pages; they do not send a ranking report. What can be counted is the prompt set, the citations found against it, and the pages restructured to be quotable.
What an AI search month actually contains
Within 10 days — Baseline visibility audit. A fixed prompt set is run against ChatGPT and Perplexity and every mention or citation of your domain is recorded with its date and the prompt that produced it. That record is the baseline.
Month 1 — Pages restructured to be quotable. Target pages are rewritten so each section opens with a direct, self-contained answer. Entity coverage and structured data are corrected on the same pass.
Monthly — Prompt set re-run. The same prompt set is re-run on the same schedule and the results are diffed against the baseline. Assistants change their behaviour constantly, so the diff is stated as observed, not as caused.
Month end — The report. Prompts run, citations observed, pages restructured, and schema changes shipped. Where a model stopped citing anyone at all, that is written down too.
Select any milestone to read what is produced at it. Every date is a published commitment from this page, not an estimate of results.
What the price buys at each tier
Billed monthly at the price shown. The yearly option is the same scope with two months off, and nothing here is quoted as a slice of your budget.
Prompt coverage is the lever, and it is the only number here that behaves like a keyword set. Nobody, us included, can promise a citation: what is being bought is the monitoring and the restructuring, stated as such.
Prices resolve from the same source as the cards above. Switch metric to compare cost against volume.
How a month actually runs
Order, brief, production, delivery, report. The same four steps every month, so a second month can be compared against a first.
- 1
Days 1 to 10
Baseline
A visibility baseline across your tracked prompts, so month two has something to be compared against.
- 2
Weeks 2 to 4
Structural fixes
Schema, headings, answer formatting and internal structure on the pages that should be the cited source.
- 3
Ongoing
Re-measurement
The same prompts re-run monthly. Changing the prompt set mid-engagement would destroy the comparison, so changes are agreed deliberately.
- 4
Month end
Citation report
Movement, displacement, and which competitor is being cited where you are not.
What the citation report contains
The report is a measurement first and a narrative second, because this channel is young enough that the measurement is the product.
Changing the prompt set breaks the comparison, so changes are agreed deliberately rather than drifting month to month.
- Every tracked prompt, marked cited, mentioned or absent
- The sources cited instead of you, named
- Movement since last month on the same prompt set
- The structural changes made, and which pages they were made on
How AEO and GEO relate to ordinary SEO
Short answer
Answer engine optimization and generative engine optimization are layers on top of conventional SEO, not replacements for it. The pages models cite are usually pages that were already crawlable, structured and specific enough to answer a question directly.
The practical overlap is large: schema, clean headings, direct answers near the top of a page, and factual specificity all serve both.
Where they diverge is measurement. Rankings are positional, citations are binary, and the reporting has to reflect that difference rather than dressing one up as the other.
What a tracked prompt looks like
- A real question a buyer would ask, not your brand name
- Checked in ChatGPT and Perplexity on the same day each month
- Recorded as cited, mentioned, or absent
- With the competing sources that were cited instead
Brand-name prompts are excluded from the tracked set by default. Being cited when someone already searched for you proves very little.
How this differs from the usual arrangement
Neither column is a verdict on quality. The left is how this service is normally sold, taken from the pages ranking beside this one.
A typical AI SEO pitch
- Price
- On enquiry
- Measurement
- Unstated
- Platforms
- "AI search"
- Guarantees
- Implied visibility
- Overlap with SEO
- Sold as a replacement
The Super Panel
- Price
- Published, by prompt count
- Measurement
- 20, 60 or 150 tracked prompts
- Platforms
- ChatGPT and Perplexity, named
- Guarantees
- Excluded in writing
- Overlap with SEO
- Sold as a layer on top, and stated
The three things buyers ask before ordering
This is the newest service on the site, and the honest answers are less exciting than the marketing around this category.
Is AI search optimization real, or a rebranded SEO upsell?
The underlying work overlaps heavily with good SEO: schema, structure, direct answers, factual specificity. What is genuinely different is the measurement, because a citation is binary where a ranking is positional. We sell the measurement and the structural work, and we do not pretend it is a separate discipline.
Can you make ChatGPT recommend us?
No. Nobody controls model output, and guaranteed citations are excluded in writing at every tier. What can be changed is whether your pages are findable, parseable and specific enough to be worth citing.
Should we buy this instead of SEO?
No. Models cite pages they can crawl and parse, so conventional SEO hygiene is a prerequisite. If your pages are not indexed, this service is premature and we will say so before taking the order.
What this pairs with
The SEO package handles indexing, crawlability and the technical base this work assumes. Buying AI search optimization for a site that is not indexed properly is spending on the roof before the walls.
Content writing turns the answer-format briefs from Growth and Scale into published pages. The briefs specify what to write; they are not the article.
The mechanics behind this service are set out in how citation actually works.
When this is too early to buy
Short answer
AI search work is premature on a site with no content worth citing, no schema, or unresolved indexing problems. Models cite pages they can find and parse, so conventional SEO hygiene is a prerequisite rather than an alternative.
- Your pages are not indexed. Fix that first; nothing here works around it.
- You have little published content. There is nothing for a model to cite yet.
- You want guaranteed citations. Excluded at every tier, by every honest provider.
- You need the content written too. Content writing is excluded at Starter.
Questions before you order
How much do AI SEO services cost?
Starter is $400/mo for 20 tracked prompts, Growth is $900/mo for 60, and Scale is $2,000/mo for 150. Yearly billing charges ten months instead of twelve.
What is a tracked prompt?
One real question checked monthly across ChatGPT and Perplexity, recorded as cited, mentioned or absent, along with which sources were cited instead of you.
Can you guarantee my site gets cited?
No, and it is excluded in writing at every tier. Nobody controls what a model returns, and a guarantee here would be the clearest possible sign of a provider to avoid.
Is this different from normal SEO?
It is a layer on top. Much of the underlying work overlaps, and the difference is what gets measured: citations rather than positions.
Which AI platforms do you track?
ChatGPT and Perplexity at every tier. AI Overviews are influenced by the same structural work, and they are volatile enough that we do not sell them as a tracked metric.
Do I need content written for this?
Content writing is excluded at Starter. Growth and Scale include answer-format briefs specifying what to write, and finished articles come from the content service.
How long before citations change?
Structural fixes can show up within a month or two on prompts where you were already close. On competitive prompts it is slower, and some prompts never move because the model favours a source type you are not.
Can I change my tracked prompts?
Yes, deliberately rather than casually. Changing the set breaks the month-over-month comparison, which is the main thing you are buying.
What is AEO and GEO?
Answer engine optimization and generative engine optimization: two names for getting cited inside AI answers. The labels are used loosely across the industry and both describe the work on this page.
Is it worth buying yet?
It depends on whether your buyers are asking models instead of searching. If you have no content and no indexing, this is premature and we will say so before taking an order.
Sources cited on this page
- Google’s guidance on AI features in Search – Cited for what influences AI surfaces.
- Schema.org – Cited for the structured-data work.
Start AI Search Optimization from $400/mo
Pick a tier and send the order. You get a brief form and a start date within one business day, and nothing is charged until that brief is agreed.
Entry point $300/mo. Cancel any time. No long contracts.
Still deciding?
Read how the process runs end to end: order, brief, production, delivery, report.
How it works