Start with something uncomfortable about this question: almost every page currently answering it was published by a company selling one of the two answers.
We looked at what ranks for this query. Agency sites conclude you need an agency. Tool sites conclude you need a tool. The headline statistics on those pages are frequently unsourced, and at least one widely repeated figure describes a Google SERP feature rather than an AI answer at all.
Pepper sells a platform with a growth team attached, so we are not neutral either. The difference we can offer is that every claim below is traceable, and we will name the claims on the other pages that are not.
The short answer
If you have people with hours to act on findings, buy a tool. If your constraint is hands rather than knowledge, buy the service. If you are not yet sure AI answers matter in your category, use a free tier for a quarter and decide with data.
And before any of that: AEO and SEO are not separate disciplines. Google says so directly, which changes what you should be willing to pay a premium for.
Key takeaways
- AEO, GEO and AI search optimisation are the same work under three names. Vocabulary differences are marketing, not method.
- Google states AEO and GEO are part of SEO. AI Overviews and AI Mode run on core Search ranking systems with no separate AI index.
- The entry price is now zero. AirOps offers a free tier tracking 100 prompts, which makes “try before you buy” the default sensible move.
- Watch for the 72% claim. A widely quoted figure about featured snippets in 90 days circulates with no study behind it, and it describes a Google SERP feature rather than an AI answer.
- Earned media is the deciding capability. Muck Rack found 84% of AI citations come from earned media, which no software tier provides at any price.
A note on where this comes from. We run organic for more than 250 enterprises and track over 10 million prompts across every major engine. What follows is shaped by that, by the client reviews we sit in weekly, and by conversations with buyers at the events we run. It is our read on a category that is unusually noisy right now.
What is AEO, and is it different from GEO?
AEO, answer engine optimization, is the work of getting your brand named and your pages cited inside the answer an AI system produces, rather than ranked in a list of blue links.
GEO, generative engine optimization, describes the same work. So does “AI search optimisation” and “LLM SEO”. The terms emerged from different corners of the market at roughly the same time and have converged on one practice.
Our breakdown of GEO, SEO and AEO covers where the distinctions genuinely carry meaning. Our answer engine optimization guide covers the mechanics.
The practical consequence matters more than the taxonomy. If a vendor prices AEO as a separate discipline from your existing search work, ask what specifically is different. Google’s own May 2026 guidance states that AEO and GEO are part of SEO. It also states that AI Overviews and AI Mode run on core Search ranking systems, with no separate AI index.
That does not make AI search irrelevant. It means the fundamentals carry over, and what is genuinely new is narrower: measuring answers instead of positions, and earning citations from sources you do not own.
The claims on this topic that do not hold up
This section exists because the SERP for this question is unusually unreliable, and you are likely to encounter these in a sales call.
“72% of target queries win featured snippets within 90 days.” This figure circulates on AEO service pages with no linked study, no sample size and no method. It also describes featured snippets, which are a Google SERP feature that predates AI answers by roughly a decade. A number about snippets is not evidence about ChatGPT citations.
“A managed programme delivers a lower effective cost-per-outcome.” Published by firms selling managed programmes. Nobody can compute a cost per outcome here, because nobody controls AI ranking systems. Google explicitly advises against providers guaranteeing rankings for that reason.
“AEO is 80% strategy and 20% technical.” A tidy split quoted with nothing behind it. Retrieval failures caused more than 70% of chatbot errors in a Stanford-led evaluation of six commercial chatbots across 2,100 questions in May 2026, which suggests the technical share is considerably larger than a fifth.
“AEO requires a completely different approach from SEO.” Contradicted by Google’s published position on its own surfaces.
None of this means AEO services are not worth buying. It means the pages selling them are currently a poor basis for deciding, and you will need to do the arithmetic yourself. Our GEO agency ranking methodology sets out the standard we hold our own claims to.
AEO agency vs AEO tool at a glance
| AEO tool | AEO agency | Platform with a team | |
|---|---|---|---|
| Published cost | Free to $399 monthly at entry tiers | Retainer, rarely published | Retainer, rarely published |
| What it produces | Measurement | Execution | Both |
| Covers earned media | No | The good ones do | Yes |
| Covers technical fixes | No | Varies by firm | Yes |
| Do you get logins | Yes | Usually not | Yes |
| Cheapest way to test | Free tier, one quarter | No cheap version exists | No cheap version exists |
| Where it falls short | Findings with nobody to act on them | You depend on their reporting | More than a small prompt set needs |
AirOps positions as “AI Search for Enterprise” and offers a free Solo tier tracking 100 prompts and pages with ChatGPT insights only. Profound publishes $99 Starter with 50 prompts and 1,500 monthly responses on ChatGPT, and $399 Growth with 100 prompts, 9,000 responses and three answer engines. Peec AI lists four tiers, publishes no figures, and covers five engines at every tier.
The service side publishes almost nothing, ours included. Treat a missing price as information rather than an oversight.
What AEO actually costs
The tool side of this market has a published floor and the service side does not.
AirOps offers a free Solo tier tracking 100 prompts and pages on ChatGPT, with its Pro and higher tiers quoted rather than published. Profound publishes $99 for Starter, covering 50 prompts and 1,500 monthly responses on ChatGPT only. Growth is $399, covering 100 prompts, 9,000 monthly responses and three answer engines. Peec AI publishes nothing across four tiers while covering five engines at each.
Two specs matter more than the headline price. Prompt volume caps what you can watch. So 50 prompts establishes whether a category matters, and does not run one. Response volume is what separates a real change from noise, because engines are probabilistic and a single run establishes nothing.
AEO service pricing is almost never published, ours included. Treat a missing price as information rather than an oversight, and note that the gated half of this market skews expensive.
The cost that decides the purchase is on neither list. A tool surfacing two hundred opportunities a quarter creates two hundred pieces of work. Priced at a loaded hourly rate, that usually lands several times above the subscription. Our cost breakdown of platforms against hiring models that line. Which gives better value works through the unit-cost arithmetic.
How we weighted this decision
| Area | Weight | What decides the score |
|---|---|---|
| Execution capacity | 35 | Whether a named person has protected hours to act on findings, with something removed from their plate |
| Earned media reach | 30 | Whether the model can get you into third-party sources, which supply the large majority of citations |
| Claim quality | 15 | Whether the vendor’s own published claims are sourced, which predicts how they will report your results |
| Data access | 15 | Whether you can read raw answers and cited URLs yourself, and export them |
| Cost transparency | 5 | Whether the price is published, which changes how fast you can build a case |
Claim quality earns its own line here, which it would not in most categories. In a market where the ranking pages carry unsourced statistics, how a vendor treats evidence about their own product is the cheapest available signal about how they will treat evidence about your results.
Why earned media decides this more than the software does
Muck Rack analysed more than 25 million cited links across ChatGPT, Claude and Gemini in May 2026, spanning 17 industries, and found earned media accounted for 84% of AI citations against 0.3% for paid and advertorial.
An AEO tool will identify precisely which publications, review sites and communities decide your category. That is genuinely useful. It cannot pitch one of them.
So the honest framing of the buying question is not “tool or agency”. It is whether you have a route into third-party sources, and if you do not, whether you intend to build one or buy one. A team that buys software and responds by publishing more of its own content is optimising the smallest share of the citation pool.
If you want that citation data pulled for your category before you decide, book a growth audit and we will show you which sources are winning your answers today.
Where does each option fall short?
The AEO tool falls short on conversion to action. It is the right first purchase for a team with capacity. At a free tier it is also the cheapest way to establish whether any of this applies to you. Where it falls short: free and entry tiers are usually single-engine, so a quiet result may be a coverage limit rather than a finding. Nothing in the subscription does the work either.
The AEO agency falls short on transparency and testability. It sells the execution hours that are the actual cost, and it is the only model that reliably covers earned media. Where it falls short: retainers are almost never published and there is no cheap version to trial. You usually do not get the instrument either, so you cannot check reporting against raw data.
The combined model falls short on price discovery. Pepper’s customers log into the platform themselves: workspace setup, brand profile, competitors and personas, GA4 and Search Console connected, themes and prompts managed, GEO analytics read directly, and their own agents built and run in the Agent Atlas.
And a growth team is attached to the account doing the work alongside them, which covers the execution and earned media gaps that break the tool-only path.
Where the combined model falls short: we do not publish pricing, so budget discovery is a conversation rather than a page. We are built for teams running organic as a long-term function, so if you want a one-off audit, a free tier is the better purchase. Our depth sits in content, authority and AI search rather than large technical migrations, so raise that early if it is your primary problem.
How to choose between an AEO agency and an AEO tool
I would start by discounting the marketing on both sides, because in this category that is most of what is available.
Google’s May 2026 guidance is the anchor. AEO and GEO are part of SEO, AI Overviews and AI Mode run on core Search ranking systems with no separate AI index, and Google advises against providers guaranteeing rankings because third parties cannot access internal ranking systems. That is correct for Google, and Google is one engine. ChatGPT and Perplexity retrieve and cite differently, so hold both: the fundamentals carry over, and the measurement and earned media work are genuinely new. Anyone selling AEO as a wholly separate discipline is contradicting the platform they are optimising for.
Score the decision on what you can verify.
| Area | Weight | What a strong answer demonstrates |
|---|---|---|
| Named execution owner | 35 | A specific person with protected weekly hours, and something explicitly removed from their plate to create them |
| Earned media route | 30 | A repeatable way into third-party publications and communities, with a target list drawn from actual citation data |
| Sourcing standards | 15 | Every statistic in their own materials links to a study with a stated sample and method, not a number in a graphic |
| Raw data access | 15 | You can read actual answer text and cited URLs, and export them if you leave |
| Prompt methodology | 5 | Someone can explain how the reported prompt set was chosen, and it is not weighted toward branded queries |
Those weights sum to 100. Score the free tier as a genuine option, because for a first quarter it frequently wins.
The live test, and it costs nothing now. Write 30 real commercial questions your buyers would ask, in their words. Four worked examples: “do we need an AEO agency or can we do this in house”, “how do we get cited by ChatGPT in our category”, “best tools for tracking brand mentions in AI answers”, “what does answer engine optimization actually involve”. Load them into a free tier and let it run, then hand the identical list to any vendor you are considering.
Ask for five things back. Where do we appear and where do we not. Which competitors appear instead, consistently. Which sources influence those answers, sorted by frequency. Why are those sources winning. What would you change in the next 90 days, and who does it. Compare their output against what you ran yourself. The discrepancies are more informative than the deck. Engines are probabilistic, so require more than one run on more than one engine.
The weak playbook against the strong one. The weaker sequence is familiar because it is cheap to deliver. Find some prompts, rewrite a few pages into question-and-answer format, add FAQ schema, sprinkle statistics, and hope for a citation. It is entirely on-site and entirely owned media, aimed at the smaller share of the citation pool.
The stronger sequence runs eight steps: demand intelligence, technical discoverability, entity and brand authority, content, earned-media authority, distribution, visibility measurement, revenue attribution. The distinction matters because the weak version can be delivered by software plus a junior writer. That is precisely why so much of this category sells it under a new acronym.
Red flags, each one a thing a vendor actually says. “72% of target queries win featured snippets in 90 days”, unsourced and about the wrong surface. “We guarantee AI citations”, which Google’s guidance advises against. “AEO is a completely new discipline”, contradicted by Google’s own position. “Here is your AEO score”, a composite with no competitor share of voice and no audit path. “We track your branded prompts”, which always looks healthy. “We optimise for ChatGPT”, where one engine stands in for the category. “We publish 40 AI-optimised articles a month.” And a proposal with no digital PR line, which given the 84% figure is structurally mismatched to how citations actually work.
Five questions for the first call.
- Which of your published statistics can you link to a study? A good answer sends the links during the call. A weak one calls it industry data.
- Do we get logins, and to what exactly? A good answer names the screens and offers a trial. A weak one describes the reporting you will receive.
- What proportion of scope is earned media, and who does it? A good answer names people and a target list. A weak one calls it content amplification.
- How did you choose the prompts in this report? A good answer describes a method and admits which segments are thin. A weak one cannot say.
- What comes off our plate? A good answer is specific about hours and throughput. A weak one talks about partnership.
It all comes down to one principle: in a category this noisy, buy from whoever sources their claims, and buy the thing that removes your binding constraint rather than the thing that measures it. The acronym on the invoice does not matter.
One closing note that costs us something. Very few providers are equally strong across measurement, execution, earned authority and attribution, ours included, and the good ones will tell you which of the four is their weakest. If you have capacity in house and an existing PR function to borrow, you do not need an agency at all, and we would say that on a call.
What nobody should promise you
Guaranteed AI citations. Google advises against providers guaranteeing rankings because third parties cannot access internal ranking systems. The same holds for AI answers.
A snippet-based success rate as evidence of AEO results. Featured snippets are a Google SERP feature. A figure about snippets is not evidence about ChatGPT citations.
That AEO is a separate discipline requiring separate budget. Google places AEO and GEO inside SEO. What is genuinely new is measurement and earned media, not a parallel practice.
A single AEO score worth reporting. Composite numbers hide the gap between Brand Visibility, how often engines mention you, and Domain Prompt Presence, how often they cite your pages.
Frequently asked questions
Do I need an AEO agency or an AEO tool?
A tool if you have people with hours to act on findings. A service if your constraint is execution rather than knowledge. If you are unsure whether AI answers matter in your category, start on a free tier for a quarter and decide with data.
What is the difference between AEO and GEO?
Very little in practice. Both describe getting cited inside AI-generated answers rather than ranked in a list. The terms emerged from different parts of the market and converged on one practice, so treat vocabulary differences as marketing.
Is AEO different from SEO?
Google states that AEO and GEO are part of SEO, and that AI Overviews and AI Mode run on core Search ranking systems with no separate AI index. The fundamentals carry over. What is new is answer-level measurement and earned media.
How much do AEO tools cost?
AirOps offers a free tier tracking 100 prompts on ChatGPT. Profound publishes $99 and $399 tiers. Peec AI lists four tiers with no figures. Service pricing is almost never published, so budget for quoted pricing there.
Can an AEO tool do the work for me?
No. Every tool in this category measures. None of them publish, pitch or fix rendering. Improving citations takes technical work, content and earned media, which is what the service side is actually selling.
Why do AEO agency pages contain unsourced statistics?
The vocabulary is new and the category is competing hard, so numbers circulate faster than studies. Ask any vendor to link the study behind their headline figure, since how they handle their own evidence predicts how they will handle yours.
What should I do first?
Run 30 real commercial prompts across the engines your buyers use, and sort the cited domains by frequency. That list tells you whether your gap is measurement, execution or authority, and it costs a day rather than a budget cycle.
Do I need AEO if my SEO is already strong?
You need the two pieces your SEO scope likely lacks: measurement that reads answers rather than rankings, and a route into third-party sources. Strong SEO covers most of the retrievability work already.
Where to go next
Do the free version first. Load 30 real commercial prompts into a free tier, let it run for a month, and sort the cited domains by frequency. That single list tells you whether you are buying measurement, execution or authority, and it is the cheapest useful thing available in this category.
Our guide to tracking brand mentions in AI search covers the five methods and what each misses. The Visibility, Citability and Retrievability framework explains which lever is stuck.
In adjacent vocabulary, GEO agency vs GEO tool covers the capability comparison. Is a GEO platform worth it without an agency covers whether your team can act on findings. AI SEO agencies covers which firms rebuilt their model rather than renaming their deck. Our SalesHood case study documents AI Overview keywords moving from 14 to 97 across five months.
The honest exit. If your team has capacity, an existing PR function you can borrow, and fewer than about 30 meaningful commercial prompts, you do not need to buy anything. Run the free diagnostic quarterly and revisit when the prompt set outgrows a spreadsheet.
Sources
Vendor details came from each company’s own site on 31 August 2026. Every study cited was published in 2026. The unsourced claims named in this article are described as unsourced deliberately, and we have not linked them, since linking would pass authority to pages carrying figures we could not verify.
- Google Search Central, AI features and your website, 15 May 2026. States that AEO and GEO are part of SEO, that AI Overviews and AI Mode use core Search ranking systems with no separate index, and advises against providers guaranteeing rankings.
- Muck Rack, Earned media still drives 84% of AI citations, 7 May 2026. More than 25 million cited links across ChatGPT, Claude and Gemini, 17 industries. Muck Rack sells PR software, so read it as an interested party with a large dataset.
- Suzgun et al., Evaluating Commercial AI Chatbots as News Intermediaries, arXiv, 21 May 2026. 2,100 questions across six commercial chatbots. Tested on news rather than commercial queries, so read the mechanism rather than the exact rate.
Latest Blogs
Almost every page answering this question was published by someone selling one of the two answers, and several of the headline statistics do not survive checking. Here is what the vocabulary actually means, which claims to discount, and how to decide without them.
The value question has a different answer than it did six months ago, and the reason is that the floor fell out of tool pricing. AirOps now offers a free tier that tracks 100 prompts and pages on ChatGPT. Profound charges $99 a month for 50 prompts on ChatGPT. Both figures are published, both were […]
Google already answered half of this: AEO and GEO are part of SEO, running on core Search ranking systems with no separate index. That settles the technical question and not the organisational one. Here are the three operating models, and why a separate GEO team usually backfires.
Get your hands on the latest news!
