Answer engine optimization: be the answer, not a result.
Your buyers ask ChatGPT, Gemini, Perplexity and Claude which vendors to consider, and those tools answer without showing a list of links. We get you inside the answer. Most engagements open with an AI visibility audit, so the work starts from a measured baseline rather than a guess.
● Start here
Tell us about the business.
It takes about a minute. It reaches the people who would do the work rather than a sales inbox, and if we are not the right fit for what you need we will say so.
Goes to hello@katama.io. No sequences, no newsletter, no list.
● What is answer engine optimization
Answer engine optimization, or AEO, is the work of getting your company cited inside AI-generated answers rather than only ranked in a list of links.
Generative engine optimization, GEO, is the same discipline under a different name. It overlaps with SEO but is no longer the same job, because the pages an engine quotes are increasingly not the pages that rank.
Ranking and being cited are now two different jobs.
The link between ranking and being cited broke.
In late 2024, roughly 76% of Google AI Overview citations came from pages ranking in the top ten. By February 2026 that overlap had fallen to somewhere between 17% and 38% depending on whose dataset you use.
The ChatGPT citations produced by fan-out queries match just 6.82% of Google’s top ten. Across four assistants and 15,000 long-tail queries, only about 12% of the URLs cited in AI answers rank in Google’s top ten for the prompt that produced them.
Citing sources
40%Visibility lift measured from citing source material, the highest performing single tactic tested.
So a company can rank well and be invisible in the answer a buyer is shown. The reverse is also true, and that is the opportunity: the pages that get cited are winnable without waiting for a domain to age.
What decides it is largely on the page, and it has been measured.
Adding statistics
37%Lift from adding relevant statistics to a page.
Adding quotations
22%
Visibility lift measured across generative engines from adding attributed quotations to a page.
Brand-controlled pages
88.3%
Of citations at the B2B consideration stage point back at pages the brand itself owns, not at forums.
Review platforms
34.5%
Of AI Overviews cite at least one review platform, which is why the off-site half of this work matters.
Listicles
69%
Of the time an AI Overview citing a self-serving listicle goes on to recommend a competitor out of it.
What we change.
Most of the work is on your own site. Brand-controlled pages account for 88.3% of citations at the consideration stage, which contradicts most of the advice being sold about seeding Reddit threads.
We start by measuring where you appear today across ChatGPT, Google AI Overviews and AI Mode, Perplexity and Gemini, for the prompts your buyers actually use. Almost nobody has this baseline, and without it the work is guesswork.
Then we restructure the pages that should be cited, mark up the entities so the engines know what you are, and go after the third-party sources they lean on for your category.
This page is built to the same specification. So are the ones we build for you.
● Included
- AI visibility baseline across five engines
- Prompt set built from real buyer language
- Share of answer tracking against competitors
- Answer-first content restructuring
- Statistic, source and quotation engineering
- Entity and schema markup
- Third-party citation building in the sources engines lean on
- Review platform and roundup placement
- AI crawler access checked and fixed in robots.txt
- Monthly share of answer reporting, including the months it falls
How the work runs.
Measure the baseline
A prompt set built from how your buyers actually ask, run across five engines. You find out what the machines currently say about you, which is often the most uncomfortable slide in the deck.
Restructure the pages
Answer-first openings, cited statistics, attributed quotations, clean entity markup. Changes to what is on the page, so they can move in weeks rather than quarters.
Build the citations
The third-party sources that engines lean on for your category: review platforms, roundups, and the publications that already rank for the terms you want.
Track the share
Monthly reporting on how often you appear, for which prompts, against which competitors, and which pages got quoted.
What we report.
Share of answer
How often you appear across the prompt set, per engine, against the competitors who appear instead of you.
Cited pages
Which of your URLs get quoted, which tells you what to build more of.
AI referral pipeline
Sessions and opportunities arriving from AI sources, tracked separately because they convert differently.
Who this suits.
- Your buyers research vendors before they contact anyone
- You are already invisible in AI answers and can prove it
- Competitors are being recommended in your category
- You can approve changes to existing pages quickly
- You want a guaranteed position, which nobody can sell honestly
- You will not publish or attribute real numbers
- You expect it to replace SEO rather than sit beside it
- Your buyers are not using AI tools yet, which is worth checking rather than assuming
Fair questions.
Is AEO just SEO with a new name?
No, and the overlap data is the reason. Strong SEO still helps: pages ranking first in Google are cited by ChatGPT about 3.5 times more often than pages outside the top twenty. But the majority of AI citations now come from outside the top ten, so treating them as one workstream leaves most of the opportunity on the table.
What is the difference between AEO and GEO?
Nothing meaningful. Answer engine optimization and generative engine optimization are used interchangeably across the industry and the academic work. We use one page for both rather than splitting them, because two near-identical pages would compete with each other for the same queries.
Do you build an llms.txt file?
Only if you ask. Google has said it does not use it, no major AI vendor has confirmed using it, and an Ahrefs study of 137,000 sites found 97% of llms.txt files were never fetched. We will ship one because it costs nothing, but we will not put it on an invoice. The file that actually controls AI crawler access is robots.txt, which GPTBot, OAI-SearchBot, ClaudeBot and PerplexityBot all respect.
How fast does this move?
Structure work can change which pages get quoted within weeks, because it changes what is on the page rather than what the web thinks of it. Authority and citation building runs on the same slow arc as SEO: meaningful movement by month three, step change by month six to nine.
Can you guarantee we get cited?
No. The engines are probabilistic and change without notice. What we commit to is the baseline, the work, and honest monthly reporting on share of answer, including the months it goes down.
Works well with.
SEO
The other half. Ranking still raises citation probability substantially, and the two workstreams share the same content foundation.
Read →Website conversion
Being cited sends a pre-qualified visitor. What they land on decides whether it was worth it.
Read →Brand & messaging
Engines describe you using the language you publish. If your positioning is vague, so is the answer.
Read →