When someone asks ChatGPT for the best med spa in their city, the best dark chocolate or a lawyer who handles their kind of case, they get a short answer with a few names. There's no page two. If you're not named, you never got a shot.
AEO (answer engine optimization) is the work of making those answers more likely to include you. Nobody can buy a spot in them or guarantee one, and anyone who promises that is guessing. But the inputs AI tools lean on are mostly things you control: what your site says, how clearly it says it, and whether the rest of the web agrees.
This is the process we run for AI Visibility clients. It's a new service, so there are no client results on this page yet. Every example is illustrative and labeled. The free PDF has the working pieces: the question list, the baseline sheet, the answer-page brief, the schema checklist, the listings audit and a 30-day plan.

AI Visibility by dependables.
- A monthly report of what ChatGPT, Perplexity, Gemini and Google AI say
- Answer-first pages written for you
- Schema, FAQ and llms.txt set up
- Listings and profiles cleaned up so they match
Run by Oksana, Nikhil and Danilo ›
01 / 06 · THE BASELINE
1. Ask the questions your customers ask

Before you change anything, find out where you stand. Write down the questions a customer would type or say to an AI assistant when they're about to choose someone like you.
Run every question through ChatGPT, Perplexity, Gemini and Google's AI Overviews, and record who gets named, in what order, and which sources each answer cites.
This baseline is the scoreboard for everything after it. It also tells you which competitors AI already trusts, and which pages it reads to decide.
Answers vary between runs and between users, so treat one run as a sample, not a verdict. Re-run the same list every month.
How to set it up
- Write 20 questions in your customers' words: best X in Y, X for a specific need, X vs Y, is X worth it.
- Run each one in ChatGPT, Perplexity, Gemini and Google (AI Overview), logged out where you can.
- Record: are you named, your position, who else is named, and the sources cited.
- Count your mentions out of the total answers. That's your baseline.
- Save the list. You'll run exactly the same questions every month.
The mistake that breaks it: Testing one flattering question once. A single answer to 'best [your brand name]' tells you nothing about how customers actually find you.
In the free PDF: The 20-question list and the baseline sheet.Get the PDF ↓
There's no page two in a ChatGPT answer.From our AI Visibility page
02 / 06 · ANSWER PAGES
2. Put the answer first on your pages

AI tools pull from pages that answer a question clearly and early. Most business sites bury the answer under a welcome paragraph, a slogan and a stock photo.
For each high-value question on your list, you want one page on your site that answers it directly in the first two sentences, in plain words, then backs it up.
That means: what you do, for whom, where, roughly what it costs or how it works, and why someone should pick you. Then the follow-up questions people ask, each with a short answer.
These pages also tend to convert better for human visitors, because they answer what people came to ask.
How to set it up
- Pick the 5 questions from your list with the most buying intent.
- Give each one a page, or a clear section of an existing page.
- Answer it in the first two sentences: what, who, where, how much, why you.
- Add 4–6 follow-up questions with two-or-three-sentence answers.
- Use the words customers use, not your internal service names.
The mistake that breaks it: Writing for keywords instead of answers. A page stuffed with 'best dentist Austin' that never actually answers a question gives an AI nothing to quote.
In the free PDF: Answer-page brief template.Get the PDF ↓
03 / 06 · SCHEMA & FILES
3. Make your site easy for machines to read

AI tools read your pages as text and structured data. If your key facts only exist inside images, scripts or PDFs, they're easy to miss or get wrong.
Structured data (schema) labels the facts: this is a business, this is its address, these are its hours, these are its services, these are questions and answers. It doesn't guarantee anything, but it removes guesswork.
An llms.txt file is a short, plain-text guide at the root of your site that points AI tools to your most important pages. It's a new convention and support varies, but it's cheap to add.
Also check your robots.txt isn't blocking the crawlers that AI search tools use. Some site settings block them by default.
How to set it up
- Add business schema (LocalBusiness or Organization) with name, address, phone, hours and services.
- Add FAQ schema to pages that have real questions and answers on them.
- Make sure prices, hours and services are text on the page, not only in images.
- Publish an llms.txt file listing your key pages with one line each.
- Check robots.txt and your security settings aren't blocking AI search crawlers.
The mistake that breaks it: Adding schema that says things the page doesn't. Structured data has to match what a visitor actually sees.
In the free PDF: Schema and technical checklist.Get the PDF ↓
Get the full playbook, with every template
Everything on this page, plus the working pieces you can copy straight into your program:
- The 20-question list
- Baseline sheet
- Answer-page brief
- Schema and technical checklist
- Listings audit
- Mentions tracker
- Monthly report
- 30-day plan
We'll email you about getting named by AI now and then. Unsubscribe any time.
04 / 06 · LISTINGS
4. Say the same thing everywhere

AI tools don't only read your site. They read listings, directories, review sites and profiles, and they're more confident naming a business when those sources agree.
Inconsistent names, addresses, phone numbers, hours and service descriptions make you look less certain. Old listings from a previous address or name are especially common.
This is unglamorous work: find every listing, fix it, and keep it fixed. It's also some of the highest-value work in the whole playbook.
Start with the big ones in your category, then the directories AI answers actually cited in your baseline.
How to set it up
- List every profile: Google Business Profile, Apple Maps, Bing, Yelp, Facebook, LinkedIn, industry directories.
- Pick one exact version of your name, address, phone, hours and main services.
- Update every listing to match it, and close duplicates.
- Add the sources your baseline answers cited, if you're missing from them.
- Check them again every quarter.
The mistake that breaks it: Fixing the site and ignoring the listings. AI answers are built from many sources, and your site is only one of them.
In the free PDF: Listings audit sheet.Get the PDF ↓
05 / 06 · MENTIONS
5. Get talked about where AI reads

Your baseline tells you which outside sources AI answers lean on for your questions: review sites, 'best of' lists, local press, forums and comparison articles.
Being present and well-described in those sources makes it more likely you'll be named. It's the same thing that makes you easier to find for humans.
The honest version of this is slow: steady reviews from real customers, being genuinely useful in the places your customers ask questions, and pitching the lists and writers that cover your category.
Don't fake it. Fake reviews and sockpuppet posts break platform rules, and they're the fastest way to lose the trust you're trying to build.
How to set it up
- From your baseline, list every outside source that was cited for your questions.
- Mark where you're missing, wrong or thin.
- Ask recent happy customers for reviews on the sites that were cited.
- Pitch the 'best of' lists and writers that cover your category, with facts they can check.
- Where your customers ask questions in public, answer as yourself and disclose who you are.
The mistake that breaks it: Chasing mentions everywhere. Start with the sources AI actually cited in your baseline.
In the free PDF: Mentions tracker.Get the PDF ↓
06 / 06 · THE LOOP
6. Re-run the questions every month

AI answers change as the tools change and as the web changes. A one-time project fades. A monthly loop compounds.
Every month, run the same question list on the same four tools, compare with last month, and pick the next fixes from what moved and what didn't.
Report it the same way every time: mentions out of total answers, positions, and which of your pages got cited. That's how you know the work is working.
This is the part most businesses skip, because nothing is broken. It's also the part that keeps you named.
How to set it up
- Same 20 questions, same four tools, same day each month.
- Record mentions, positions and cited pages next to last month's.
- Pick next month's 2–3 fixes from the questions that still don't name you.
- Add a new answer page or update an old one every month.
- Share the one-page report with whoever owns marketing.
The mistake that breaks it: Judging it on one month. Answers move slowly and unevenly; read the trend across a quarter.
In the free PDF: Monthly report template.Get the PDF ↓
THE HONEST PART
Your first 30 days, if you do it yourself
Here's how we'd sequence it. It's our plan, and there are no client results behind it yet.
If you have someone who can run the questions, fix the pages and keep the listings clean every month, run the playbook. If you don't, that's the job we do: AI Visibility puts one team on it.
Get the full playbook, with every template
Everything on this page, plus the working pieces you can copy straight into your program:
- The 20-question list
- Baseline sheet
- Answer-page brief
- Schema and technical checklist
- Listings audit
- Mentions tracker
- Monthly report
- 30-day plan
We'll email you about getting named by AI now and then. Unsubscribe any time.
Questions
Can anyone guarantee ChatGPT will name us?
No. AI answers change and nobody can buy a spot in them. You can make it more likely, and measure it every month.
Is this just SEO?
It overlaps: clear pages, structured data and consistent listings help both. The difference is the scoreboard: you measure whether AI tools name you, not where a page ranks.
Which AI tools matter?
Start with ChatGPT, Perplexity, Gemini and Google's AI Overviews. Your customers may use others, so ask them.
What's llms.txt?
A short plain-text file at the root of your site that points AI tools to your key pages. It's a new convention and support varies, but it takes minutes to add.
What's in the PDF?
The 20-question list, baseline sheet, answer-page brief, schema checklist, listings audit, mentions tracker, monthly report and the 30-day plan.
Can you do it for us?
Yes. That's AI Visibility: the monthly report, the answer pages, schema and llms.txt, and the listings cleanup, done for you.
AI Visibility is a new dependables. service: there are no client results on this page, and every example box is illustrative. The steps and templates are our own process. No one can guarantee placement in AI answers.
Sources