The AEO library

Fix negative AI brand sentiment

From What Is Brand Sentiment in AI Answers?

1. Collect the negative answers
2. Open the cited sources
3. Decide whether each claim is wrong or fair
4. Correct your own pages
5. Update outdated third-party sources
6. Track sentiment per engine over weeks

1. Collect the negative answers. Save the full text of qualified and negative answers, grouped by engine and prompt.
2. Open the cited sources. Read the sources cited on those answers and note which ones carry the negative framing.
3. Decide whether each claim is wrong or fair. Treat an inaccurate claim, such as a wrong price or a feature you do support, as a correction problem, and share a fair criticism with the product team.
4. Correct your own pages. Make pricing, features and limitations explicit and current on your own site, because engines take your own words at face value.
5. Update outdated third-party sources. Ask publishers to update outdated articles and listings, and respond where old community threads still circulate.
6. Track sentiment per engine over weeks. Score sentiment per engine, only on answers that name you, and judge the change over several weeks.

Track your brand across AI engines

From How to Track Your Brand Across AI Engines (Step by Step)

1. Build and freeze a buyer-worded prompt set
2. Ask every prompt from a clean session, on a schedule
3. Save the full answer
4. Record five fields from each answer
5. Read the trend over weeks

1. Build and freeze a buyer-worded prompt set. Write the questions buyers ask AI engines about your category, in their own words, keep branded prompts in a group of their own, and stop changing the set.
2. Ask every prompt from a clean session, on a schedule. Log out or turn off memory and personalisation, paste each prompt exactly, and run the full set on fixed days.
3. Save the full answer. Keep the full text of every answer with its date and engine, so you can re-score it or check a surprising result later.
4. Record five fields from each answer. Note whether your brand is named, its position, the competitors named, the sources cited and the sentiment, scoring sentiment only when you are named.
5. Read the trend over weeks. Read each engine on its own before combining them, and judge changes on trends over several weeks rather than single runs.

Rank in ChatGPT

From How to Rank in ChatGPT: What 2,545 AI Answers Show (2026)

1. Write pages for the short queries ChatGPT sends
2. Track mentions and citations separately
3. Allow OAI-SearchBot in robots.txt and at your CDN
4. Decide on GPTBot and ChatGPT-User separately
5. Serve your text in the HTML
6. Put the answer in the first lines
7. Add tables, dates and sources
8. Earn accurate mentions on third-party sites
9. Measure a fixed set of prompts over time

1. Write pages for the short queries ChatGPT sends. When ChatGPT searches, it usually rewrites a question into shorter queries, so give each product category, comparison and named problem a page that matches one of them.
2. Track mentions and citations separately. Being named in an answer and being one of the sources it links are separate results, so record both for every prompt you measure.
3. Allow OAI-SearchBot in robots.txt and at your CDN. OpenAI says that to appear in ChatGPT search, a site must let OAI-SearchBot crawl it and let through traffic from the IP addresses OpenAI publishes for its search bot.
4. Decide on GPTBot and ChatGPT-User separately. GPTBot collects content that may be used for training, and ChatGPT-User visits a page when a person asks about it. Blocking GPTBot does not remove a site from ChatGPT search.
5. Serve your text in the HTML. Check that your words are in the page before any JavaScript runs, because a crawler that does not run it sees an empty page.
6. Put the answer in the first lines. Open each page with a one-sentence answer to its question, then the detail, and use headings phrased the way buyers ask.
7. Add tables, dates and sources. Put comparisons in tables, say when each fact was checked, and link where it came from.
8. Earn accurate mentions on third-party sites. Review sites, comparison articles, industry publications and community threads are where a rewritten query is likely to land, so make sure you are described accurately there.
9. Measure a fixed set of prompts over time. Run the same buyer questions on a schedule and record whether your brand is named, where it appears, how it is described and which sources are cited.

Engines named: ChatGPT

Track brand mentions in ChatGPT by hand

From How to Track Brand Mentions in ChatGPT (2026 Guide)

1. Write the prompts your buyers ask
2. Ask each prompt from a clean session
3. Save the full answer
4. Record whether you are named and what is cited
5. Repeat on a schedule and read the trend

1. Write the prompts your buyers ask. List category, comparison and problem prompts in your buyers' own words, plus a few branded prompts kept in a group of their own, and stop changing the list.
2. Ask each prompt from a clean session. Log out, or turn off memory and personalisation, and paste each prompt exactly, so your own history does not colour the answer.
3. Save the full answer. Keep the full text of every answer with its date, so you can re-score it or check a surprising result later.
4. Record whether you are named and what is cited. For each answer, note whether your brand is named, where it appears, which competitors are named, which sources are cited and how you are described.
5. Repeat on a schedule and read the trend. Run the same prompts on fixed days and judge changes over several weeks, because ChatGPT's answers vary from one run to the next.

Engines named: ChatGPT

Build a GEO strategy

From GEO Strategy for 2026: A Step-by-Step Playbook to Win AI Search

1. Choose the buyer questions you want to win
2. Baseline every engine your buyers use
3. Make your site readable by AI crawlers
4. Write answer-first pages for the gaps
5. Earn mentions on the sources engines cite
6. Correct what the engines get wrong
7. Measure monthly and report share of voice

1. Choose the buyer questions you want to win. List 25 to 100 questions buyers ask an AI assistant before they choose a product like yours, grouped by stage: learning about the problem, comparing options and deciding.
2. Baseline every engine your buyers use. Ask each question in each engine and record whether you are named, your position, which competitors are named and which sources are cited, so every later change has a starting point.
3. Make your site readable by AI crawlers. Check that robots.txt and your CDN let search and AI crawlers fetch your pages, that the main content is in the HTML, and that structured data matches the visible text.
4. Write answer-first pages for the gaps. For each question where competitors are named and you are not, publish or rewrite a page that answers it in the first lines, then gives the detail, facts and dates an engine can cite.
5. Earn mentions on the sources engines cite. Find the review sites, lists, forums and publications the engines cite for your questions, and work to be listed on them and described accurately.
6. Correct what the engines get wrong. Where answers misstate your prices, features or category, fix the source pages they draw on and publish a clear, dated statement of the facts on your own site.
7. Measure monthly and report share of voice. Re-run the same questions on a schedule and report visibility, position, share of voice and citations per engine, alongside the visits that arrive from AI assistants.

Estimate AI prompt volume for your category

From AI Prompt Volume: How to Estimate What People Ask AI Engines

1. Start from the search demand behind each question
2. Add the long questions from your own data
3. Check what people ask in public
4. Group prompts and score them in bands
5. Weight by value and revisit quarterly

1. Start from the search demand behind each question. Map every prompt to the short Google search it grows out of and record that keyword's monthly search volume as a relative measure of interest in the topic.
2. Add the long questions from your own data. Pull long, question-shaped queries from Google Search Console, site search, sales calls and support tickets; they show how your buyers phrase questions when they have room to.
3. Check what people ask in public. Read People Also Ask, autocomplete, Reddit and community threads for your category to find the questions and qualifiers that recur.
4. Group prompts and score them in bands. Group similar prompts under one topic and give each topic a demand band, such as high, medium, low or unknown, rather than a precise number you cannot verify.
5. Weight by value and revisit quarterly. Rank topics by demand band and by how close each question is to a purchase, track the top ones in AI answers, and revisit the bands every quarter as the data moves.

Set up AEO governance

From AEO Governance: How to Run AI Search Across Teams and Regions

1. Name one owner for AI search
2. Agree one shared question set
3. Publish a brand facts page
4. Set content standards for AI search
5. Run a corrections process
6. Track each region and language separately
7. Report one set of numbers on one cadence

1. Name one owner for AI search. Give one person or team responsibility for the question set, the engines tracked, the standards and the monthly report, with named contacts in content, PR, product marketing, legal and each region.
2. Agree one shared question set. Keep a single list of the buyer questions you track, grouped by product line, stage and market, so every team measures against the same questions and results can be compared.
3. Publish a brand facts page. Maintain one public page with the facts AI assistants should use about your company: what it is and is not, products, prices, coverage and official links, owned by one team and dated.
4. Set content standards for AI search. Write down the rules every team follows: answer-first openings, one page per question, sourced and dated figures, structured data that matches the visible text and crawler access that is never blocked by accident.
5. Run a corrections process. When an AI answer gets a fact wrong, log it, find the source it came from, assign the fix to the team that owns that source and re-check the answer once the source is corrected.
6. Track each region and language separately. Measure the same questions in each market's language and location, because engines can answer differently by country, and let regional teams add local questions under the shared standards.
7. Report one set of numbers on one cadence. Produce one monthly report with visibility, share of voice, sentiment and citations per engine and region, and review the question set and standards every quarter.

Check whether AI answers cite YouTube for your category

From YouTube in AI Answers: How to Check If Your Videos Get Cited

1. List the questions that matter
2. Ask them in engines that show sources
3. Count YouTube citations per engine
4. Note whose videos are cited
5. Repeat on a schedule

1. List the questions that matter. Write down 20 to 50 questions buyers ask about your category, including how-to and comparison questions, which are the kinds a video can answer.
2. Ask them in engines that show sources. Put each question to the AI engines your buyers use, from a fresh session, and keep the list of sources each answer cites.
3. Count YouTube citations per engine. Mark every source on youtube.com or youtu.be, and count the answers that cite YouTube for each engine and each question.
4. Note whose videos are cited. Open each cited video and record the channel: yours, a competitor's, a creator's or a publisher's, and what the video covers.
5. Repeat on a schedule. Ask the same questions again over several weeks, because AI citations change from run to run and one check can mislead.

Get on the vendor shortlists AI writes

From How to Get on the Vendor Shortlists AI Writes (B2B SaaS)

1. Find the shortlist questions buyers ask
2. Ask them in every engine your buyers use
3. Publish pages that answer the shortlist question
4. Get onto the lists the engines read
5. Keep comparisons honest and current
6. Measure the shortlists over time

1. Find the shortlist questions buyers ask. List the 'best X for Y' and 'which X should I use' questions for your category, with the qualifiers buyers add: team size, budget, industry and stack.
2. Ask them in every engine your buyers use. Record which vendors each engine names, in what order, how it describes them and which sources it cites, because engines agree on a few names and differ on the rest.
3. Publish pages that answer the shortlist question. Give each use case and buyer type its own page that says plainly who you are for, with prices, limits and integrations in the HTML, so an engine can match you to the qualifier in the question.
4. Get onto the lists the engines read. Find the comparison articles, directories and review pages cited for your questions and make sure you are on them, described accurately and up to date.
5. Keep comparisons honest and current. Publish dated comparisons against the vendors engines already name, with sourced figures, so your page is a fair source for the question buyers ask.
6. Measure the shortlists over time. Re-ask the same questions on a schedule and track how often you are named, your position and which competitors appear instead.

What an AEO programme actually does

From The AEO guide, chapter 1: What AEO is, and what it is not

1. Decide what you want engines to say about you, specifically enough to check (chapter 2).
2. Choose the questions that test whether they say it (chapter 3).
3. Measure where you stand today, per engine and per claim (chapter 4).
4. Change the pages and sources that shape the answer (chapter 5).
5. Keep measuring, and spread the work across the teams that own the sources (chapter 6).

1. Decide what you want engines to say about you, specifically enough to check (chapter 2).
2. Choose the questions that test whether they say it (chapter 3).
3. Measure where you stand today, per engine and per claim (chapter 4).
4. Change the pages and sources that shape the answer (chapter 5).
5. Keep measuring, and spread the work across the teams that own the sources (chapter 6).

Your own pages: make the claim easy to lift

From The AEO guide, chapter 5: Turn the baseline into work

1. Is there a page whose main job is this claim? If not, that page is the first piece of work.
2. Which pages already get cited or fetched? Those are the places where stating the claim reaches an engine soonest.
3. Is the claim stated plainly, near the top, in a sentence that still makes sense quoted on its own?

1. Is there a page whose main job is this claim? If not, that page is the first piece of work.
2. Which pages already get cited or fetched? Those are the places where stating the claim reaches an engine soonest.
3. Is the claim stated plainly, near the top, in a sentence that still makes sense quoted on its own?

Other people's pages: change the consensus

From The AEO guide, chapter 5: Turn the baseline into work

1. Fix the mentions you already have. Out-of-date pricing, old product names and wrong descriptions on pages that already name you are the cheapest wins. Start with partners, integrations and customers, where you have a relationship.
2. Get onto the pages that are already cited without you. Comparison articles, lists and directories from your baseline's source list. Earn the place with evidence and a clear reason to include you.
3. Create new sources. Original data others want to reference, practitioner writing on credible sites, and genuine participation in the communities your buyers use.

1. Fix the mentions you already have. Out-of-date pricing, old product names and wrong descriptions on pages that already name you are the cheapest wins. Start with partners, integrations and customers, where you have a relationship.
2. Get onto the pages that are already cited without you. Comparison articles, lists and directories from your baseline's source list. Earn the place with evidence and a clear reason to include you.
3. Create new sources. Original data others want to reference, practitioner writing on credible sites, and genuine participation in the communities your buyers use.

Correcting what an engine gets wrong

From The AEO guide, chapter 5: Turn the baseline into work

1. Record the exact prompt, the engine and the full answer, so the error can be reproduced.
2. Find where it came from: look at the sources cited, and search for the wrong claim itself.
3. Fix your own pages first, so the correct version is clearly stated somewhere authoritative.
4. Ask the owners of third-party sources to correct theirs, with your reference page attached.
5. Re-run the prompt on a schedule rather than once, because answers vary from run to run.

1. Record the exact prompt, the engine and the full answer, so the error can be reproduced.
2. Find where it came from: look at the sources cited, and search for the wrong claim itself.
3. Fix your own pages first, so the correct version is clearly stated somewhere authoritative.
4. Ask the owners of third-party sources to correct theirs, with your reference page attached.
5. Re-run the prompt on a schedule rather than once, because answers vary from run to run.

Where to start tomorrow

From The AEO guide, chapter 6: Run it as a programme

1. Write the sentence you want an engine to say about you.
2. Pick the three claims that matter most to buyers.
3. Write a handful of buyer-worded prompts for each, and ask them of every engine you care about.
4. Read the answers and the sources, and write the three lists.
5. Take one finding to the team that owns it.

1. Write the sentence you want an engine to say about you.
2. Pick the three claims that matter most to buyers.
3. Write a handful of buyer-worded prompts for each, and ask them of every engine you care about.
4. Read the answers and the sources, and write the three lists.
5. Take one finding to the team that owns it.

What scheduling software suits a small physio clinic?

From The AEO guide, chapter 3: Choose prompts that measure something

Category prompt

What scheduling software suits a small physio clinic?

Is Brand A or Brand B better for a two-person clinic?

From The AEO guide, chapter 3: Choose prompts that measure something

Comparison prompt

Is Brand A or Brand B better for a two-person clinic?

How do clinics cut down on missed appointments?

From The AEO guide, chapter 3: Choose prompts that measure something

Problem prompt

How do clinics cut down on missed appointments?

Does Brand A send text reminders?

From The AEO guide, chapter 3: Choose prompts that measure something

Direct prompt

Does Brand A send text reminders?

How much does Brand A cost per practitioner?

From The AEO guide, chapter 3: Choose prompts that measure something

Fact check prompt

How much does Brand A cost per practitioner?

Best tools for X

From Which Prompts Should You Track to Monitor AI Search?

Category prompt

Best tools for X

What should I use to do Y

From Which Prompts Should You Track to Monitor AI Search?

Category prompt

What should I use to do Y

Top platforms for Z

From Which Prompts Should You Track to Monitor AI Search?

Category prompt

Top platforms for Z

X versus Y

From Which Prompts Should You Track to Monitor AI Search?

Comparison prompt

X versus Y

Alternatives to X

From Which Prompts Should You Track to Monitor AI Search?

Comparison prompt

Alternatives to X

Is X better than Y for Z

From Which Prompts Should You Track to Monitor AI Search?

Comparison prompt

Is X better than Y for Z

How do I stop X happening

From Which Prompts Should You Track to Monitor AI Search?

Problem prompt

How do I stop X happening

Why is Y so slow

From Which Prompts Should You Track to Monitor AI Search?

Problem prompt

Why is Y so slow

What do I do about Z

From Which Prompts Should You Track to Monitor AI Search?

Problem prompt

What do I do about Z

What is [you]

From Which Prompts Should You Track to Monitor AI Search?

Branded prompt

What is [you]

How much does [you] cost

From Which Prompts Should You Track to Monitor AI Search?

Branded prompt

How much does [you] cost

Does [you] support Z

From Which Prompts Should You Track to Monitor AI Search?

Branded prompt

Does [you] support Z

What is the best tool for [job]?

From How to Track Brand Mentions in ChatGPT (2026 Guide)

Category prompt

What is the best tool for [job]?

Engines named: ChatGPT

[Competitor] versus [competitor]

From How to Track Brand Mentions in ChatGPT (2026 Guide)

Comparison prompt

[Competitor] versus [competitor]

Engines named: ChatGPT

Alternatives to [your brand]

From How to Track Brand Mentions in ChatGPT (2026 Guide)

Comparison prompt

Alternatives to [your brand]

Engines named: ChatGPT

What is [your brand] and what does it do?

From How to Track Brand Mentions in ChatGPT (2026 Guide)

Branded prompt

What is [your brand] and what does it do?

Engines named: ChatGPT

Best CRM tools for a B2B SaaS startup with a 10-person sales team

From How to Get on the Vendor Shortlists AI Writes (B2B SaaS)

Shortlist question

Best CRM tools for a B2B SaaS startup with a 10-person sales team

Engines named: ChatGPT, Gemini, Claude, Perplexity

Best customer support software for a 50-person B2B SaaS company

From How to Get on the Vendor Shortlists AI Writes (B2B SaaS)

Shortlist question

Best customer support software for a 50-person B2B SaaS company

Engines named: ChatGPT, Gemini, Claude, Perplexity

Best product analytics tools for a B2B SaaS product team

From How to Get on the Vendor Shortlists AI Writes (B2B SaaS)

Shortlist question

Best product analytics tools for a B2B SaaS product team

Engines named: ChatGPT, Gemini, Claude, Perplexity

Best project management tools for a remote B2B engineering team

From How to Get on the Vendor Shortlists AI Writes (B2B SaaS)

Shortlist question

Best project management tools for a remote B2B engineering team

Engines named: ChatGPT, Gemini, Claude, Perplexity

Best AI visibility (AEO) tools for a B2B SaaS marketing team

From How to Get on the Vendor Shortlists AI Writes (B2B SaaS)

Shortlist question

Best AI visibility (AEO) tools for a B2B SaaS marketing team

Engines named: ChatGPT, Gemini, Claude, Perplexity

How visible is my brand in AI answers today, and how has that changed this month?

From Noma MCP Server for Claude and Cursor

Ask in Claude, Claude Code or Cursor with the Noma MCP server connected

How visible is my brand in AI answers today, and how has that changed this month?

Engines named: Claude

Which AI answers mentioned us this week, and what did they say?

From Noma MCP Server for Claude and Cursor

Ask in Claude, Claude Code or Cursor with the Noma MCP server connected

Which AI answers mentioned us this week, and what did they say?

Engines named: Claude

Which questions do AI engines answer with my competitors but not with us? Draft a brief for the top one.

From Noma MCP Server for Claude and Cursor

Ask in Claude, Claude Code or Cursor with the Noma MCP server connected

Which questions do AI engines answer with my competitors but not with us? Draft a brief for the top one.

Engines named: Claude

What is my site’s AEO readiness score, and what should I fix first?

From Noma MCP Server for Claude and Cursor

Ask in Claude, Claude Code or Cursor with the Noma MCP server connected

What is my site’s AEO readiness score, and what should I fix first?

Engines named: Claude

Re-check the prompt about pricing tools and tell me when it has run.

From Noma MCP Server for Claude and Cursor

Ask in Claude, Claude Code or Cursor with the Noma MCP server connected

Re-check the prompt about pricing tools and tell me when it has run.

Engines named: Claude