Which websites ChatGPT and Gemini cite when an American buyer asks for marketing help, from 19,793 citations

On 2026-10-22 we asked ChatGPT and Gemini 318 buying questions an American marketing buyer would actually type, recorded every source each answer returned, and got 19,793 citation URLs across 635 readings. Those citations point at 3,307 distinct hosts, and half of all 19,793 of them come from just 51. Our own domain appears in 0 of them.

That last sentence is the reason to keep reading. We sell the work that moves a brand into these answers, we ran the measurement, and the measurement says we are not in it. The question set is printed in full at the end of this page so anybody can send the same questions and check the result.

What was measured, and on what date

Item	Value
Date of run	2026-10-22
Market	United States, fixed in the prompt text
Engines	ChatGPT and Gemini, both through the same provider. Two engines, not five. Perplexity, Claude and Google AI Overviews were not in this run.
Readings scored	635
Readings that ran a live web search	550 of 635
Citation URLs recorded	19,793
Distinct hosts behind them	3,307
Hosts holding half of all citations	51
Readings naming our own company in the answer text	0
Citation URLs on our own domain	0

The concentration is the finding, and it makes the work smaller

3,307 hosts is a long tail and it is the number most people would quote. The useful figure is the other one. Half of every citation these two engines returned, on 318 American buying questions, came from 51 hosts. Every other host in the dataset shares the other half, which means most of them were cited once or twice.

For a marketing team that changes the shape of the job. An off site plan that lists a hundred places to get mentioned is spreading a budget across the half of the dataset where single citations live. A plan built on the short list at the top is working on the half that carries the weight. The short list is also stable enough to audit: ask any vendor quoting you an AI visibility number which hosts sat behind it, and the answer either names them or it does not exist.

The 25 most cited hosts in this run

Counts are citation URLs, not readings. The fourth column is a warning, explained under the table.

#	Host	Citations	Note
1	developers.google.com	1,360
2	semrush.com	744
3	support.google.com	695
4	help.openai.com	467
5	developers.google.cn	447	Google own regional hostname, same documentation as row 1
6	blog.google	394
7	ahrefs.com	354
8	peec.ai	328
9	techradar.com	286
10	reddit.com	277
11	otterly.ai	255
12	bing.com	241
13	searchengineland.com	237
14	ccn.web.tr	233	the data provider fetch infrastructure, not a publisher
15	tryprofound.com	227
16	developers.openai.com	191
17	github.com	171
18	developers-google-com.hckseu1-xqad.com	171	the data provider fetch infrastructure, not a publisher
19	blogs.bing.com	158
20	help.otterly.ai	143
21	g2.com	125
22	hubspot.com	119
23	business.google.com	119
24	openai.com	116
25	learn.microsoft.com	112

Platform documentation leads it, and on these questions that is permanent rather than competitive. When somebody asks what a crawler directive does, the company that wrote the directive publishes the answer, and nothing is going to outrank it. We put a number on that pattern in an earlier run, where one documentation host was a source in 446 of 1,260 readings, and the reading that produced an action was about our question selection rather than about Google.

Three rows in that table are one body of writing under another name

Rows 14 and 18 are not publishers. Open the addresses and they resolve to Google's developer documentation served through our data provider's own fetching setup, an iframe proxy and a reverse proxy mirror with the same path structure. Nobody at an engine picked them as a source. Row 5 is a different case and the distinction matters: that one is Google's own regional hostname carrying the same documentation as row 1, so the settled rule on this site is to keep the row and label what it serves.

All three are left in the table because the table is the raw tally the generated evidence file holds, and quietly merging them would put a second, undocumented number on the page. The merge rule, and the correction that produced it, is written up where a second hostname served the same documentation in 325 readings. Counting a proxy as a publisher is how a source table overstates the breadth of a field, and it cost us a published correction once already. If you build this measurement yourself, open the cited addresses before you name a domain in a report.

What one citation means here, because the rule changes the total

19,793 is every citation URL the two engines returned. It is not one per host per reading, and the two rules give very different totals on the same raw data. We published that arithmetic on an earlier dataset, where the same 1,260 readings produce 20,143 or 36,576 depending on one scoring rule nobody states. Any AI citation figure you are shown, including this one, is only comparable to another figure counted the same way.

The two totals on this page follow the same discipline. 19,793 is the measurement arm. A second arm ran the same questions in batches through a three message pipeline and recorded 2,583 citations across 38 readings. Those two numbers measure different runs and are never added together.

550 of 635 readings ran a live search

A reading that searched the web is a reading about the live web. A reading that did not is a reading about what the model already holds, which is a different thing with a different fix: the first responds to a page you publish this month, the second responds to the corpus the model was trained on. Both are in the 635, and the split is on this page so that a rate taken from it can be qualified properly.

This matters most when somebody quotes you a single AI visibility percentage. If the readings behind it were not checked for a live search, the number mixes two populations.

Our own row, published because the rest is worth less without it

In 635 readings our domain appeared in 0 of 19,793 citation URLs and in 0 answers by name, and what we published between this pass and the one before it is set out in whether 507 new pages changed what these engines say about us. Those are two separate findings with two separate causes, and we never add them, average them or set one against the other. Cited means a URL on our domain was in the sources. Named means our name was in the prose a reader sees.

Rivals were named 1,053 times across a roster of 24, counted once per reading in which a name appears, and the roster itself is published in which brands these assistants put forward for an American buyer. 451 readings named a rival and not us. 184 readings named nobody at all, and that last group is the one with room in it. The earlier version of this result, from a narrower bank in three markets, is published as our own domain was cited 3 times in 597 readings.

How to use this dataset if you are buying marketing help

• Ask for the host list. A vendor who measures AI citations can hand you the hosts behind the number. If the deliverable is a score with no host list, there is nothing in it you can act on.
• Ask how one citation was counted. Every URL, or one per host per reading. The answer changes the total by a wide margin.
• Ask how many of the readings searched. A reading answered from training data will not respond to a page published last week.
• Measure the category, not only your own name. A brand that is absent from every answer in its category has a different problem from one that is present and placed last, and measuring what assistants tell buyers about your industry is how that is separated.
• Put the off site work on the short list. 51 hosts carried half of these citations, and the table above shows what kind of host they are. One review directory does sit in the top 25, g2.com at 125 citations, so a directory listing is a part of this picture rather than the whole of it.
• Freeze your question set before you measure anything. A host total describes your questions at least as much as it describes the engines, which is why ours is printed below. Choosing the questions worth tracking is the method we use for that.

The method, written so a stranger can repeat it

Each question was sent once per engine as a single turn with a live web search requested, and every source URL the engine returned was recorded. The prompt is three parts in this order: a persona line, a market lock line, and one instruction to search the web and name the specific tools, platforms, agencies or websites the engine would genuinely recommend and in what order, with the real source URLs it retrieved.

The market is set in the prompt text, not by an API parameter. That choice was measured rather than assumed. The provider accepts country, location, gl and hl on these engines, returns success for every value, and ignores all four. The prompt based lock does work: the United States prompt answers for the United States in dollars.

No question in the measurement arm names our own company. Naming a brand inside a question changes the answer, so the only honest way to find out whether an engine names you is to ask without naming yourself. The evidence file for this run records 324 questions written and 19 batches for the pipeline arm, and the 318 questions the measurement arm actually sent are printed in full at the end.

One limit is a property of the provider rather than of the method. The middle step of our three message pipeline asks an engine to audit its own answer, which needs that answer back in the prompt. One batch answer measured 21,711 characters against a 4,000 character prompt cap, so that step ran by hand in the product interface instead of by API. Truncating a 21,711 character answer to fit would produce an audit of something the engine never said.

What this dataset does not cover

• Two engines, not five. Our product reads five surfaces. This dataset covers ChatGPT and Gemini, so nothing here describes Perplexity, Claude or Google AI Overviews.
• One date. 2026-10-22. This category moves month to month and every figure above carries that date for that reason.
• One reading per question in this arm. A single reading per question measures a tendency rather than a fixed ranking, and we put a number on that drift in how often two identical prompts name the same brands.
• The market was set in the prompt, not by a connection in the country. That is weaker than reading from a real American connection, and a buyer there may see something different.
• 550 of 635 readings ran a live search. The rest sit inside the totals and describe training data rather than the current web.
• Host names are counted as returned. 2 of the top 25 are the data provider fetch infrastructure and 1 more is a regional hostname of the same publisher, all labelled in the table. A merged total is not published here, because the evidence file holds the raw tally and a figure assembled by hand is a figure nobody can reproduce.
• We sell the remedy. This measurement is published by a company that charges money to improve the thing it measures. That is a reason to repeat the run from the question set below rather than to take the table on trust.

Who should not buy the product behind this page

Three kinds of buyer should not. If you want a tool that watches all five surfaces, our cheapest plan that reads all five is Growth at $299 a month. SEO Start at $149 renders 0 questions, 0 credits, 0 engines, so it is the wrong plan for anybody who came here for AI answer monitoring.

If your procurement list requires single sign on, SAML, SOC 2, ISO certification, a data processing addendum or a white label, our pricing page renders none of those words, checked by word boundary on 2026-10-22. Profound's own pricing page, read on 2026-10-22 at tryprofound.com/pricing, lists up to 9 answer engines tracked and SSO, SAML and SOC 2 compliance on its Enterprise tier, with no published price. On that requirement they are ahead of us and the honest answer is to say so.

And if you want one blended visibility score, this is the wrong supplier. Everything above is reported as counts with their denominators, split by engine and by arm, because a single blended number would have hidden both halves of our own result.

Questions about this citation dataset

Why only ChatGPT and Gemini

Both were reachable through one provider by one method on the day of the run, which keeps the comparison between them clean. Adding a third engine through a different route would mix two methods and make the engine comparison meaningless.

Can I repeat this myself

Yes, and that is why the question set is printed below. You will not get identical numbers, because these engines return different answers to the same question. If your result is a long way from ours, we would like to see it.

Does 51 hosts holding half of it mean the tail is worthless

It means the tail is where single citations live. A mention on a long tail host is still a real mention, and on this evidence it is a poor place to spend the first half of a budget.

Why is there no share figure for review directories here

Because the source type split for this run is not in the generated evidence file this page is built from, and a figure assembled by hand is a figure nobody can reproduce. What this run does show is one directory inside the top 25, g2.com at 125 citations. We published a directory share for two earlier datasets, each with its own date and denominator.

How often will this be repeated

The question set is frozen so the next run is comparable to this one. The run date sits on every figure above, and an earlier pass of the same method, on a narrower bank in three markets, is published as 597 readings across ChatGPT and Gemini, alongside what happened to the domain count when the question bank doubled.

The 318 questions, printed so this run can be repeated

Each question was sent with a persona line, a market lock line and the search instruction described above. The batch subjects are the groups the bank is organised in.

Batch	Subject	Questions
1	The category by its own names	6
2	Tools and software	6
3	Agencies and services	6
4	The problem, diagnosis intent	6
5	Measurement and reporting	6
6	Technical surfaces	6
7	Platform specific	6
8	The adjacent search buyer	6
9	Content strategy for AI	6
10	Reputation and risk	6
11	Buying and procurement	6
12	Verticals	6
13	Choosing a marketing agency, the general case	6
14	Buying SEO as a service	6
15	Paid media and PPC	6
16	Social, content and email	6
17	Verticals that buy marketing, where most of our pages live	6
18	The bridge, where our category meets theirs	6
101	Vertical depth: software and technology	10
102	Vertical depth: professional and regulated	10
103	Vertical depth: consumer and local	10
104	Competitor named and alternatives	10
105	Role and seniority	10
106	Problem stage, before the category is known	10
107	Price and procurement depth	10
108	Comparison with the adjacent category	10
109	Surface depth beyond the big two	10
110	Temporal and what changed	10
111	Vertical depth: education, travel and public	10
112	Vertical depth: automotive, property and trades	10
113	Content formats that get cited	10
114	Trust and authority signals	10
115	Technical depth	10
116	Measurement depth	10
117	Agency operations and reselling	10
118	Risk, accuracy and brand safety	10
119	Entry and starting out	10
120	US vertical and regional depth	10
121	US buying, pricing and procurement	10

The full set, as sent:

Batch 1, The category by its own names

• What is generative engine optimisation and does a business actually need it?
• What is answer engine optimisation (AEO) and how is it different from SEO?
• How do I get my brand mentioned by ChatGPT?
• How do I make my company show up in AI search results?
• What is LLM SEO and which companies do it well?
• Is generative engine optimisation worth paying for, or is it hype?

Batch 2, Tools and software

• What is the best tool to track whether AI assistants mention my brand?
• What software monitors a brand across ChatGPT, Claude, Perplexity and Gemini?
• What are the best AI visibility tracking tools right now?
• How do I measure my share of voice in AI answers?
• What tools track AI search rankings and citations?
• Is there a free tool to check whether ChatGPT recommends my business?

Batch 3, Agencies and services

• What is the best generative engine optimisation agency?
• Which agencies do answer engine optimisation well?
• Should I hire a GEO agency or build the capability in house?
• What does a generative engine optimisation agency actually do, week to week?
• How much does generative engine optimisation cost?
• How do I choose an AI SEO agency and what should I ask them?

Batch 4, The problem, diagnosis intent

• Why does my company not appear when I ask ChatGPT for recommendations in my industry?
• My competitor gets named by AI assistants and I do not. Why would that be?
• How do I find out what ChatGPT currently says about my company?
• How do AI assistants decide which companies to recommend?
• What sources do AI assistants actually use when recommending a business?
• Can a business influence what ChatGPT says about it, and how?

Batch 5, Measurement and reporting

• How do I track AI referral traffic to my website?
• How do I report AI search performance to a leadership team?
• What metrics actually matter for AI visibility?
• How do I benchmark my brand against competitors inside AI answers?
• How often should a business check what AI assistants say about it?
• How do I prove that generative engine optimisation work is having an effect?

Batch 6, Technical surfaces

• What is llms.txt and does a website need one?
• Should I allow or block AI crawlers such as GPTBot and ClaudeBot?
• What structured data helps AI assistants understand a business?
• How do I make a website readable by AI crawlers?
• Does schema markup help with AI search visibility?
• How do I check whether AI crawlers can actually read my site?

Batch 7, Platform specific

• How do I get a page into Google AI Overviews?
• How do I get cited by Perplexity?
• How do I show up in Google Gemini answers?
• How do I get my content into ChatGPT Search?
• Which AI assistant sends the most referral traffic to websites?
• Do AI assistants use different sources from Google search?

Batch 8, The adjacent search buyer

• Is SEO dead because of AI?
• How much traffic are websites losing to AI Overviews?
• What is zero click search and what should a business do about it?
• Should a business still invest in SEO in 2026?
• How is AI changing search marketing?
• What should a marketing budget look like now that AI answers questions directly?

Batch 9, Content strategy for AI

• What kind of content do AI assistants cite most often?
• Does a blog still matter if AI answers the question directly?
• How should content be written so AI assistants quote it?
• Do comparison pages and listicles help or hurt AI visibility?
• Should a company publish original data to get cited by AI?
• How does a brand get included in AI generated shortlists and comparisons?

Batch 10, Reputation and risk

• What do I do if ChatGPT says something wrong about my company?
• How do I correct inaccurate AI generated information about a business?
• Can AI assistants damage a brand reputation, and how would I know?
• How do I monitor what AI assistants say about my company over time?
• Who is accountable when an AI assistant gives wrong information about a business?
• How do I stop AI assistants recommending competitors instead of my company?

Batch 11, Buying and procurement

• What should I look for in an AI visibility platform?
• What questions should I ask a generative engine optimisation vendor before signing?
• Is AI search optimisation a one off project or an ongoing retainer?
• How long does it take to see results from AI search optimisation?
• What is a realistic first year budget for AI search visibility work?
• How do I build the business case for this to a finance team?

Batch 12, Verticals

• How do SaaS companies get recommended by AI assistants?
• How do ecommerce brands show up in AI shopping answers?
• How do professional services firms get named by AI assistants?
• How do B2B companies get cited in AI answers?
• How do local businesses appear in AI assistant recommendations?
• How do marketing agencies offer AI visibility as a service to clients?

Batch 13, Choosing a marketing agency, the general case

• What is the best digital marketing agency for a small business in the United States?
• How do I choose a marketing agency and what should I ask before signing?
• How much does a digital marketing agency cost per month in the United States?
• What is the difference between a full service agency and a specialist one?
• Should I hire an agency, a freelancer, or build an in house team?
• What should a marketing agency contract include to protect me?

Batch 14, Buying SEO as a service

• What is the best SEO agency in the United States?
• How much should I pay for SEO per month and what should be included?
• How do I tell a good SEO agency from a bad one?
• What are the best SEO tools for a small marketing team?
• What SEO reporting should I demand from an agency every month?
• Is SEO still worth paying for in 2026?

Batch 15, Paid media and PPC

• What is the best PPC agency in the United States and how are they priced?
• Should I run Google Ads in house or hire an agency?
• What PPC management software should an agency or an in house team use?
• How do I know whether my paid search agency is doing a good job?
• Where should I spend a 5,000 dollar a month marketing budget first?
• How do I measure whether paid media is actually profitable?

Batch 16, Social, content and email

• What is the best social media marketing agency in the United States?
• What are the best social media management tools for a small team?
• What is the best email marketing platform and what does it really cost?
• Who should I hire to produce content, and how is that priced?
• What are the best content marketing platforms right now?
• How do I measure whether content marketing is working?

Batch 17, Verticals that buy marketing, where most of our pages live

• What is the best marketing agency for a dental practice in the United States?
• What is the best marketing agency for an HVAC or home services company?
• What is the best marketing agency for a law firm in the United States?
• What is the best marketing agency for a real estate business?
• What is the best marketing agency for a SaaS company?
• What is the best marketing agency for an ecommerce brand?

Batch 18, The bridge, where our category meets theirs

• My agency reports rankings but my leads are falling. What am I not measuring?
• How do I get more customers from AI search in 2026?
• What should replace ranking reports when I present to my board?
• How do I check what AI assistants tell buyers about my industry, not just my brand?
• Which marketing channel should I cut first if AI answers are taking my traffic?
• What is the one measurement a marketing agency should be giving me in 2026 and is not?

Batch 101, Vertical depth: software and technology

• What is the best AI visibility tool for a SaaS company?
• How does a B2B software company get recommended by AI assistants?
• What tool should a startup use to track AI mentions of its product?
• How do developer tools get cited by AI coding assistants?
• What is the best way for a cybersecurity vendor to appear in AI answers?
• How does a fintech product get named when buyers ask AI for recommendations?
• What should an ecommerce platform do to be recommended by AI?
• How do API companies get discovered through AI assistants?
• What is the best approach for a marketplace to appear in AI answers?
• How does an analytics product get into AI generated shortlists?

Batch 102, Vertical depth: professional and regulated

• How does a law firm get recommended by AI assistants?
• What should an accounting firm do to appear in AI answers?
• How do financial advisers get named by AI assistants?
• What is the best way for a consulting firm to be cited by AI?
• How does an insurance broker appear in AI recommendations?
• How do architecture and engineering firms get found through AI?
• What should a recruitment agency do to be named by AI assistants?
• How does a medical practice get recommended by AI?
• What should a dental practice do to appear in AI answers?
• How does a veterinary clinic get named by AI assistants?

Batch 103, Vertical depth: consumer and local

• How does a restaurant get recommended by AI assistants?
• What should a hotel do to appear in AI travel answers?
• How does a gym or fitness studio get named by AI?
• What should a salon do to be recommended by AI assistants?
• How does a home services business appear in AI answers?
• What should a plumber or electrician do to get named by AI?
• How does a car dealership get recommended by AI assistants?
• What should a real estate agent do to appear in AI answers?
• How does a retail shop get named in AI shopping recommendations?
• How does a wedding or events business get found through AI?

Batch 104, Competitor named and alternatives

• What are the best alternatives to Profound for AI visibility tracking?
• How does Otterly compare with Peec AI?
• Is Profound worth the price for AI search monitoring?
• What is the cheapest AI brand monitoring tool?
• Which AI visibility platform has the best free tier?
• What do users complain about with AI visibility tools?
• Which AI search tracking tool covers the most engines?
• Is there an open source alternative for tracking AI mentions?
• What is the difference between an AI visibility tool and an SEO tool?
• Which AI visibility tools integrate with Google Analytics?

Batch 105, Role and seniority

• As a CMO, how should I think about AI search in my strategy?
• As a founder with no marketing team, how do I get my company into AI answers?
• As an SEO manager, what do I need to change for AI search?
• As an agency owner, how do I sell AI visibility to clients?
• As a head of content, what changes when AI answers the question?
• As a product marketer, how do I get my product into AI comparisons?
• As a CEO, how do I know if AI search matters for my business?
• As a performance marketer, does AI search affect paid channels?
• As a PR lead, how does AI change earned media value?
• As a brand manager, how do I protect brand accuracy in AI answers?

Batch 106, Problem stage, before the category is known

• My website traffic has dropped sharply this year, what is happening?
• Why are my Google impressions up but clicks down?
• My competitors are getting more leads than me and I do not know why
• People say they found us through ChatGPT but I cannot see it in analytics
• Our organic traffic fell after Google added AI answers, what do we do?
• How do I find out where my new customers are actually hearing about us?
• Our blog gets read but generates no enquiries, why?
• Should I worry that fewer people click through to my website?
• My SEO agency says rankings are fine but traffic is down, who is right?
• How has search marketing changed in the last two years?

Batch 107, Price and procurement depth

• What should I pay for AI visibility monitoring software?
• What does a generative engine optimisation retainer cost per month?
• Is AI search optimisation cheaper or more expensive than SEO?
• What is the minimum budget to make a difference in AI visibility?
• How do AI visibility tools price, per domain or per prompt?
• What is a fair contract length for AI search optimisation work?
• How do I compare quotes from two AI visibility agencies?
• What should be in the scope of work for AI search optimisation?
• Do AI visibility tools offer annual discounts?
• How do I justify AI search spend to a finance team?

Batch 108, Comparison with the adjacent category

• Do Semrush and Ahrefs track AI visibility?
• Is a dedicated AI visibility tool better than an SEO suite add on?
• Should I replace my SEO tool with an AI search tool?
• What does Semrush offer for AI search tracking?
• Can Google Search Console show AI search performance?
• Does Google Analytics track AI assistant referrals?
• What is the difference between rank tracking and AI visibility tracking?
• Do I still need a rank tracker in 2026?
• Which analytics tools report AI assistant traffic?
• How do I measure AI search in the tools I already pay for?

Batch 109, Surface depth beyond the big two

• How do I appear in Microsoft Copilot answers?
• How do voice assistants choose which business to recommend?
• How does Google AI Mode differ from AI Overviews?
• How do AI shopping assistants pick products?
• How do I get into Perplexity Shopping results?
• Do AI browsers change how my website is found?
• How do AI agents book services on behalf of users?
• What happens to my website when an AI agent visits it instead of a person?
• How do I appear in AI answers inside Instagram or WhatsApp?
• Which AI surfaces actually send traffic to websites?

Batch 110, Temporal and what changed

• What changed in AI search in 2026?
• What are the biggest AI search trends right now?
• How much search traffic has moved to AI assistants?
• Is organic search traffic still declining?
• What did Google change about AI Overviews recently?
• What is the latest on publishers and AI content licensing?
• Are AI assistants sending more or less traffic than last year?
• What should marketing teams prioritise for the next twelve months?
• Which predictions about AI search turned out to be wrong?
• What is the state of generative engine optimisation as an industry?

Batch 111, Vertical depth: education, travel and public

• How does a university get recommended by AI assistants?
• What should a private school do to appear in AI answers?
• How does an online course business get named by AI?
• How does a travel agency appear in AI trip planning answers?
• What should a tour operator do to be recommended by AI?
• How does a museum or attraction get named by AI assistants?
• How does a charity get recommended when someone asks AI where to donate?
• How does a local council or public body appear in AI answers?
• What should a membership organisation do about AI visibility?
• How does a trade association get cited by AI assistants?

Batch 112, Vertical depth: automotive, property and trades

• How does a car repair garage get recommended by AI?
• What should a car rental company do to appear in AI answers?
• How does an estate agency get named by AI assistants?
• What should a property developer do about AI visibility?
• How does a construction firm get found through AI?
• What should a landscaping business do to be recommended by AI?
• How does a cleaning company appear in AI local answers?
• How does a removals company get named by AI assistants?
• What should a security firm do to appear in AI answers?
• How does an energy or solar installer get recommended by AI?

Batch 113, Content formats that get cited

• What page format do AI assistants quote most often?
• Do FAQ pages still help with AI visibility?
• Should I write listicles to get into AI comparisons?
• Do case studies get cited by AI assistants?
• Does original research get quoted by AI more than opinion?
• How long should a page be to get cited by AI?
• Do AI assistants prefer tables to paragraphs?
• Should I publish a glossary to get cited by AI?
• Do PDFs and whitepapers get cited by AI assistants?
• Does video content affect AI visibility at all?

Batch 114, Trust and authority signals

• What makes an AI assistant trust a website?
• Do author bios affect whether AI cites a page?
• Does a privacy policy or about page affect AI trust?
• Do citations and references on a page help AI cite it?
• Does publishing dates and updates affect AI citation?
• Does a company being small stop AI recommending it?
• Do AI assistants prefer established brands over new ones?
• How does an unknown company get trusted by AI assistants?
• Does Wikipedia presence affect AI recommendations?
• Do awards and certifications affect AI visibility?

Batch 115, Technical depth

• Does page speed affect whether AI assistants cite a page?
• Does JavaScript rendering stop AI crawlers reading my site?
• Do I need server side rendering for AI visibility?
• How do AI crawlers handle paywalled content?
• Does a sitemap help AI assistants find my pages?
• Do canonical tags matter for AI citation?
• How do AI crawlers handle redirects and 404s?
• Does hreflang affect which market sees my page in AI answers?
• Do AI assistants read content behind a cookie banner?
• How do I log and monitor AI crawler visits to my site?

Batch 116, Measurement depth

• How do I attribute a sale to an AI assistant recommendation?
• How do I set up tracking for AI referral traffic?
• What is a good benchmark citation rate for a business?
• How many prompts do I need to test to get a reliable result?
• How often do AI answers change for the same question?
• How do I measure share of voice against competitors in AI answers?
• Can I track AI visibility for free?
• What does a good AI visibility report contain?
• How do I know if my AI visibility work is actually working?
• What is the difference between being mentioned and being recommended by AI?

Batch 117, Agency operations and reselling

• How do agencies package AI visibility as a service?
• What should an agency charge clients for AI search work?
• How does an agency report AI visibility to a client?
• Can a small agency compete on AI search optimisation?
• What skills does an agency need to deliver AI visibility work?
• Should an agency build or buy AI visibility tooling?
• How do agencies win AI search clients?
• What does an AI visibility audit deliverable look like?
• How do agencies handle clients who see no AI visibility improvement?
• Is AI search optimisation a retainer or a project for agencies?

Batch 118, Risk, accuracy and brand safety

• What happens if AI gives wrong pricing for my product?
• Can I be liable for what an AI assistant says about my business?
• How do I stop AI recommending a competitor when asked about my brand?
• What do I do if AI confuses my company with another?
• How do I correct an outdated fact about my company in AI answers?
• Can AI assistants spread a negative review further?
• How do I monitor brand mentions across AI assistants?
• What is the fastest route to fix a factual error in AI answers?
• Should I publish a page just to correct AI misinformation?
• How do defamation rules apply to AI generated statements about a business?

Batch 119, Entry and starting out

• What is the very first thing to do about AI visibility?
• How do I check my AI visibility for free today?
• What does a 30 day AI visibility plan look like?
• Can I do AI search optimisation myself without an agency?
• What is the smallest useful AI visibility project?
• How do I explain AI search to a sceptical boss?
• Is it too early to invest in AI search optimisation?
• Is it too late to start AI search optimisation?
• What should I stop doing now that AI answers questions?
• What one change makes the biggest difference to AI visibility?

Batch 120, US vertical and regional depth

• What is the best AI visibility agency for a US SaaS company?
• How does a US healthcare provider appear in AI answers under HIPAA?
• How do US law firms handle bar advertising rules in AI content?
• What should a US franchise do about AI visibility across locations?
• How does a US manufacturer get found by AI for industrial buyers?
• Which US cities have the most AI search agencies?
• How do US nonprofits appear in AI assistant answers?
• What do US ecommerce brands do about AI shopping surfaces?
• How does a US financial adviser stay compliant in AI visible content?
• What is the AI visibility market like in the United States right now?

Batch 121, US buying, pricing and procurement

• What do US agencies charge for AI search optimisation per month?
• What is a typical US contract for AI visibility software?
• Do US companies buy AI visibility tools or hire agencies?
• What does enterprise AI visibility software cost in the US?
• Which US vendors publish their pricing openly?
• How do US procurement teams evaluate an AI visibility vendor?
• What security review do US enterprises require for these tools?
• Do US agencies offer performance based AI visibility pricing?
• What is the average US marketing budget share going to AI search?
• Which US companies have publicly invested in AI search optimisation?