Question 1

What is answer engine optimisation (AEO)?

Accepted Answer

Answer engine optimisation is the practice of getting your brand returned as the answer when someone asks a question of an AI system such as ChatGPT, Perplexity or Google AI Overviews, rather than ranking a page in a list of blue links. Classic SEO competes for position on a results page. AEO competes to be the source the answer is built from, which means being the clearest, most specific and most quotable answer to a real question.

Question 2

How do I get my brand cited by ChatGPT or Perplexity?

Accepted Answer

Answer specific questions properly, in public, in a form a machine can extract. Popularity is close to irrelevant: in Otterly.AI's 2026 study of more than 100 million AI citation instances, the correlation between citation frequency and view count was -0.03, and 40.83% of cited videos had fewer than a thousand views. What earns citations is clarity, specificity and original material a language model could not have generated on its own — your own data, your own client results, your own judgement.

Question 3

Which platforms do AI engines cite most?

Accepted Answer

Brand-owned domains still dominate at 52.2% of all AI citations. Social and video platforms together account for around 5.54%. Within that social slice, Reddit takes 46.4% and YouTube 31.8% — between them 78.2% of all social citations in AI search. The practical read is that your own site is still the single most important asset, and that Reddit and YouTube are where the social opportunity sits.

Question 4

Do AI search engines read video?

Accepted Answer

Yes, but they read it rather than watch it. AI systems ingest the transcript, the description, the chapter titles and the surrounding metadata. That is why long-form video accounted for 94% of YouTube citations against 5.7% for Shorts, and why 78% of videos carrying chapter timestamps were cited more than once, usually across two to five different chapters. A well-structured video behaves like an article with headings.

Question 5

Does my brand need an llms.txt file?

Accepted Answer

Probably not yet, and nobody should be selling it to you as essential. llms.txt is a proposed standard for telling AI crawlers how to use your content, but as of 2026 there is no clear evidence that major AI systems consistently honour it, and support across providers remains incomplete. It costs almost nothing to add, so add it if you like — but do not build a visibility strategy on it, and treat anyone who guarantees results from it with caution.

Question 6

What is social search?

Accepted Answer

Social search is what happens when buyers use a social platform as a search engine — typing a question into TikTok, Instagram, Pinterest, Reddit or YouTube instead of Google, and choosing from whoever answers. It matters because the answer usually comes from a person rather than a brand, which means the brands that appear are the ones whose creators answered the question.

Question 7

Why are buyers searching on social instead of Google?

Accepted Answer

Because for a whole class of question, a person who has actually used the thing is a better answer than a page written by the company selling it. Questions about whether something works, what it is like in practice, and who it is not right for are answered more convincingly by a creator on camera than by a product page. Buyers have not stopped using Google — they use both, often in the same week and for the same purchase.

Question 8

Does FAQ schema still work in 2026?

Accepted Answer

Not for Google rich results. Google removed FAQ rich results from Search on 7 May 2026, dropped the Rich Results Test and reporting support in June, and ends Search Console API support in August 2026. Existing FAQ markup is harmless and can stay, but it will not produce a visible Google feature. Structured question-and-answer content is still worth writing — for readers, and because it is the form AI systems extract most easily — but anyone still selling FAQ schema as a Google ranking play is selling you an obsolete tactic.

Question 9

What is a zero-click world?

Accepted Answer

It describes search behaviour where the buyer gets their answer without ever clicking through to a website — from an AI summary, a featured snippet or a creator's video. Combined with the finding that roughly 54% of Google searches never mention a brand at all, it means a large share of category demand is settled before anyone visits your site. Whoever supplied that answer influenced the purchase.

Question 10

What is Share of Search?

Accepted Answer

Share of Search is your brand's share of all the branded search queries in your category. Les Binet's analysis of the UK car, energy and mobile markets found it led market share by roughly six to twelve months, with the longer leads in categories that have longer purchase cycles. It is the most reliable leading indicator most marketers have access to, and Google Trends gives you a directional version free.

Question 11

What is Answerz Share and how is it different?

Accepted Answer

Answerz Share is the proportion of your buyers' questions where your brand comes back as the answer, measured across AI answers, Search, Social, Retail and Video. The difference is direction: Share of Search measures demand — how many people are already looking for you by name — while Answerz Share measures presence, whether you are there when somebody asks a question that mentions no brand at all. One is an outcome. The other is the input that moves it.

Question 12

How do I find out if my brand appears in AI answers?

Accepted Answer

Start by testing the questions your buyers actually ask, not your brand name. Take the ten questions your sales team fields most often, run each one through Google, an AI assistant and the social platform your category lives on, and record whether you appear and who does instead. That is the manual version of an Answerz Share baseline, and it takes an afternoon.

Question 13

Can creators improve my AI visibility?

Accepted Answer

Yes, and through a route most brands are not accounting for. AI crawlers filter out standard advertising units, so a pre-roll ad has essentially no influence on what an AI system says about your product. What these systems read is the organic narrative in the transcript — the words a creator actually speaks. Because a paid partnership does not necessarily surface in a video's metadata, sponsored content mentioning your brand can be read and cited as ordinary organic information.

Question 14

Should I use AI-generated video to scale this?

Accepted Answer

We would advise against it. YouTube renamed its “repetitious content” policy to “inauthentic content” on 15 July 2025, explicitly targeting mass-produced and templated material, and channels have been demonetised under it since. The deeper problem is that what makes content citable is containing something a language model could not have produced on its own. Generated video is a model writing for another model to read, with nothing in the middle that was not already in the training data.

Question 15

How many questions does a brand need to answer?

Accepted Answer

Fewer than most people expect, chosen better than most people choose them. A category typically has a few dozen questions that carry real purchase intent, and the same handful recur across every buyer. The work is identifying which ones you are currently absent from, then answering those properly — not publishing volume. Google has said around 15% of the searches it sees each day have never been asked before, so no brand can answer everything; the discipline is answering the questions closest to a purchase.

The questions we get asked.

AI answers and citations