Skip to content

How to rank in AI search and answer engines: 7 tactics that work

Author: Matteo Pellegrini

How to rank on answer engines

Here is how to rank on answer engines (our advice comes from the hundreds of campaigns we have worked on):

1. Understand how answer engines work

Search engine optimisation (SEO) does not start with optimising a website but with understanding how search engines work; the same applies to answer engine optimisation (AEO).

So how do answer engines work? In short, they generate answers from data:

Source of the answerDetails
Training dataTraining data is the knowledge base an answer engine such as ChatGPT or Claude was trained on. It can include third-party data, data taken from the internet (up to a specific date) and internal data.
Internet dataInternet data is a real-time stream of information, such as results from the Google or Bing indexes. Most answer engines need search to be switched on to use it.

Some answer engines combine or extend these options, for example:

  • Perplexity, which limits its internet data to authoritative, trustworthy domains.
  • AI Overviews, which uses Google's index to generate summaries in Google search results.
  • Apple Intelligence, which draws on the Apple and Google ecosystems to generate results.

Bear in mind that no answer engine gives the same answer twice. You and a friend could ask an AI the same question and get the same general answer, but the way it is presented and worded will be different. The same happens if you ask two different answer engines the same question, as in the example below with Claude and Gemini.

same query answered differently by Claude and Gemini

Because users can personalise their AI experience, with personal information (dietary preferences, for example) and answer preferences (short or longer), answers can become even more individual, and so can the brands they mention.

For marketers, this means optimising for access to both training data and internet data across all engines. It also means understanding where answer engines get their internet data (hint: from the biggest search engines).

Answer engineInternet data source
ChatGPTBing
Meta AIBing
GeminiGoogle
AI OverviewsGoogle
Apple IntelligenceGoogle
PerplexityBing
Microsoft CopilotBing

Read more:

2. Optimise for search engines

As well as using internet data to generate answers, answer engines use past training data, including historical information from the internet. Businesses that have been doing SEO for years are likely to find themselves mentioned in answers generated from the knowledge base. This is why omnichannel SEO matters: optimisation work extends from traditional search engines to AI platforms to build a complete digital presence.

example Claude query with a brand mention

Some marketers think SEO is dead (in response to Gartner's prediction that businesses will lose 25% of their organic traffic by 2026), but it is not. Experiments show that websites ranking well in traditional search engines are more likely to appear in AI answers.

  • 73% is the overlap between Bing search results and ChatGPT search answers.
  • 61% is the overlap between AI Overviews and the first page of Google Search results.

That is why businesses keep investing in traditional search engine optimisation. It provides the foundation for a solid optimisation programme, even if marketers have to explain to leadership how the value of SEO has changed.

So what does search engine optimisation include when you are thinking about AEO? The essentials, such as:

  • Making the website discoverable and crawlable by search engines and answer engines
  • Creating a smooth user experience with intuitive layouts and fast-loading pages
  • Meeting user needs with well-written, well-presented content
  • Attracting mentions and backlinks from trustworthy sites relevant to your sector.

You can read about all these essentials in our beginner's guide to SEO.

3. Target conversational phrases

Answer engines encourage people to search differently. Because they can give personalised answers based on dietary preferences, travel plans, company size and more, they invite users to have a conversation with them.

Search engine queryAnswer engine query
autumn family holiday ideas lake districtI'm looking for ideas for an autumn family holiday somewhere in the Lake District
gluten free restaurants near meWhat are some good restaurants with gluten-free options in my area?
freeagent alternativesMy cleaning company needs new accounting software. We're using FreeAgent. What are some good alternatives in the same price range?

Much like the "People also ask" box in Google Search, answer engines prompt users to add more information. In the gluten-free restaurant example above, the answer engine might ask which cuisine the user prefers, such as Indian, Italian or Thai.

The next section covers how to produce content that meets these specific needs, but the point here is to think about conversational phrases to target alongside traditional keywords.

Talking to customer-facing colleagues, such as customer support, sales or service providers, tells you what people ask about your solution or the problem you solve.

4. Produce excellent content

Excellent content is what ranks in answer engines such as AI Mode, ChatGPT, Perplexity and Gemini. Businesses cannot rely on AI-generated or "good enough" content to succeed in answer engines over the long term, where the same extraction criteria apply as for featured snippets.

This is because the AI models behind these answer engines assess:

ElementExample
Data-backed details"This restaurant has a 4.5 rating, based on more than 200 reviews over the last 8 months."
Specific recommendations"Their onion, mushroom and pepper pizza gets the balance of vegetables and cheese just right."
Well-organised content"Make this pizza at home with these ingredients: pizza dough, grated mozzarella, onions..."
Easy-to-read contentReadability score: 65
Relevant structured dataRecipe schema markup
Trustworthy domainDomain Rating (DR): 80

Producing content at this level takes a few key steps:

  1. Research: copying competitors' content from the search results is not enough for answer engines. Businesses need to do more research to give their content more context and originality, for example through anecdotes, studies and so on.
  2. Editing: correcting grammar, spelling and branding errors is essential. But editors now also have to prioritise the content's user experience. How can writers use headings, lists, tables and more to make the content easy to use (for people and AI)?
  3. Repurposing: with answer engines, the user journey is increasingly omnichannel, from chatting with Meta AI on social media to reading AI Overviews in search. Repurposing content for different channels helps organisations create touchpoints along these journeys.
  4. Promotion: adding insights from internal studies or first-hand experience makes content much easier to promote (and to earn mentions and backlinks). Outreach, paid ads and repurposing strong content all help here.
  5. Analysis: most businesses already measure content performance, but now that measurement has to follow user journeys that include answer engines. For example, brands can run manual searches to track their AI visibility.

For more on producing excellent content, see our SEO content guide.

5. Build assets

As well as good written content, businesses that want to rank in generative engines such as ChatGPT need other types of content (especially if they want to compete with zero-click searches). These assets can range from visuals, such as graphics, to tools, such as calculators.

This is because answer engines share images, videos and interactive options in their answers.

videos in Gemini AI answers

For video, YouTube is the most cited site, especially when users are looking for instructions, demonstrations or examples. Since August 2024, YouTube citations in Google's AI Overviews have grown by more than 300%, which shows how much weight this content carries.

There are several ways to produce other content, including:

  • In-house team members
  • Freelancers
  • AI tools such as Canva, OpenAI's Sora or Cursor

To get the most from this work, focus on what would help the user most. For example, do not create an image just for the sake of it. Create one to explain a topic, such as cell division, that often confuses people without a visual explanation.

6. Get mentioned on third-party sites

Answer engines use a range of sources from their training data and the internet to generate answers, which is why being mentioned on third-party sites is an effective way to appear in AI-generated answers.

Think about questions asking for the "best" provider of something. Answer engines are more likely to refer to third-party review sites than directly to brands, because those sites include ratings, reviews, rankings and more.

ChatGPT answer to a "best" query

In the example above, ChatGPT cites several websites. The sites mentioned often do not sell the product being searched for; they list possible options and explain why each one is worth considering.

You can appear on third-party sites in several ways, including:

  • Claiming your profiles on Google Business Profile and Bing Places
  • Contacting publishers to offer experiences or free samples
  • Producing excellent content (for example with internal data) that gets you mentioned.

7. Prioritise review management

Reviews also matter for ranking in AI engines: nobody is looking for a mediocre product or service. People want well-rated products, whatever the price. So businesses need to prioritise review management.

Review management takes a few steps:

  1. Create profiles on sites such as Google Business Profile, G2 and Bing Places
  2. Collect reviews, for example through incentive offers*, business cards and emails
  3. Manage reviews, for example by replying to them.

Building an authoritative reputation across websites and listings is where this strategy pays off most, because it gives answer engines more sources pointing to a positive experience with your brand, its support and its offers.

* Sites such as Google Business Profile prohibit incentivised reviews, so check the platform's policies before sending offers.

More tips:

How to track answer engine rankings

Answer engines such as ChatGPT, Gemini and Claude are starting to cite your brand. How do you track those citations (and share them with the team for analysis)? For most businesses, the answer is a Google Sheets or Microsoft Excel file.

You can also use dedicated AI Overviews tracking tools.

Improve your ranking on the most popular answer engines

You now know how to rank on answer engines such as ChatGPT, Perplexity AI, Claude and others. The next step is to start optimising your presence and measuring what it brings: brand awareness, referral traffic and, eventually, sales.

Need professional help? Contact us to find out how our AI-driven SEO services can help.

Matteo Pellegrini

Matteo Pellegrini

I’m a Business Developer, and at Visilay I focus on developing data-driven SEO, Google Ads, and CRO strategies. I love historical museums, have been practicing Karate for as long as I can remember, and on weekends I enjoy exploring Italian villages in search of authentic local food.