Generative Engine Optimization for insight agencies: Top tips to get found by AI
AI is changing how we search. If you want an answer to a burning question, holiday inspiration, or to find a recipe, chances are you’re turning to ChatGPT, Claude, or the search engine’s AI for a curated answer.
As these LLMs increasingly become the first stop for research, insight agencies, like many other businesses, face a new online visibility challenge. It’s not enough to rank on page one of Google, you need your thinking to be the one that AI cites, quotes and recommends. In this article, I share some top tips for how to get there.
What is Generative Engine Optimization (GEO)?
GEO is the practice of structuring your content so that AI-powered search engines and large language models surface, cite and recommend it. Unlike traditional SEO which optimises for crawlers and keyword ranking, GEO optimises for comprehension, authority and trustworthiness in the eyes of AI systems trained to synthesise the best available knowledge.
For market research and insight agencies, this is a significant opportunity because you produce exactly the kind of expert, evidence-based content that AI systems are designed to surface.
Here are my 10 top tips:
1. Lead with the answer
AI systems favour content that gets to the point immediately. Structure every article, report and thought leadership piece so the core insight appears in the first paragraph rather than buried in a conclusion. Think top-down research reporting style.
Start doing this: Begin with a clear, declarative statement, then go on to build the argument beneath it.
2. Use Clear, Quotable Language
AI models extract and paraphrase content more readily when it is written in plain, direct English. So avoid the jargon so that models can confidently cite you.
Start doing this: Write in short, clear sentences and make your key points so that they can stand on their own. Think also about the use of acronyms – write them out in full so they can be understood.
3. Structure content with clear headlines
AI systems use headings to understand what a piece of content is about. Well-structured content with descriptive headings are more likely to be surfaced for specific queries.
Do this: Use headings for your article that answer questions directly (e.g. "Why traditional brand trackers are failing" rather than "The problem").
4. Build genuine topical authority
AI models prioritise sources that cover a topic consistently and in depth. A single viral post won’t build lasting GEO visibility, but a body of interconnected, expert content will.
Do this: Develop a clear content territory for your agency and publish consistently within it.
5. Ground claims in data and evidence
Content that backs up claims with specific statistics, proprietary data, or expert quotes is significantly more likely to be cited.
Do this: Publish your own primary data wherever possible. Avoid vague claims like "many consumers say..." and replace them with specific results.
6. Optimise for question-based queries
A large proportion of AI search queries are phrased as questions so content that directly answers specific questions is far more likely to be picked up.
Do this: Include an FAQ section at the bottom of key articles. Write perspectives framed as answers to questions your clients are likely to want to ask.
7. Ensure your content can be found by LLMs
If your best thinking is locked behind gated PDFs or password-protected portals, AI systems simply will not find it.
Do this: Publish a publicly accessible version of your key thought leadership and ensure your website renders content.
8. Build Citation-Worthy Credentials
If other reputable sites reference and link to your work, AI models are more likely to treat you as an authoritative source.
Do this: Pitch findings to trade press (Research Live, ESOMAR, Campaign). Speak at conferences and publish your presentations online. Regularly share content and perspectives on Linked In – and have your experts write on their profiles. Linked In is one of the most cited sources for LLMS.
9. Keep Content Fresh
AI systems, particularly those with live web access, favour current content.
Do this: It helps to add a "Last updated" date to evergreen pieces and refresh the data annually if you can.
10. Use Structured Data (Schema Markup)
Schema markup tells search engines exactly what type of content a page contains, helping AI systems categorise and surface your work more accurately.
Do this: Make sure you have implemented Article, FAQ Page, Person and Organization schema across your site.
The Bigger Picture
GEO is not a replacement for good thinking - it is the discipline of making good thinking findable. For insight agencies, GEO is less a technical challenge than a communication one: say what you know, say it clearly, say it consistently, and say it where AI systems can find it.
The agencies that invest in this now will own the category when a client asks an AI who the best insight agencies are and gets a confident, well-cited answer.
Want to audit your agency's current GEO visibility? Start by asking your favourite LLM a question your ideal client would ask and see who comes up. Then ask why it is not you.