A practical guide to answer engine optimization (AEO) and generative engine optimization (GEO): what research says, and the checklist we use.
Search is splitting in two. People still type queries into Google, but more of them now ask ChatGPT, Gemini, Claude or Perplexity for a recommendation and act on the answer. If your brand is not in that answer, you are invisible at the moment of decision.
Two terms describe the work. Answer engine optimization (AEO) is making your brand the source an AI answer draws on. Generative engine optimization (GEO) is the wider discipline of earning citations, mentions and recommendations from generative systems. A June 2026 research paper on measuring brand visibility treats AEO as a part of GEO, and we do the same here.
Why this matters now
Pew Research Center tracked the browsing of 900 US adults in March 2025. When a Google search page showed an AI summary, users clicked a traditional result link in 8% of visits. When no summary appeared, they clicked in 15% of visits. Few clicked links inside the summary itself. Around 18% of the searches in the study produced an AI summary.
That was one market and one month, and the share of searches with summaries has changed since. The direction is what counts: more questions are answered on the results page or inside a chatbot, and fewer ever reach your website.
What the research says
The June 2026 arXiv paper "Generative Engine Optimization at Scale" is one of the first large attempts at a baseline. Its main conclusions are modest and useful. Brand visibility in AI engines can be measured, it differs from one platform to another, and it varies strongly with how established a brand is. The authors also propose follow-up experiments to test which tactics actually cause gains, which is a polite way of saying nobody should claim guaranteed results yet.
Ogilvy's Australia and New Zealand arm has pointed to the same two levers in its Generative Impact offer: earned media and reputation on one side, and a website structured so models can read, rank and recommend its content on the other.
The checklist we use
1. Be unambiguous about who you are. Use one consistent name, description and set of facts across your site, LinkedIn, Google Business Profile and directories. AI systems resolve entities, and inconsistent facts lower confidence.
2. Answer questions directly. Write pages that open with a plain, quotable answer to a real buyer question, then add detail. Short definitions, comparison tables and FAQs are easy for a model to lift.
3. Add structured data. Use schema.org markup for your organisation, services, articles and FAQs, so machines do not have to guess what a page is about.
4. Let the crawlers in. Many sites block AI crawlers by accident. Check that your robots.txt allows the search and retrieval bots you want, such as OAI-SearchBot, ClaudeBot, PerplexityBot and Google-Extended, and that your pages are fast and readable without JavaScript.
5. Earn mentions elsewhere. Models lean on third-party sources. Press coverage, expert quotes, credible reviews and industry lists matter as much as your own copy. This is where PR and SEO meet.
6. Publish original material. Data, case studies and clear explanations that other sites cite give models something worth citing. Rewriting what everyone else says does not.
7. Offer a machine-readable summary. An llms.txt file that points to your key pages is cheap to add. To be straight about it, no major AI provider has confirmed that it relies on the file, so treat it as a low-cost extra rather than a ranking lever.
8. Measure it. Keep a fixed set of prompts that mirror how your customers ask, run them across the main assistants each month, and record whether you are mentioned, cited, or recommended, and in what position. Track the trend, not single answers, because outputs vary between runs.
What to expect
AEO and GEO build on good SEO. (Our SEO, AEO and GEO service covers both.) They do not replace it, and they reward the same long-term habits: clear content, trusted sources, a fast site. The gains compound slowly, and anyone promising a ranking inside AI answers by next week is guessing.
For Indian brands there is also an opening. Many competitors have not looked at how they appear in AI answers at all. A few months of disciplined work on the checklist above can put you ahead of them.
Sources
- Pew Research Center, Google users are less likely to click on links when an AI summary appears
- arXiv, Generative Engine Optimization at Scale: Measuring Brand Visibility Across AI Search Engines
- LBB, Ogilvy ANZ launches Generative Impact
- Adweek, Google is helping WPP build websites that show up on AI search engines