Generative Engine Optimisation, in Plain English
- 1 hour ago
- 5 min read
A founder in Singapore types her own company name into ChatGPT one evening, half curious and half cautious. A tidy paragraph appears, describing what her business does, who it serves, and how it works. Much of it is right. Some of it is a year out of date. She sits back and realises that a machine has been quietly introducing her business to buyers all week, using whatever it could find about her.
This is the world generative engine optimisation is built for. Generative engine optimisation, or GEO, is the practice of becoming the clear, well-evidenced source that AI engines choose to quote when someone asks a question in your field. Your next customer often meets a machine's summary of you before they ever meet you. The work is to make that summary accurate, useful, and unmistakably yours.
What is generative engine optimisation?

Generative engine optimisation is the practice of making your business the clear, well-evidenced answer that AI engines such as ChatGPT, Google AI Overviews, Gemini, and Perplexity quote when a buyer asks a question in your category. When someone asks one of these tools a question, the engine reads across many pages, writes a single reply, and names a few sources. GEO is the craft of being one of those named sources.
It helps to picture the change in plain terms. For years, search handed people a shelf of links and let them choose. Now the engine reads the shelf, forms a view, and speaks it aloud as a recommendation. GEO is how your business earns a fair hearing in that recommendation, so the spoken answer reflects what you truly do.
The idea has a clear origin. Researchers at Princeton and IIT Delhi published the first formal GEO study, and found that structuring content well can lift a source's visibility inside AI answers by up to 40 percent, with the largest gains coming from adding relevant statistics, direct quotations, and citations from credible sources. That single finding holds the whole discipline: AI engines reward content that is easy to trust and easy to quote.
Why does AI search matter for a Singapore business now?

Something has shifted in how people look for answers, and all of us are adjusting to it at the same time. In the first four months of 2026, around 68 percent of Google searches in the United States ended without a click, because the answer already sat on the results page. AI Overviews now appear on more than a fifth of Google searches, and when they appear, click-through to the pages below falls by close to 60 percent.
People are also going straight to the AI tools. Monthly visits across the major AI platforms grew 70 percent year on year, reaching 9.5 billion by May 2026. Your buyers are among them. A Forrester survey of roughly 18,000 business buyers found that 94 percent used AI during their most recent purchase, up from 89 percent a year earlier, and named AI answer engines their single most meaningful research source.
This carries extra weight across Southeast Asia's eleven markets, where buyers move fluidly between languages, platforms, and borders, and will happily ask an AI assistant a question in one language about a business rooted in another. For a company in Singapore serving the wider region, the AI answer is often the first handshake a buyer has with the brand.
Put simply, a growing share of first impressions now form inside an answer the engine writes, drawing on sources you can shape.
How do AI engines decide which sources to trust?
They favour content that reads like a clear, honest answer to a real question. The Princeton study tested this directly, and the sources that carried credible statistics, direct quotations, and citations to authoritative material earned a place in the answer far more often. The lesson is friendly: the engines reward the same generosity that good writing has always shown.
The attention is also spreading across more homes. ChatGPT's share of generative AI web traffic moved from about 76 percent in mid 2025 to 53 percent by May 2026, as Gemini grew to roughly a quarter and Claude to close to a tenth. Being quotable in one engine is a fine start. Being quotable across several is the real goal, and the habits that earn it stay the same everywhere: clarity, evidence, and structure.
What can you do this week to become more quotable?

A handful of habits do most of the work, and you can begin them today.
Write the answer first. For each question your buyers ask, place a clear, self-contained answer in the opening lines of the page, so an engine can lift it cleanly.
Show your evidence. Add a current statistic, a source link, or a concrete example under each claim, since evidence is what the research shows moves visibility most.
Structure for people and machines together. Headings phrased as real questions, short paragraphs, and plain words help both kinds of reader.
Keep your facts current. Refresh figures, prices, and service descriptions often, because the engine will repeat whatever it finds.
Tell one coherent story everywhere. Your website, your profiles, and your listings should agree with each other across every market you serve.
Two kinds of visibility worth planning for

Classic search and AI answers reward overlapping habits, and a healthy plan feeds both at once. Here is how they sit side by side.
Classic search visibility | AI answer visibility | |
What the buyer sees | A ranked list of links to choose from | A written answer with a few named sources |
What earns a place | Relevance, quality links, healthy pages | Clarity, evidence, and quotable structure |
The main signal to watch | Rankings and clicks to your site | Mentions and citations inside answers |
How you check it | Track your rankings and search traffic | Ask the engines your buyers' questions and see who they quote |
What helps both | Genuinely useful, well-evidenced content | Genuinely useful, well-evidenced content |
What we believe at Eysy
We believe the businesses that do well in AI search are often the ones that were already generous with clarity, and that generosity is simply worth more now. The engines have raised the reward for being genuinely helpful. We are learning this alongside our clients, week by week, as the engines keep changing.
Our part is to carry the judgment. Which questions your buyers actually ask. Which claims deserve evidence. How to structure a page so it earns a quote. How to keep a brand voice warm and human while a machine repeats it. We do that through hands-on work, through teaching and workshops so your own team grows the skill, and through video and content production that gives the engines something worth quoting. The tailored execution changes with every business, and that judgment is the reason to have a conversation.
Where to start this week
Here is one honest thing you can do on your own. Open ChatGPT, Gemini, and Perplexity, and ask each of them the three questions your best customers ask before they buy. Read how the engines describe your field, and notice which sources they quote. That single hour will show you where you already appear, where the answer is still thin, and which question is worth answering clearly first.
When you would like a partner to turn that picture into a plan, or to teach your team to run it themselves, we would be glad to talk.
Sources
Search Engine Land, Google zero-click searches reach 68% in early 2026 (SparkToro and Similarweb data): https://searchengineland.com/google-zero-click-searches-2026-study-479717
Similarweb, AI Search Stats in 2026 (platform visits and traffic share): https://www.similarweb.com/blog/marketing/geo/gen-ai-stats/
Aggarwal et al., GEO: Generative Engine Optimization, Princeton University and IIT Delhi: https://arxiv.org/abs/2311.09735
Forrester 2026 Buyers' Journey Survey, summarised (94% of B2B buyers used AI): https://machinerelations.ai/research/b2b-ai-vendor-research-2026




Comments