Try this before you read any further. Open ChatGPT, Gemini or Perplexity and type: “What is the best [your category] brand in Indonesia?”
Whatever comes back is your new shop window.
For most brands the answer is uncomfortable. Either a competitor is named, or a brand nobody on your team has heard of, or the assistant explains the category without naming anyone at all. You are simply not in the conversation.
That is the problem Generative Engine Optimization — GEO — exists to solve.
Search did not die. It moved.
For twenty years the game was simple. Rank on page one of Google, win the click, and the visit was yours to convert.
Now a growing number of people never see page one. They ask an assistant a question and get one answer, written as a paragraph, with two or three brands named inside it. No blue links. No scrolling. No second place.
That changes the goal. SEO asks how do we rank? GEO asks a harder question: how do we get mentioned, quoted and recommended inside the answer itself?
And look at what kind of person is asking. Someone typing “which CRM should a mid-sized retailer in Jakarta use?” is not browsing. They are building a shortlist. A mention there lands at the exact moment a decision is being made.
Why the old playbook is not enough
Ranking and being recommended are not the same skill.
A search engine matches your page to a query. A language model does something different — it assembles an answer out of everything it has read about you, from all over the internet, plus whatever it pulls in live while it is writing.
So a few things behave differently:
- You get described, not just listed. The model summarises what your brand is. If your positioning is vague or inconsistent across the web, the summary will be vague too.
- Other people’s words count as much as yours. Reviews, forums, news, comparison articles, directories, video transcripts. The model reads the whole room, not only your homepage.
- Clear beats clever. Beautiful but vague brand copy gives a model nothing to repeat. A plain sentence — what you do, who for, where, at what price level — is quotable.
- Being absent is worse than ranking low. On Google, position eight still exists. In an answer that names three brands, you are either in it or invisible.
What actually moves the needle
GEO is not a trick or a new tag you paste into your header. In practice it comes down to a handful of unglamorous things done properly:
- Answer real questions in plain words. Pages that deal with what buyers actually ask — comparisons, how pricing works, “is this right for a business like mine” — get pulled into answers. Brochure copy does not.
- Make your facts easy to lift. Clear headings, short definitions, specifics, dates, numbers you can stand behind. Models quote what is easy to quote.
- Say the same thing everywhere. One consistent description of your business on your site, your LinkedIn, directories and press coverage. Contradictions make a model hedge, and hedging means no recommendation.
- Earn mentions you did not write. Being cited in credible articles, listings and communities is the closest thing GEO has to backlinks.
- Stay machine readable. Structured data, clean HTML, pages that are not buried under heavy scripts. If a crawler cannot read it, a model cannot repeat it.
- Then measure it. This is the step almost everyone skips. You can track how often each assistant names you, for which questions, and who it names instead of you.
The question worth raising in your next marketing meeting
Not “how is our SEO doing?” You already have a dashboard for that.
Ask this instead: when a customer asks an AI about our category, what does it say about us — and who does it recommend in our place?
Very few marketing teams can answer that today. The good news is that it is measurable, and it is usually fixable well within a quarter.
How BIG helps brands get visible inside AI answers
GEO sits exactly where BIG has always worked: between technology and performance marketing. Web and mobile builds that machines can actually read. Communication strategy that says something specific enough to be worth repeating. Digital performance work judged on outcomes rather than impressions.
Our approach is deliberately unromantic:
- Audit. We ask the major assistants the questions your buyers ask, and record what they say about you, about your competitors, and about your category.
- Diagnose. We find where the gap comes from — thin content, unclear positioning, a site the models cannot parse, or simply no third-party evidence that you exist.
- Fix. Content, structure, technical foundations and your off-site footprint, tackled in the order that moves visibility fastest.
- Track. We measure again, month after month, so you can watch your share of AI answers change instead of guessing.
The brands that get named by assistants two years from now will be the ones that started making themselves legible to these systems today. It is the same advantage early SEO handed out — and the same window, closing at the same speed.
Curious what the assistants are saying about your brand right now? That is the first thing we check, and it only takes a short conversation to show you. Email us at sales@bigindigital.com or reach us through our contact page. We will run your category through the major AI assistants and walk you through exactly what comes back — including who is being recommended instead of you.