When a customer asks AI, who does it recommend?
Some people no longer ask Google, but ChatGPT or Perplexity. They get one answer, and it usually names a handful of companies. Either you're among them, or you're not. It goes by GEO or AEO — the substance is the same.
For brands that don't want to find out after five years that their customers were looking elsewhere. And for companies that see competitors in AI answers, but not themselves.
- GEO (generative engine optimization). Optimising for engines that assemble an answer instead of listing links: ChatGPT, Perplexity, Google's AI overviews.
- AEO (answer engine optimization). Optimising for the answer — originally for featured snippets and voice search, now used mostly about AI assistants.
- In practice the two overlap so heavily that almost nobody separates them, and we do not need to. The work is the same: be a source that can be read, verified and quoted. That is why this is one service rather than two.
- What it is not: SEO. That competes for a position in a list of links and has its own service here. The technical groundwork is shared, the rest is not.
- We find out how things look today. We test what the main AI tools say about you and your field, and why they cite someone else.
- We adjust your content and website. Structured data, content that genuinely answers questions, wording that can be quoted. Not keyword stuffing.
- We measure. We track how often and in what context you appear, and compare it with the competition.
Whoever settles into AI answers as a source is harder to push out. Much like the top positions in search fifteen years ago, only it's happening faster.
What people ask us.
Is GEO the same as AEO?
In practice, yes. GEO is the name for optimising towards generative engines, AEO for optimising towards the answer. They come from different moments and different directions, but today both mean getting into the answer a model assembles. Anyone selling you GEO and AEO as two services with two prices is selling one thing twice.
Is it the same as SEO?
It overlaps, but it isn't. SEO competes for a position in a list of links. Here you compete to have a language model use you as the source of its answer. Different rules, different work with content.
Can it even be measured?
It can, though differently than positions in Google. We test a set of questions your customers actually ask and track how often you appear in the answers versus the competition.
How long until it shows results?
Usually weeks to months, depending on how much content needs reworking and how often the models update. It isn't a switch.
Won't it disappear when AI changes?
The essence of the work is to make you a trustworthy and readable source. That will hold for the next generation of models too.