Questions, answered plainly.

Getting your brand named and recommended inside the answers AI tools give. ChatGPT, Claude, Perplexity, Gemini, Google’s AI Overviews, Copilot. A search engine hands your buyer a list of options. An answer engine hands them a decision.

Ranking third on Google still gets you seen. Coming third in an AI’s judgement usually means it doesn’t mention you at all. The work differs too. SEO is mostly about your own site. AEO is mostly about whether the AI knows who you are, and whether the sources it trusts talk about you.

Yes, and we wouldn’t tell you to drop it. Good search fundamentals help an AI find you and trust you, and plenty of buyers still use the ordinary results page. The problem is that a top ranking stops earning when an answer appears above it and nobody scrolls. Treat this as something you add to SEO, not something you swap it for.

Broadly, yes. The field is new and nobody has settled on a name. GEO, LLMO, AI SEO and AEO all describe the same problem. We say AEO because the thing you’re trying to get into is the answer.

It depends on your category, and that’s worth finding out before you spend anything. In some sectors buyers are already choosing suppliers this way. In others it hasn’t arrived yet. The free report will tell you which one you’re in, and if the answer is “not yet”, we’ll say so.

ChatGPT, Claude, Perplexity, Gemini, Google’s AI Overviews and Copilot, each reported on its own. They disagree with each other far more than people expect, so averaging them into one score hides the part you actually need.

We agree them with you up front, then leave them alone. They’re the questions someone types when they’re close to buying. Comparisons, alternatives, “best X for Y”. Not the ones they type when they’re reading around a subject. If the questions change between runs, you can’t compare the results.

Only if your buyers do. If nobody types “in Dubai” when they look for what you sell, adding it gives you a different test rather than a sharper one. B2B software is the clearest case. Put a city in and you get a local-services answer that has nothing to do with how anyone buys it. We work this out with you rather than guessing.

Every run is signed out, with no history and no personalisation. If we stayed logged in we’d be measuring our own browsing rather than your market. We also note the model version each time, because an AI update can look exactly like a change in your position.

Monthly on a retainer, using the same questions every time. These answers shift on their own, sometimes a lot, so one reading tells you where you were on that day rather than where you stand.

A group of your questions that we deliberately leave alone for the whole engagement. If the ones we work on improve and the ones we left alone improve by the same amount, your category moved and we didn’t cause it. That’s how you get a net figure rather than a number nobody can explain.

More than once, and we print the number in the report. Ask the same AI the same question twice and you can get two different lists of companies, so one run tells you very little. If a report doesn’t say how many runs are behind it, ask.

Three things. Whether the AI can work out who you are and what you sell. Whether the sources it draws on mention you. And whether your own pages are written so an answer can be lifted straight out of them. Most of the weight sits on the second one, which is the awkward one, because you can’t publish your way to it.

On its own, usually not. Studies of tens of millions of AI citations keep finding the same thing. Most point at sites the brand doesn’t own. Press, directories, review platforms, forums. You can add pages to your own domain all year and watch the answer stay where it was.

It helps the AI work out who you are, which is a real problem worth fixing. On its own it won’t get you named. We do it early because it’s quick and cheap and it takes one of the three causes off the table. Then we get on with the part that actually moves the answer.

Not fake ones. These models lean heavily on community sites, so being genuinely useful there does matter. Planted posts get spotted, deleted, and traced back to the brand that paid for them. It’s not a risk worth taking with your name on it.

The technical fixes can show up within a few weeks. Getting mentioned on other people’s sites takes months, because it depends on people who don’t work for you. We’d rather re-measure every month and show you where it has got to than promise a date we can’t control.

No, and be careful of anyone who says they can. These tools change how they build answers without telling anybody. What we can do is show you where you stand now, work on the causes we can influence, then measure again. You will see what changed and why.

Sometimes that settles it and often it doesn’t. A competitor who owns ChatGPT in your category is frequently missing from Perplexity or Gemini entirely. The first measurement shows you which. If it turns out the category really is closed, we’ll tell you and turn the work down.

Then we tell you and turn the work down. Some categories really are sewn up by two incumbents with no way in. If our measurement always happened to conclude that you should hire us, it wouldn’t be worth much to you.

Your category run across all six engines: who gets named, in what order, and which sources each answer was built from. It’s the real thing, not a sample. Take it to your own team and act on it if you’d rather.

Not for the report. Once we start making changes we need either CMS access or a developer on your side. If you’d rather keep us out of your site entirely, we’ll write the changes up and hand them to your team.

Yes, and it usually works better that way. Getting mentioned on third-party sites is most of what a PR team already does, and the technical fixes sit close to SEO. We’d rather brief the people you already pay than charge you twice for the same work.

Every engine separately, never one blended score. Raw counts next to any summary figure. The questions we left alone shown beside the ones we worked on. And the prompt list, market, date and run count attached, so anything in the report can be traced back and checked.

Anyone who wants a guaranteed position. Anyone who wants a set number of blog posts a month. And anyone in a category where buyers aren’t using AI to choose suppliers yet. Better to find that out in the free report than three months into a retainer.