AI search & GEO.
Be understood, cited and represented when AI answers the question.
AI search is another place customers discover, compare and verify brands.
We measure how your brand is represented across AI answers, find where competitors or incorrect information appear instead, then improve the technical, content and entity signals we can actually influence. What a model chooses to say is not ours to control. What it has to work with, and how well that is measured, is.
Talk to an AI search specialistRecent work
Ampcontrol
An Australian electrical engineering manufacturer. We track how the brand appears across a fixed set of category prompts and AI engines, and what AI referrals do once they reach the site.
- 4
- ChatGPT-attributed conversionsObserved · GA4 · March 2026, the first converting AI month
- 5 engines
- Citation trackingIn-house AI visibility tracking
Our view
AI search is another decision surface, not a separate universe.
Customers use AI assistants to find options, compare products and services, check claims, understand a category and build a shortlist. That puts AI search inside discovery, consideration and preference: the same decisions the rest of search already shapes.
It does not run on its own. AI answers can reflect information found across websites, search engines, publishers, communities and structured information, so what is said about a brand elsewhere can shape what an answer says about it.
Nor is it SEO under a new name. The surfaces differ, the output is an answer rather than a ranked list, the measurement differs and so do the risks: a brand can be left out, described inaccurately, or see a competitor recommended in its place.
What the work covers
Three areas of work, run together.
Not stages. Each runs alongside the others, and most engagements weight them differently.
Where we measure
Five answer environments, monitored and tested.
ChatGPT
Where many customers now ask for recommendations and comparisons directly. We test the category prompts that matter to you and record who is named, who is cited and what is said.
Google AI Overviews and AI Mode
AI answers inside Google, shown above or in place of the usual results for some queries. We track which queries trigger them, which sources they cite and whether you are one of them.
Perplexity
An answer engine that shows its sources openly, which makes it a useful view of what gets referenced for a question and how that changes over time.
Gemini
Google’s assistant, used on its own and inside Google’s products. Tested on the same prompt set, so the result is comparable with the other engines rather than read in isolation.
Claude
Included where it is part of how your customers research. Monitored on the same prompts, with the answer recorded and the brand’s presence and accuracy noted.
Scope
What AI search & GEO can include.
Not every engagement needs all of it. Scope follows the problem.
-
Baseline and research
An AI visibility baseline across a fixed prompt set, prompt and query research, and competitor citation analysis.
-
Monitoring
Citation and mention monitoring, answer accuracy and misinformation monitoring, and measurement across answer engines.
-
Understanding and readiness
Entity clarity, and structured data and technical readiness where they are relevant to the gap.
-
Content and opportunity
Optimising content for AI answers, citation opportunity analysis, and analysis of AI referral behaviour on the site.
Full SEO, broader content production, digital PR and paid AI advertising are scoped through the capabilities that own them: SEO & search performance, Content & search experience, Digital PR & authority, and Paid search & media.
Measurement
Measure what AI search actually gives us.
AI answers vary from one run to the next, so a single prompt proves very little. We measure across a fixed prompt set, repeat it on a schedule, and report each engine separately, with the sample size and period beside every figure.
Some of what follows is observed directly. Some can only be inferred. We keep the two apart, and we do not roll them into one AI visibility score or present them as revenue.
Observed 7
Measured directly · sample size and period on every figure
- Presence across a fixed prompt set, per engine
- Citation and source rate
- Competitor presence in the same answers
- Answer accuracy and misinformation frequency
- The context a brand or product is mentioned in
- Referral traffic, where the engine passes it
- AI-attributed conversions, where they can be observed
Inferred 4
Reasoned from the observed evidence · always labelled as inference
- Influence on decisions made inside the answer, with no click
- Contribution to branded search and direct visits
- Why a change happened, when the engine does not say
- Commercial impact beyond what attribution can see
Where a change follows the work, we report it as a change observed after the work, not as proof that the work caused it.
Selected clients
Before you brief us
Questions, answered plainly.
01 What is GEO, or generative engine optimisation?
Work intended to improve and measure how a brand is understood, represented, cited and surfaced in generative search and AI answer environments, such as ChatGPT, Google AI Overviews, Perplexity, Gemini and Claude.
02 Is AI search different from SEO?
Yes, though they are not separate worlds. They can share foundations: the website, its content and how clearly the brand is described. The outputs and the measurement differ. SEO is judged on a ranked list and the clicks it earns; AI search is judged on whether the brand is present, cited and accurately described inside an answer.
03 Will AI search replace SEO?
We do not know how the balance between traditional search and AI answers will evolve. Today both matter, and the right investment depends on how customers in your category actually search and decide.
04 What is the difference between GEO, AEO, AI SEO and LLMO?
The market has not settled its terminology, and the terms overlap more than they differ. AEO tends to mean answer surfaces such as AI Overviews, LLMO tends to mean large language models, and AI SEO is often used for all of it. We use AI search & GEO as the name of the capability.
05 How do you measure AI search performance?
We run a fixed set of the prompts that matter commercially, repeat it on a schedule and record, per engine, whether the brand appears, whether it is cited, which competitors appear and whether the answer is accurate. Referral traffic and AI-attributed conversions are added where the platforms pass them. Each figure carries its sample size and period, observed results are kept apart from inference, and a single run is never treated as a result.
06 Is GEO worth investing in for every brand?
No. It depends on whether the questions that matter to your customers are actually being answered in AI environments, and whether there is a meaningful commercial opportunity there. We will check that before recommending any work.
Ready when you are.
Be found first. Be chosen naturally.
Tell us which questions matter commercially. We can show you where your brand appears, where competitors appear instead and what the available evidence says about why.