Your buyers are asking AI which firm to hire. Right now, it is not saying your name.

SEO earns rankings in traditional search results, AEO (Answer Engine Optimisation) structures your content so it becomes the extracted answer, and GEO (Generative Engine Optimisation) builds the third-party authority that generative models cite. Think of them as three layers of the same programme: SEO is the foundation, AEO restructures your own pages to be answer-ready, and GEO earns presence across the external sources — directories, reviews, industry publications — that account for roughly 85% of AI citations. A complete programme runs all three at once.

Yes — traditional SEO still matters, because AI systems disproportionately cite pages that already rank well and crawl sites that are technically sound. AI Overviews draw a large share of their citations from top-ranked pages, so strong rankings feed AI visibility rather than competing with it. The mistake is treating SEO as sufficient on its own; in 2026 it is the necessary foundation that answer and citation work build on top of.

The only reliable way to know is to test your buyer queries directly across multiple AI engines, repeatedly, and measure how often you are named. A single check is meaningless — the same query returns different brands on roughly 99 of 100 runs and monthly citation churn runs 40 to 60%. Our baseline audit runs your top buyer queries 10+ times each across ChatGPT, Perplexity, Gemini and Google AI Overviews to establish a statistically honest picture of where you appear today.

Early movement in citation frequency typically appears within 60 to 90 days, with compounding gains over the following six to twelve months. The first month is baseline and architecture; answer restructuring and schema changes start influencing retrieval within weeks, while third-party authority building accrues more slowly. Because citation share compounds, the businesses that start earlier build advantages that later entrants find increasingly expensive to overturn.

AI-referred visitors convert at dramatically higher rates — up to 23x traditional organic in some measurements — because the AI has already pre-qualified and pre-sold them before they arrive. They land already knowing you were recommended for their specific need, so the visit is lower in volume but far higher in intent and economic value. Judging AI channels on raw traffic misses the point; the correct metric is revenue per visitor, where AI referrals lead decisively.

Share of model is the percentage of relevant AI answers in your category that cite your business, and it is to AI search what share of voice is to advertising. It matters because it is the single clearest measure of whether you are winning or losing the AI visibility race against named competitors. We track it weekly across multiple engines with 10+ runs per query, because anything less mistakes day-to-day model noise for real movement.

The first-mover advantage is real and structural, because AI models are trained on and cite existing authority — so early citation share becomes self-reinforcing over time. Once a model consistently associates your entity with a category, that association is expensive for competitors to displace, much like an established backlink profile. The 12-to-18-month window is not a sales tactic; it reflects how quickly the citation graph in most categories is being locked in.

Considered-purchase and research-heavy industries — professional services, healthcare, legal, financial services, B2B and high-value home services — are most exposed, because their buyers ask AI detailed comparison questions before choosing. Any category where a customer would previously have researched several providers before deciding is now being intermediated by AI shortlists. Low-consideration, impulse or purely local-foot-traffic businesses feel it less, though even they are increasingly affected by AI-powered local answers.

Yes — smaller businesses can compete strongly, and often more easily than in traditional SEO, because AI citation depends on clarity, structure and topical authority rather than sheer domain size. A focused mid-market firm that answers specific buyer questions cleanly and earns the right third-party mentions can out-cite a larger, vaguer competitor. This is precisely why the current window favours decisive smaller players who move before their category consolidates.

llms.txt is a simple text file at the root of your site that tells AI systems which content is most important and how to interpret it, similar in spirit to robots.txt. It does not guarantee citation, but it helps models find and correctly attribute your most relevant pages, and it signals that your organisation is deliberately structured for AI consumption. Most sites do not have one yet, so implementing it is a low-cost, early-mover hygiene step we configure as part of the programme.

You measure AI SEO ROI by tracking citation share, AI-referral traffic, and — most importantly — the conversion and revenue attributable to AI-referred visitors against your programme cost. Because AI referrals convert at a large premium, even modest visit volumes can produce meaningful pipeline, so the honest calculation is revenue and qualified leads per dollar, not clicks. Our executive dashboard reports citation frequency, share of model, competitive share and AI-referral conversion monthly so the return is visible at board level.