Frequently asked questions

Answer-engine optimization is the practice of getting a brand named and cited inside AI-generated answers from systems like ChatGPT, Perplexity, Google AI Overviews, Gemini and Claude. Because an answer engine names a small number of options rather than returning a ranked list, the objective is inclusion in that shortlist rather than position on a page.

SEO competes for position in a ranked list, where being eleventh still means being present. AEO competes for inclusion in a set of three or four names, where being excluded means being invisible. There is no page two of an AI answer.

The second difference is where the leverage sits. SEO is driven substantially by your own site's authority and content. AEO is driven more by what trusted third parties say about you, because models weight external corroboration heavily. You can hold position one and still not be named.

In practice, yes. Generative engine optimization and answer engine optimization describe the same work; GEO is more common among practitioners from a traditional SEO background. Treat anyone drawing a sharp methodological line between them as selling a taxonomy rather than a service.

ChatGPT, Google AI Overviews and AI Mode, Perplexity, and Gemini as standard — four engines in every Index Report and retainer. Claude and Copilot are available on request where your buyers actually use them.

Yes. The mechanism is identical: a buyer asks for the best product in a category and the model names a few. Consumer categories often move faster than B2B, because review content and community discussion — both weighted heavily in retrieval — are more abundant.

No, and we would advise against it. Traditional search still sends substantial traffic, and the disciplines overlap on content quality, technical health and authority. We work alongside your existing SEO agency or in-house team. We are not trying to replace them and will say so on the call.

A proposed plain-text file at your site root giving language models a curated map of your most useful content, loosely analogous to robots.txt. Adoption by model providers is not guaranteed and anyone telling you it is a ranking factor is guessing. It costs almost nothing to publish, so the expected value is positive. Ours is here.

Between 120 and 200, composed roughly of 40% category shortlist prompts, 25% head-to-head comparisons, 20% use-case phrasings and 15% alternative-seeking prompts. Brand-name prompts are excluded entirely — asking a model about you after naming you proves nothing.

Large language models are non-deterministic. Ask the same question five times and you can get five different vendor sets. A single-run figure is a sample of size one presented as a fact.

We run every prompt at least five times per engine and report the mean with its variance band. The practical consequence is that we will sometimes tell you a month-on-month change is not significant — which vendors reporting a clean single number have no basis to know.

It will happen. When a major provider shipped an update in late 2025, citation counts moved across every dashboard in the industry for reasons unrelated to any brand's relevance. Every data point we record carries its engine and model version, so a provider update appears as a labelled model event separated from your trend line. We set that expectation in the proposal, not in an apology.

You help write it. The panel is built jointly and signed off before any measurement happens, and it is printed inside every report. This is also the question we would suggest you put to every other vendor you are evaluating.

We charge $1,500 for a one-off Index Report, $3,000/month for Core and $6,000/month for Category. Across the market, mid-market retainers run roughly $2,000–$8,000/month, one-off audits $1,500–$5,000, and enterprise engagements $10,000–$30,000/month. Full breakdown on the pricing page.

If measurement is all you need, buy the tool. Licences run roughly $85–$500/month and they are genuinely good products — we use one in delivery and the cost is inside our fee.

A dashboard tells you that you are losing. It does not write the comparison article you are absent from, negotiate the third-party placement, fix the entity data confusing the model, or run disclosed community presence. That work is people.

The Index Report has no commitment. Core has a three-month minimum, then rolls monthly with 30 days' notice. Category has a six-month minimum on the same terms. The minimums exist because the work genuinely takes that long to show, not to trap you.

Yes, in full, against your first month, provided you start within 60 days of delivery.

No, and we will not claim it. Attribution from an AI answer to a closed deal is unsolved across the whole industry. Anyone promising you traceable revenue today is selling something they cannot deliver.

What we commit to is measurable movement in Recommendation Rate against a panel you approved, with variance shown. We agree that success metric in writing before we start, so nobody is redefining success at the review.

No. Answer engines are third-party systems whose behaviour nobody outside the model providers controls. Any agency guaranteeing inclusion is either misunderstanding the mechanism or misrepresenting it, and in both cases you should keep looking.

Never. Community discussion carries real weight in retrieval, which is exactly what makes the temptation real and the rule absolute. We disclose affiliation every time, and we would end an engagement rather than run an undisclosed campaign. A detected manipulation campaign ends your standing in that community permanently.

Not yet. We launched in 2026 and we will not publish results until they are real, client-approved, and reported with the variance bands left in.

We know that costs us deals against firms with three years of published outcomes. We would rather lose those than put an unverifiable claim on a site whose entire argument is about auditable measurement. What we offer instead is a methodology you can inspect and a free finding that shows you the work before you pay for it.

Nairobi, Kenya, working with clients across the US and UK. East Africa Time covers the entire European working day and every US East Coast morning; we hold two fixed call windows so scheduling is never a negotiation. We invoice in USD. Judge us on the method — it is published in full precisely so that you can.