AI Search Strategy
Do ChatGPT, Claude, Gemini and Perplexity recommend you at the end of a buying conversation, and why or why not? We measure your site, your market and what AI believes about your brand, and an experienced strategist turns it into one strategy. Including the AI Search Website Audit and one re-measurement.
In short
Add the workshop later? You only pay the difference.
The question this answers
Why are we not recommended, and what do we do first?
Marketing teams that want to get recommended, with AI Search as the weight, and can carry out a plan across brand, content, PR and tech.
Anyone who only wants a daily visibility chart. That is a thermometer, and there are cheaper ones.
Our approach
Our platform measures your site, your market and how the AI assistants talk about you, up to the end of the buying conversation. An experienced strategist turns it into one strategy.
Nothing goes out unread
What is in it
In the brand profile we set down together who you are, for whom, against whom and which associations you want, per product group where needed. Then Topics maps the buying moments in your market: the situations in which someone starts thinking about the category, the topics underneath and the questions buyers ask. The questions we measure come from there. Per topic you see search volume, trends and how you do in live AI answers.
The AI Search Website Audit is fully included: five layers, a score per page and actions per page. It is the part of your evidence you control yourself.
How your site does with people and machines: structured for search, parseable for agents and persuasive for buyers.
Editorial pages: substance, clarity and light E-E-A-T.
Commercial pages: persuasion, positioning and E-E-A-T.
Technical SEO: crawl, performance, schema, semantics, authority.
How discoverable, parseable and citable the site is for AI agents.
42 pages across two templates ship without Organization markup, so AI engines can't resolve who you are.
Market trends, share of search, traffic and links, and a card per competitor: where you win or lose and what you can learn from it.
Many buyers look at the sites of the named brands after an AI recommendation, and that is where you can still lose the choice. A panel of your own ideal customers puts your homepage and landing page next to your competitors', without seeing the brand names. Per page type you read who wins, on which of the six dimensions, and what you learn from the winner.
Velora names the service interval and the price per year above the fold. Your page makes the buyer look for both.
We read reviews, communities, trade media and social, the way the AI assistants do too. You see how customers judge you and your competitors and which pain points in the market nobody solves yet. You also see which media write about your competitors and not yet about you, and which reviews and forums call for a response.
We ask ChatGPT, Claude, Gemini and Perplexity open questions about your brand and set the answers next to who you want to be: identity, facts, associations and the sources behind them. You see right away what is wrong and what to work on. The score carries the label Interpretation, because it is a reading by AI models.
To AI assistants, Acme Cycles is the specialist in commuter e-bikes: long range and little maintenance, but smaller than Velora.
Per topic and per assistant we measure repeatedly whether you are mentioned, whether you are the recommendation and whether your site is used as a source. Every number carries the 95% band and a label that says how much it is worth. Memory and live answers are kept apart, with per topic the share of live answers where the model actually searched.
| Live | Memory | |||
|---|---|---|---|---|
| Platform | Named | Recommended | Named | Recommended |
| ChatGPT | 71% | 42% | 58% | 33% |
| Claude | 40% | n=3 | 35% | 21% |
| Gemini | 58% | 31% | 44% | 22% |
| Perplexity | 84% | 55% | n/a | n/a |
The searches a model fires during live answers, across all your topics, per engine, clustered into what they actually research. Plus the checks against your own site: do they land, and which words your pages are missing. This is the knob closest to your hand.
1,984 of 4,203 captured sub-questions are checks against a site (47%).
A second model plays a buyer from your audience, built from your buying moments, competitors and personas, and keeps asking: orienting, comparing, deciding. You see where you drop out, who displaces you and on which criterion. From all conversations we distil why you are or are not recommended, and what to do about it.
Differentiation brief
The topic brief reads the ten pages Google shows for a topic and the pages AI assistants cite for it, finds what none of them answers and lays your brand profile beside it. It opens with a verdict: a new page, improve the page you have, a different kind of page, or do nothing. For your editorial key pages, the differentiation brief does the same per page.
Example: the topic e-bike maintenance
The brief
Make a new page
The brief opens with a verdict, and that can be that there is little to win here.
Every recommendation names the page, the question or the brand fact it rests on. Whatever rests on nothing is removed before you see the brief.
An illustration with a made-up bike brand. In the platform this runs on your own pages and topics, in the language of your market.
A quote is looked up in your page, a question has to be on the list Google itself shows, and in the topic brief every recommendation names the page, the question or the brand fact it rests on. Whatever rests on nothing is removed before you see the brief.
What the top pages do well is listed as context. The recommendations come from what they leave open and from what only you can say: your own figures, experience from practice, a point of view.
When the top pages cover the topic well and your brand has nothing of its own to say about it, the brief says so. That saves you a page nobody would miss.
What you get
Where you stand, where the decision falls in your market, where your evidence falls short and what the order is. Read, interpreted and written by us.
The strategy and the main lines in a deck for your team or board.
| A | B | C | D | E | |
|---|---|---|---|---|---|
| 1 | Action | Page | Impact | Effort | Owner |
| 2 | Service interval above the fold | /e-bikes/city | High | Low | Content |
| 3 | Add Organization schema | templates | High | Low | Dev |
| 4 | Comparison with Velora | /vergelijk | High | Medium | Content |
| 5 | Merge dealer pages | /dealers | Medium | Medium | Dev |
| 6 | Give GPTBot access | robots.txt | High | Low | Dev |
| 7 | Respond to reviews | reviews | Medium | Low | Marketing |
By impact and effort, with what you change for each action, on which page or channel, why it matters and who picks it up. Ready to hand to your team or agency.
Peter personally goes through the key findings with you in an hour, so you know what to do first and why.
At a moment you choose, after you have put the recommendations in place. On the same instrument, and a change only counts when the bands no longer overlap. After that, a re-measurement costs 495 euro.
How it works
This is how we deliver a world-class AI Search Strategy, step by step.
You pick your package and fill in the form.
Through the intake form you fill in your brand profile, so we get a clear picture of your audience, your market and your goals. Then we go through it together in a short call.
We analyse your site, your market and your performance in AI Search. An experienced AI strategist reads every finding, connects them and writes the strategy. Human insight and expertise, on top of the measurements.
We present the findings and recommendations in a one-hour online call. You receive the advice report, the presentation and the action plan with every recommendation.
At a moment you choose, after you have put the recommendations in place, we measure again and you see what really changed and what can still improve. One re-measurement is always included in your package. After that, a re-measurement costs 495 euro.
Questions
The AI Search Website Audit reads your own site. The AI Search Strategy adds everything outside your site and inside the assistants: your market, how often you are recommended, what the models look up, whole buying conversations, your reputation and what the market says. The AI Search Website Audit is fully included.
ChatGPT, Claude, Gemini and Perplexity, through their official APIs, with the model each app serves by default. On top of that we read per topic whether Google shows an AI overview or an AI Mode answer, and whether your site is cited in it.
We ask every question repeatedly, and every number carries the band the real value sits in plus a label. A failed measurement is not a zero. Under five usable answers we say it was not measured. How that works exactly is on the methodology page.
At a moment you choose, after you have put the recommendations in place, we measure the same topics on the same instrument. We only report a change when the two ranges no longer touch, so you know whether it was real. Then we go through together what changed and what can still improve. One re-measurement is included in your package; after that a re-measurement costs 495 euro.
Other packages
Request the package directly. We confirm your request and plan the intake together.
Fixed price · no subscription · results within two weeks
