Why weekly tracking when other tools scan every day?
+
Ask yourself what actually moves in 24 hours. Ask an AI the same question twice and you will not get exactly the same answer: the models introduce their own variability. Over a single day, that variability weighs far more than any change in your visibility, which depends on content published, pages edited and citations earned on third-party sites. Those movements take 3 to 8 weeks.
A daily curve rises and falls every day without anything changing on your side. You end up making decisions on noise, and believing in progress that is not there. Weekly tracking smooths that variability and lets the underlying trend appear.
And when you need an immediate reading, after a release or a fix, you trigger an on-demand scan. You choose the moment, instead of being handed an automatic refresh that teaches you nothing. We would rather sell you fewer data points that actually mean something.
How are my questions split across my projects?
+
Freely. Your plan gives you a total number of questions and a number of projects, and you decide how to distribute them. An Agency account has 75 questions for 5 projects: nothing forces you to put 15 on each. A strategic client can take 40, a smaller one just 5.
You change that split whenever you want from your account, with no input from us and without changing plan.
I can check in ChatGPT myself, why pay?
+
You can, and it is a good way to form a first impression. But manual checking hits three walls. Your ChatGPT account knows your history and your brand: it cites you more readily than it would to a stranger. A single answer tells you nothing, since the same question asked tomorrow gives a different result. And you will not know whether your score is rising or falling, with no comparison point measured under the same conditions.
That is exactly what we industrialise: identical conditions at every scan, no account and no history, across three platforms, with a history that makes the trend readable. The equivalent manual work runs to several hours a month, and it stays incomparable from one time to the next.
My AI traffic is low today, is it worth it?
+
That is true of almost every site, and it is not the right question. AI volumes will not catch up with search volumes for a long time. What matters is the direction of the curve and the concentration: a results page distributes visibility to ten players, an AI answer cites three.
That concentration is mechanical, and it produces lasting gaps. Brands that settle into those answers do not gain a little, they take the place. Those that are not there do not lose a little either. Starting to measure now costs 29 EUR excluding VAT per month; catching up on two years will cost considerably more.
Why only three AI platforms, and not AI Overviews or Claude?
+
ChatGPT, Gemini and Perplexity are included in every plan, at no extra cost per model. Each covers a distinct use: ChatGPT concentrates most questions from the general public and decision makers, Gemini sits at the heart of the Google ecosystem, Perplexity dominates sourced research.
Other surfaces are under review, and we will add them once their tracking is reliable and reproducible. That is the point we do not compromise on: an unstable measurement shown in a dashboard is worth less than no measurement at all, because it makes you decide on something false. On each platform we query the conversational version, the one people actually use.
What sets you apart from other tracking tools?
+
Three things. Most tools stop at diagnosis: you learn you are not cited, without knowing what to fix. Here a full GEO audit is included every month in all three plans, with a prioritised action plan and a 33-point checklist.
Second, all three AI platforms are included everywhere, with no per-model surcharge, and no feature is locked behind a higher tier: only the volumes change.
Third, the method is stated and documented. We explain publicly why we measure through APIs rather than by driving the interfaces, why the frequency is weekly, and what our figures do not measure.
Our view on GEO sets out those choices.
How long before the results are usable?
+
From the first scan you have a snapshot of your current visibility and your starting score. The first GEO audit gives you an immediate action plan. To see the score move after your fixes, allow 3 to 8 weeks: AI platforms take time to pick up new or edited content.
Do you measure through APIs or by driving the ChatGPT app?
+
Through the official AI APIs. A bot driving the web interface returns answers shaped by the session, the conversation history and the personalisation of whichever account it uses, and it breaks with every interface change. APIs ask the question under identical conditions at every scan, with no account and no history. That is what makes your curve comparable week after week.
How does the 14-day free trial work?
+
No card required. You create one project and track 5 questions across the three AI platforms. A first scan runs at sign-up, a second seven days later, and you get one on-demand scan. On the audit side, you receive the technical part: whether AI platforms can reach your site, any blocks in place, and content hidden behind JavaScript. The full action plan comes with a paid plan. At the end of the 14 days nothing is charged and your project is kept.
What is the Cockpyt Score?
+
A score out of 100 combining three dimensions: your presence in the answer, your position inside that answer, and whether a link points back to your site. If no AI cites you on your tracked questions, the score is zero: the other dimensions compensate for nothing, because they have nothing to measure.
Full detail here.
What is the GEO audit included every month?
+
A monthly deliverable, present in all three paid plans, available once a month per project. It brings together a technical audit of your site from an AI standpoint, an in-depth analysis of 10 questions you choose, tested under several phrasings, an analysis of your AI traffic, then a prioritised action plan and a 33-point checklist. You do not walk away with a dashboard, but with what to fix and in which order.
See the GEO audit in detail.
What is an on-demand scan?
+
A scan you trigger yourself, without waiting for the automatic weekly scan. Useful after a launch, a redesign or a fix, when you want to check the impact right away. It covers a project of your choice, always in full: every tracked question is measured, you have nothing to select. Your plan includes a monthly allowance of questions to spend as you like, from 15 on Solo up to 75 on Scale, then €5 excl. VAT per pack of 15 questions.
What is Query Fan Out?
+
Before answering, an AI breaks your question into several separate searches, launched without being displayed, and builds its answer from what they bring back. A brand can be well positioned on the question as asked and still be absent from the answer, simply because it covers none of those sub-searches. Cockpyt AI records them at every scan.
See sources and Query Fan Out.
Does Cockpyt AI work for my sector?
+
Yes. The tool adapts to your activity when you create the project: local business, e-commerce, SaaS, agency, independent practitioner, national service. Tracking questions are suggested for your sector, and for your area if you run a local business.
Can I switch plans or cancel?
+
Yes. You move up or down at any time and the new quotas apply straight away. On monthly billing you cancel whenever you want. Annual plans run for 12 months and renew automatically unless you turn renewal off before the renewal date from your account.
Are prices shown excluding VAT?
+
Yes, every price shown excludes VAT. Applicable VAT is added at checkout depending on your country and your status. An invoice is issued at each billing date and stays available in your account.
See the full comparison.
Do I own my data?
+
Yes. Your projects, your tracked questions and your results belong to you, and you can export them whenever you want while your account is active. Your data is processed in line with the GDPR and hosted in Europe.
