Four situations that come up in the first visibility study. None of them shows in Search Console, because it has no column for model answers. Until you count them, the questions in your category go to a competitor the model already names.
The query ends with a ready answer above the results. Impressions in Search Console grow, clicks stay flat, and it looks like a drop in CTR rather than a change in the market.
Asked a question about your industry, it names three competitors. Not because they are better, but because their content is in a form the model can quote.
An old offer, an address that no longer exists, a price from two years ago. Models learn from copies of the web, and your site is not the only source.
The customer asks about one thing, and the answer is scattered across five paragraphs. A person pieces it together, a model does not. It quotes the passage that sits closest to the question, even if it is the weakest.
The first stage is a visibility study: a set of questions from your category put to four engines, a record of what the models say about you today, and a list of facts to correct.
The same editorial work, a different unit of success and a different reader of the content. That is why one does not replace the other, it adds a layer.
| Aspect | Classic SEO | AI SEO |
|---|---|---|
| Goal | A high position in the list of results | Being quoted in the model's answer |
| Unit of success | Position, a place from 1 to 10 | Share of answers, the percentage of questions in which the brand appears |
| Reader of the content | A person clicking a link | A model choosing a passage to quote |
| Form of content | A page built around a keyword | Self-contained passages: the question in a heading, the answer in the first sentences |
| Structured data | Meta tags, heading hierarchy | Plus schema.org, a consistent business entity, a text version for crawlers |
| Measurement | Search Console, keyword positions | A set of test questions put to the models on a regular cycle |
| Data freshness | Stable, rare corrections | Needs the facts about the brand watched in external sources |
| Time to effect | Weeks to months | Varies by engine: AI Overviews and Perplexity faster, ChatGPT and Gemini over quarters |
| Traffic | A click from the list of results | Often no click, the value is in the mention and the citation |
We do not start by writing content, but by checking what the models already say and where they get it from. We do not promise a place in a model's answer, because nobody controls it: not us, and not a supplier who makes such a promise. We agree on a number and on the work to change it over time.
3-4 weeks
The set of questions from your category goes to four engines, with each question asked several times so that the answer is not a fluke. The result: how often you appear, how often your competitors do, which sources are quoted and what the models say about you that is untrue.
from month two
First, what can be fixed fastest: a consistent business entity in the structured data, up-to-date information in external sources, and cleaning up what the model finds about you outside your site.
ongoing
Customer questions as headings, answers in the first sentences, figures with a source. The same work usually lifts classic SEO too, because Google has long been looking for the same thing.
ongoing
The same set of questions asked on a regular cycle in a monitoring tool that you order separately, outside the PROFIT plans. This way the share of answers can be compared over time.
We quote after the visibility study, because only the study tells us how much work there is to do. The pricing and the scope of the packages are below.
Visibility in model answers cannot be shown with one screenshot, because the answer changes with every query. What can be shown is a repeatable measurement: every week we put the same set of seven Polish-language buying questions about agencies to ChatGPT, Gemini and Google, and count how often our brand appears in the answer. Below is the reading from 2 September 2026.
of ChatGPT answers to seven questions about agencies name JustIdea, roughly one in five
of Gemini answers to the same questions name JustIdea
of Google AI Overviews answers to the same questions name JustIdea, roughly one in four
The service is young, so promises are cheap, and there is nothing yet to compare us against: no rankings and no certifications in this field. What remains is how people have rated our work on visibility over the years. This is the only place on this page where JustIdea is not describing itself.
weighted average of 224 reviews across three platforms. Clients gave them, and each site below links to the profile with the full text of every review.
Read the reviewsFor each one we set out what it gives in the work on visibility in the models.
AI Overviews is a Google product and changes faster than its documentation. Partner status gives earlier access to information about changes in the results, and a route to ask questions when the search engine stops behaving the way it is publicly described.
Copilot and Bing search draw on the same index, and it is an index most businesses overlook. Partner status gives access to tools on Microsoft's side, including data on how that index sees your site.
Models also learn about brands from social profiles and from what people write about the brand there. Tidying up the profiles is part of correcting the facts, and partner status shortens the route when business data has to be fixed at the source.
We start with a visibility study, because without it the rest cannot be priced honestly. Further work is a subscription and usually runs together with SEO, not alongside it. One-off products for single models are on the pages about ChatGPT, Gemini, GEO, Bing and the AI Overviews audit.
What the models answer to questions from your category, who they quote and what outdated things they say about you.
The same set of questions in a model-monitoring tool, with a reading of your brand's and your competitors' share. You pay once, for setup and the first reading.
Classic SEO and visibility in the models run as one piece of work, because to a large extent they are.
Three articles, in Polish, on what the results page looks like today and where models get the sentences they quote. Without them it is hard to judge what in this service is real.
The call is led by a new business specialist. When we get into the details of an account, the specialist for that channel joins in. We reply within one business day.
Before the meeting we review your website, your visibility and what your campaigns show from the outside. We will not open with “so, what does your company do?”.
The reply comes from the same new business specialist who will run the call. No qualification form and no call from an unknown number.
We go through your numbers and your questions. If it turns out we are not the right fit, we will tell you straight away.
