Off-site signals: what Click Made was missing in AI

Below, we describe the second measurement clickmade.pl, a graphic and web design studio from Bielsko-Biała, and what it means for the work plan. We described the May starting point in the article A graphic design studio in AI chatbots: the Click Made case, May 2026.

What did the second measurement show?

Still zero, but on a broader sample. This time we checked five models: ChatGPT, Claude, Gemini, Perplexity and Mistral. None of them mentioned Click Made when answering customer questions, and the citation rate was still 0%. The only trace was in ChatGPT: the brand appeared on average in position 12.8 on the list, i.e. outside the top ten, which customers usually never get to.

The recognition check showed something else. One of the AI engines confused Click Made with another company. This is typical of brands with little independent information online: the model fills the gaps with whatever seems similar. What to do then is described in the article ChatGPT doesn’t know my company.

Was the problem the amount of content?

No. The report listed what was already on the website: 86 blog posts, service descriptions, a portfolio, a price list and a local page for Bielsko-Biała. Most of the topics that customers enquire about already had their own subpage. Adding further articles wouldn’t have changed the result, as the models had no trouble understanding what the studio does.

Real content gaps the report found only a few. Visual identity was in the portfolio and on the blog, but not as a separate service with a scope and a price. The same applied to print materials and web graphics. There was also no page explaining how remote collaboration works, even though it is one of the studio’s main strengths.

What are off-page signals?

Everything that websites other than the brand’s own say about it: customer reviews on Google and LinkedIn, directory listings, portfolio profiles, contractor rankings, guest articles and media mentions. A model that is supposed to recommend a company looks for confirmation that it is not only the company itself that speaks well of it.

The Click Made report said it plainly: the models mainly saw business directories, the Google profile and reviews, while independent rankings, comparisons and trade press were missing. That is why the result was zero despite a good website. We describe the mechanism in the article Why does AI recommend the competition?, and the difference between working for search engines and for AI chatbots in the article SEO, GEO, AEO: the differences.

In this category the models were keen to mention website builders and websites with contractors’ offers. They are the default answer to questions about a website or graphics, because there are a huge number of independent mentions of them online. A small studio will not beat them on general questions. It can win where a customer is looking for one person to handle everything. We describe this kind of advantage in the entry USP.

What are we implementing on the website?

All the on-site tasks from the Aelo plan. These are small and medium tasks that do not require new articles:

• separate service pages: visual identity, print design and web graphics, with the scope, process and portfolio examples,
• a page on remote collaboration: what the process looks like, from the brief through the mock-up to implementation and ongoing support,
• a stronger page about the author: experience, number of projects, a photo and client reviews,
• structured data describing the company, the person, the services and questions with answers,
• address and local information on the pages for Bielsko-Biała,
• improvements to existing service pages instead of new subpages on the same topics.

The last point matters. A new subpage on the same topic as an old one scatters the signals and makes it harder for the model to choose a source, so it is better to strengthen the page that already exists. How to describe a company so that models read it unambiguously is covered in the article Identity, and the rules of structured data are explained in the glossary.

What are we doing off-site?

We create directory listings for the company on the directories and websites that models cite in this category, and tidy up the existing ones. The name, address, phone number and service description must read the same everywhere, otherwise a model may treat them as two different companies. For a brand that one AI engine has already confused with another, this is especially important.

The report also suggested collecting reviews from portfolio clients and guest articles on industry websites. These are bigger tasks, spread over the following months. We plan them in the module Tasks, and the brand information used by Aelo’s advisors is updated in Brand Knowledge.

What does the second measurement tell us?

• The number of articles does not measure AI visibility. 86 articles gave 0% mentions.
• Position on the list matters: a brand in position 12.8 is invisible to the customer.
• A model’s mistake is a sign that consistent company data is missing online.
• First we close the gaps and tidy up the data, only then do we write new articles.

How does Aelo help you find missing signals?

Aelo compares what is on the website with what models say about the brand, and shows whether the gap lies in the content or off-site. The measurement is done by the module Visibility in AI, the plan is created in Tasks, and for details you can ask the Advisors. How to read the results is described in the article AI Visibility: how to read your scan results.

Small businesses will find more on the page Aelo for small businesses, personal brands on the page Aelo for personal brands, and agencies that measure their clients, on the page Aelo for agencies. You can take your first measurement on the Free account, without a card.

Try it for free

FAQ

Sources

• Aelo, visibility reports
• Aelo, a methodology for measuring visibility in AI: https://aeloapp.io/wiedza/metodologia/