What did we change?
The rebuild answers each cause found in the before score.
- Both services now have their own pages, written around the questions buyers ask: AI visibility audit, generative engine optimization and custom AI development.
- Thirteen answer pages each take one buyer question as the title and answer it in the first paragraph. All of them are on the answers page.
- The home page and the About page now carry one description of the company in their code, so the engines read the same facts each time.
- Claims we could not support with evidence were removed, including the old list of seven "systems in production".
What was the before score?
On 20 October 2026, before the rebuild, we asked 28 questions that buyers use when they look for AI visibility work or a custom AI build. Fourteen were about being found in AI answers, such as "AI visibility audit" and "GEO agency Toronto". Fourteen were about AI builds, such as "custom AI agent development Toronto" and "build vs buy AI agent". Each question went to ChatGPT, Claude and Gemini with web search available, and to Google Canada, with Toronto set as the location for questions that named a place. Every answer was saved.
The separate run used the audit engine Kin-etiq.ai runs for clients, with questions such as "Best AI agency Toronto?".
Where was Kin-etiq.ai named before the rebuild?
Once. ChatGPT, asked "build vs buy AI agent", listed Kin-etiq.ai as a Toronto option and cited the home page. That question matched an article on the old site, but ChatGPT cited the home page, not the article. No answer cited any of the old site's four articles, and no other answer named Kin-etiq.ai.
Why was Kin-etiq.ai missing?
The evidence points to four causes.
The old site did not describe the AI visibility service at all. No page used the words "AI visibility", "GEO" or "AI search", so none of the 14 visibility questions had a page to land on.
The pages were named after products. Six of the seven system pages used a product name as the title and main heading, and buyers search with questions. None of those product names appears in any of the 28 buyer questions.
The engines were not sure who Kin-etiq.ai is. A Google search for "Kin-etiq" showed other brands with similar names in all nine regular (unpaid) results. Claude could not identify the company in 5 of 6 questions and twice described a different firm.
The answers were built from pages Kin-etiq.ai was not on. Asked "Best AI agency Toronto?", ChatGPT and Claude both built their answers from Clutch's Toronto rankings. Whether Kin-etiq.ai had a Clutch profile was not checked in this audit.
What happens next?
On or after 20 November 2026 we will re-run the same 28 questions on the same engines and publish the result here with its date. The number to watch is the share of answers over several runs, because AI answers change from one run to the next.
If you want the same audit on your business, start with the free assessment, a Level 1 check of your website.