Test case
We asked ChatGPT who to hire in Montreal. Our own name wasn't in the answer.
Here's a test we ran on ourselves, because honesty is the product. We put a real buyer's question to ChatGPT, Claude and Gemini: which agency should a Montreal business hire to show up in AI recommendations? The assistants answered confidently — and named three competitors. Be Preferred, our own company, appeared zero times.
We recorded every answer, dated it, and filed it as our baseline. Our own report card gave us an F on "found by AI assistants." We publish that grade because it teaches the most important lesson in this field: nobody is entitled to be recommended. Not even the people who do this for a living. The assistants name businesses whose track record is written down where machines can read it — and until recently, ours wasn't.
So we're doing for ourselves exactly what we do for clients: stating the facts of the business in machine-readable form, publishing pages that answer the questions buyers actually ask, and measuring again on a schedule. Every measurement is recorded, so the day our name appears in that answer, we'll be able to show the before and the after, with dates.
That's the whole method. Measure honestly, fix what the measurement shows, measure again. If an agency won't show you its own baseline, ask why.