You can guess whether AI recommends you, or you can find out. Guessing is comfortable and usually wrong. The good part is that checking takes very little time once you have a simple routine, and the routine matters more than any single result.
The most common mistake is asking the AI about yourself. "Is my company good" tells you nothing, because a real buyer never asks that. They ask category questions. Your job is to write down the ten or so questions a genuine customer would type before they know your name exists. For example:
These are the moments where a recommendation actually changes a decision. Those are the ones worth tracking.
ChatGPT, Claude, and Gemini pull from different mixes of sources and will not agree. You might be recommended by one and absent from another. Checking only your favorite gives you a misleading picture. Run the same questions across all three so you see the full spread.
One assistant is an anecdote. Three assistants, tracked over time, is data you can act on.
Consistency is what turns spot checks into a trend. For each question and each assistant, note:
That last one is quietly the most useful. Over a few checks you learn exactly who the AI thinks your rivals are, which is often not the list you would have written yourself.
AI answers move. Models get updated, the web changes, and a mention you had last month can disappear without warning. A single check is a snapshot. Running the same set every few weeks turns it into a trend line, and the trend is what tells you whether your work is paying off. Put a recurring reminder on the calendar so it actually happens.
You can run all of this by hand, opening each assistant and copying answers into a spreadsheet. It works, but it is tedious and easy to abandon after week two. The point of a tool is to remove the friction so the habit survives: same questions, all three assistants, results laid out side by side, position and description scored for you. However you do it, the discipline is the same. Ask real questions, record the same fields, and repeat.
Run your questions across ChatGPT, Claude, and Gemini and get the results in one report.
Check my AI visibility