Here's the answer, and it's the actual instrument behind the QYRAI™ Diagnostic, published in full. Write 25 questions in your buyers' language, five per category: direct (what is this brand, is it legit), best-of (the money prompts: are you named unprompted), problem-solution (where journeys start), comparison (do you survive the conversation), and trust (whose evidence gets cited). Run all 25 on ChatGPT, Gemini, Perplexity, Claude, and Grok in clean sessions. That's 125 responses, recorded verbatim.
Score each response on four dimensions from 0 to 3: presence, accuracy, sentiment, and evidence. Weight by commercial value, best-of counts double, comparisons 1.5 times. The weighted average scales to a 0 to 100 Primacy Score™, banded Fragile, Moderate, or Strong. Score the founder and the brand separately, because the gap between them is often the headline finding.
The integrity rules keep the number honest: clean sessions, the same prompts verbatim every run, full responses recorded, never cherry-picked, everything date-stamped because models change monthly.
Then the low scores translate into named gaps, each with a fix: known but not chosen, founder authority not transferring to the brand, competitors' evidence being cited instead of yours, AI describing you wrong, or losing every comparison. Your lowest-scoring, highest-weight categories set the build order.
And one rule above all: never promise a score, to a client, a boss, or yourself. Anyone selling guaranteed AI rankings is selling weather.
The full method is Volume 3 of our Primacy Playbook.
You can't manage what you haven't measured.