← PersonRank blog

Aim it at the job: getting AI to see you for the role you want

AIs answer differently depending on how the question is framed, so the useful test is how you look for a specific role, industry and place.

By PersonRank5 min read

GlobalRegionCountry

To see how AI sees you for a specific job, ask the way a recruiter would: name the role, the industry and the place, and test at country, region and global scope. A role-specific question draws on different evidence from "tell me about this person", and a single comparison proves little, so repeat it in varied wording and order.

Ask an AI "Tell me about this person" and you get a general profile. Ask "Is this person a strong candidate for a chief financial officer role at a mid-sized manufacturer in Germany?" and you get a different answer, built from different evidence. The framing changes what the AI looks for.

That is useful. It means you can test yourself against the job you actually want, rather than against a vague idea of "being visible". Here is how.

Start with the job posting

Take a posting for the role you would like, and read it for requirements rather than adjectives. Most postings contain a long list, but only four to six requirements really matter. They are usually the ones that appear in the first lines, repeat in different words, or are tied to a specific outcome, such as "has led a post-merger integration" or "experience selling into regulated markets".

Write them down as plain statements. Then, for each one, ask a blunt question: if someone searched the public web, what would they find that shows I have done this?

  • A leadership page that names the role and scope.
  • A press article describing the project.
  • A conference talk on the topic.
  • A register entry showing a board seat.

Where a requirement has no evidence you can point to, that is a gap in the public record, even if it is not a gap in your career. Closing it usually means publishing something, not changing something.

Scope changes the answer

"Among the strongest candidates" is a claim that only makes sense inside a boundary. An executive who is well known nationally may be unknown in the next region and invisible globally. The reverse also happens: someone with a strong international record may not be recognized at home.

So test at several scopes: your country, your region, and globally. If you are named at home but not beyond, the next step is evidence that travels: international speaking, cross-border press, a profile on a pan-regional organization's site. If you are named globally but not nationally, local coverage may be the missing piece.

Comparisons need citable evidence

When an AI compares you with peers for a role, it can only weigh what it can find and cite. A strong peer with five well-sourced facts will usually beat you if your best achievements live only in your head, or on a page nobody links to. A win in a comparison that rests on no cited source is fragile: it can flip the next time the question is asked.

One comparison proves little

Here is a caution we would rather be honest about. In head-to-head questions, AIs can favor the name that is mentioned first. Ask "Who is stronger, A or B?" and then "Who is stronger, B or A?", and the answers may not agree. So a single comparison, however flattering, tells you very little.

What does tell you something is repetition and balance: the same question asked several times, with the order swapped, in different wording, across more than one AI. If you come out ahead consistently, that is a signal. If the result flips with the order, treat it as a draw. And because AIs change over time, repeat the check every few months.

An example of the gap between career and record

Imagine a senior finance leader who spent two years leading a cost program across several countries. Their career clearly contains that achievement. But if the only public mention is a one-line job title, an AI asked about "cost transformation experience" has nothing to cite. A short interview, a conference panel or a clear dated line in a bio would fix that. The experience was always real. What was missing was a page that said so.

Notice what this does not require: a bigger reputation. It requires that the reputation you already have is written down somewhere a reader, human or machine, can find and cite.

A simple way to work

  1. Pick one target role, in one industry, in one place.
  2. Extract the four to six requirements that matter from a real posting.
  3. For each one, find the public page that shows your evidence. If none exists, note it.
  4. Choose the peers you are realistically compared with, not the most famous names in the field.
  5. Ask balanced questions at country, region and global scope, and swap the order.
  6. Fix the biggest evidence gap first, then check again.

Where PersonRank fits

This is the workflow PersonRank is built around. You set the target: the role, the industry, the peers you want to be measured against, and, if you have one, the job posting itself. PersonRank then asks ChatGPT, Gemini and Claude the questions a recruiter would, checks the facts in their answers, shows whether you are named among the strongest candidates in your country, region and globally, and suggests concrete next moves.

It is a measurement, not a promise. Nobody controls what an AI says. But knowing where you stand, and why, is far better than guessing.