Asking the agent for what you want
You describe the people you want in plain language, and the habits that make the answer better are exclusions, specific sources and correcting the agent in the same chat.
You describe the people you want in plain language. There is no query language, no filter builder and no syntax to learn, so the useful skill is not phrasing. It is being precise about who counts and who does not.
Say who to exclude
This is the highest-leverage habit on this page. Noisy lists are noisy for predictable reasons, and one sentence of exclusions removes those people before any work is spent on them.
| Exclude | Why they show up |
|---|---|
| Recruiters and agencies | They match almost any role or industry description |
| Students and interns | Their titles read like the job you are targeting |
| Freelancers and consultants | They list many companies, so they match broad criteria |
| Competitors | They follow the same topics and posts you do |
Find heads of operations at mid-sized logistics companies in Europe.
Exclude recruiters, agencies, students, interns and freelancers.
Be specific about the source, not the syntax
A good source beats a clever phrasing. Naming a post, a company or a hashtag gives the agent a real starting set of people instead of an approximation of one.
Take everyone who commented on this post, keep the ones who run a support team, and tell me who you dropped.
Argue with the result
The agent tells you why it kept or set aside each person. When it gets one wrong, say so in the same chat. The correction carries into the rest of the list, so you are fixing a rule rather than editing rows one by one.
You set aside the agency owners. Keep them if they run delivery in-house, and re-check the rest of the list on that basis.