The company has already published its own roster
When you know which firm you want people from, the usual move is a company search in a people database. For a large employer that works. For the firms whose team pages recruiters tend to paste, the boutique consultancy, the owner-led engineering practice, the small specialist, the databases are thin or out of date. Some of the team is missing and the titles lag behind.
Meanwhile the company has printed the answer itself. A team page is the firm's own statement of who works there and what each person does, and it is more current than a database row. So this play does not search for the roster. It reads it, and then does the two things a page cannot: it works out which profile belongs to each name, and it finds out what each person has really done.
It is the narrow sibling of source from named competitors, which searches inside the companies you name. Use that one for larger firms. Use this one when the page is better than the data.
What you get back
A CSV with one row per person on the page: name and title as printed, the LinkedIn profile and how it was matched, the dated work history, the score against your brief with the evidence quoted, and a personal email for the people who passed.
With it comes a short note of what could not be read: pages behind a login, names held in images, and people listed without a title. Nothing is sent and nothing is written to your ATS.
Variations worth knowing
One firm, several pages. Bigger teams are split by department or spread over numbered pages. The site is mapped first, so the whole roster is read and not only the first screen.
Names you typed yourself. If a page cannot be read, paste the names and titles. Everything after the page read runs the same way.
Profile links only. If all you want is a link against each name, add the LinkedIn profile URL to a list is that job by itself.
Where this goes wrong
Login walls and picture rosters. They are reported and left alone. The play reads public text and nothing else.
A page with nobody on it. A stock-photo or contact page lists no one. The agent falls back to a search inside that company's domain and tells you.
Bios read as experience. A blurb that claims deep sector expertise is copy. Sector experience is only ever taken from dated history.
People who have left. A page can lag behind a resignation. The dated history shows the current employer, and a mismatch with the page is flagged, not hidden.
Leadership-only pages. Large companies publish the top table and not the team. You get those names and no more, so a company search is the better tool there.
