Try it free
  1. Home
  2. Help
  3. The numbers in a role

The numbers in a role

Updated

A role's candidates narrow through five numbers, widest first. You'll find them in the Talent pool in the right-hand panel and on the tabs of the List view, and each word means one thing only.

An example, a senior controller in Aarhus: about 1,840 matches, 986 fetched, 207 on the longlist, 41 recommended and 9 on the shortlist.

Matches

How many profiles in the market fit the filters, before anything is fetched. It comes from the preview, which is free, and it's always shown as an estimate: "about 1,840 matches". Remove a filter and you get a new figure without using a search.

Fetched

The unique candidates fetched for the role: about 1,000 per search. The count runs live: "Fetched 640 of about 1,000".

  • Duplicates. The same person can appear more than once in search data. A duplicate, meaning the same person or the same LinkedIn profile, is dropped before it's stored or counted. While the market has more, Navigent fetches replacements at its own cost, never for another search.
  • Candidates you already have. Someone the workspace knows from another role isn't a duplicate. They're reused and they count.
  • People who are never fetched. Anyone who has asked to be removed from Navigent, and the people and companies on your do-not-contact list, are left out.
  • A smaller market. If the market has fewer, the search fetches what there is, and it counts as usual. If it stores no candidates at all, you get the search back.

Replacements are capped per run, so Fetched can land a little under 1,000, at 986 say.

Longlist

The best of the fetched candidates after the gates and the ranking: about 200.

The gates are the four kinds of must-have that can rule someone out, when the ad asks for them, and each has some tolerance. The function has to appear somewhere in the job history. Seniority can be at most one level off. Experience can be at most 12 months short of the minimum. The candidate has to be in the search area and, if the ad states a commuting distance, within it. Missing data never excludes anyone, and industry and nice-to-haves set the order without ruling anyone out. Whoever a gate did exclude sits under Excluded in the List view, with the reason.

The number isn't fixed. Everyone who scores as high as number 200 makes it, so a tie at the cut-off gives you a slightly longer list, 207 for instance. If fewer than 200 pass the gates, the longlist is simply those who did.

The genuinely relevant at the top of the longlist, up to about 50, assessed requirement by requirement first. To be recommended, a candidate needs the band Strong or Good and a clear fit on function, so a high score from nice-to-haves alone is never enough.

It's never a quota. If 23 fit, 23 are recommended. If none do, the thread says so and offers you the longlist instead. The rest of the longlist stays visible, and you can have it assessed 50 at a time, free.

Shortlist

Only the candidates you've said yes to. Nothing else is called the shortlist.

When the numbers move

The longlist and the recommended are worked out again whenever the candidates are re-ranked, after calibration, say, or when you change a preference in the chat. So a candidate can move on or off the list. A re-rank never changes a decision, and an assessment stays put.

The tabs in the List view show the same numbers as the Talent pool for the same words.

Why they say "about"

The market decides how many people there are, so Navigent never promises an exact count. Numbers about the process say "about" and "up to". A count of something that has already happened, like "Fetched 640", is exact.

Good to know

  • About 1,000 more for the same role costs one more search, once you've confirmed. The new run picks up where the last one stopped and skips anyone the role already holds.
  • The roles list shows each role's longlist and shortlist, and how many searches it has used.