How scores and fit bands work
Scout gives each candidate a suitability score from 0 to 100 for each job, and puts it in one of four fit bands. The score is there to help you decide who to look at first. It never decides anything on its own.
The four fit bands#
| Band | Score |
|---|---|
| Strong | 85 and above |
| Good | 70 – 84 |
| Possible | 50 – 69 |
| Weak | Below 50 |
| Not scored | No call has scored them yet |
On a call these read as Strong fit, Good fit, Possible fit and Weak fit. On a job’s board, the (i) beside the fit filters opens the same key (What the colours mean).
A missed requirement caps the band#
Some answers are checked against hard requirements the job states: the notice period, minimum years of experience, and work mode. If a candidate’s answer misses one, their band is capped at Possible, whatever the number. The score itself is unchanged, so you can still see how they did everywhere else.
What the score is built from#
The score comes from what the candidate says on their calls. Each call is analysed, and every area the job’s agents are set up to assess becomes part of the score:
| Area | Comes from |
|---|---|
| Screening | The answers to the job’s screening questions |
| Technical | A technical question block, when the agent has one |
| Communication | A communication probe, when the agent has one |
| Custom areas | Each custom question block set to assess |
For each area Scout uses the latest call at each stage, averaged across stages. An area no call has assessed yet counts as “not assessed”, never guessed, and the score is worked out over the assessed areas only.
Where the resume fits in#
Before the first call there is no score, only a resume check: the candidate panel says Not screened yet and shows Resume check — not a score. The resume match can hold back the first call (see Set up a job for screening), and the resume can raise flags for you to look at, but it doesn’t add to the number.
The score mix: how areas are weighted#
Each job weighs the areas differently. Scout suggests weights from the job description; you can change them in the job’s Job setup tab under How it’s scored; the Score mix tile at the top of the tab sums them up. A button beside the heading puts the suggested weights back. The candidate panel’s How the score is made shows each area’s weight and score for that candidate.
How each call is scored#
A call’s analysis has a section for each kind of question the agent asked: Fit for screening questions, Communication and Technical when the agent used those blocks, and one entry per custom block. Each section is scored from 1 to 5 and comes with strengths, concerns and quoted moments, so you can check the reasoning. Open it from Full report on the call. See Review a screening call.
A call that isn’t linked to a candidate on a job (a preview, say) shows its own Fit score on the same 0–100 scale and says it is based on “this call”.
With the scores, Scout writes a summary, what counted for and against the candidate (with the call each point came from), follow-up questions, and flags such as a gap or a claim to check. The model writes these words; the numbers come from the analyses and the job’s weights.
Score or re-score candidates#
- One candidate: open them on the job and click Re-score. Scout runs again even if nothing changed.
- A whole job Scout doesn’t screen itself (an ATS job with no workflow): click Score candidates on the job’s Candidates tab. Anyone already scored on unchanged information is skipped.
- One call: Score this call, or Try again after a failure, on the call.
Scores also update on their own when a new call is analysed. When a follow-up call changes an answer, the call says Re-scored on the answers this call changed.