Sales Hire Scorecard

Sales hire scorecard

Runs in two modes. Build runs once, before the first interview, and produces a scorecard the whole panel agrees to in advance. Score runs after every interview and produces a ranked candidate list where every score points to the answer that earned it.

The prime rule: score the motion, not the logo. A rep who sold a famous product into inbound demand at a company everyone wanted to buy from has a different skill set from a rep who built pipeline from nothing for a product nobody had heard of. Both look great on a CV. Only one of them matches most hiring briefs. The scorecard asks how they sold, and to whom, before it asks where.

Template placeholders

Replace every {{...}} before enabling. The setup checklist in references/setup-customization-checklist.md covers each one.

• {{ATS}}: where the role, candidates and interview feedback live
• {{NOTES_TOOL}}: interview recordings or notes, if you record interviews
• {{REVIEW_CHANNEL}}: where the scorecard and summaries are posted
• {{HIRING_MANAGER}}: who owns the ranking and the final call
• {{PANEL}}: the interviewers who sign the scorecard
• {{CURRENCY}}: currency for deal sizes in the seat motion

Trial mode (recommended for the first role)

Build the scorecard and score candidates, but write nothing to {{ATS}}. Post every summary to {{REVIEW_CHANNEL}} marked TRIAL and let the panel run their usual process alongside it. At the end, compare who each process would have hired, and why. Switch writes on once the panel trusts the scores.

Diagnostic questions (before the first build)

Ask these once per team.

1. Think of your best and worst sales hires of the last two years. What did the interviews miss about each?
2. Where does pipeline in this seat really come from, as a rough percentage?
3. Does the panel agree what "good" looks like before meeting candidates, or after?
4. How often does the final call go to the candidate the most senior interviewer liked?
5. Which criterion do you suspect you over-weight because it is easy to see in an interview?

Build mode

Step 1. Pull the inputs

• From {{ATS}}: the role, the job description, the target start date and the panel.
• From the hiring manager: the numbers behind the seat. If they are not known, ask for them before going further. A scorecard built on a guess at the motion scores the wrong things.
• If the team has a past scorecard for a similar seat, load it as a starting point, never as the answer.

Step 2. Write the seat motion

One line each: average deal size in {{CURRENCY}}, sales cycle length, who the buyer is, where pipeline comes from, how complex the product is, and whether this is new business or growing existing accounts. This is the target every candidate is measured against.

Step 3. Pick six to eight criteria

From references/motion-fit-factors.md. Choose the ones where a mismatch would actually sink the hire. Drop anything a good rep learns in their first month.

Step 4. Force-rank and weight

No ties. The hiring manager decides the order. Then spread 100 points across the criteria in rank order using references/scoring-method.md, and mark up to two must-haves. The weights belong to the team running the hire, because they encode what this seat needs. There is no universal answer.

Step 5. Assign questions and owners

Each criterion gets one primary question and one follow-up from the factor library, and one interviewer who owns it. No criterion goes unasked and none is asked three times.

Step 6. Get it signed and lock it

Post the scorecard to {{REVIEW_CHANNEL}} with a Yes/No: "Agree scorecard v1 before the first interview?" Every {{PANEL}} member answers. Once agreed, it gets a version number and is locked. A change after interviews start creates a new version, and every candidate scored so far is re-scored against it.

Score mode

Step 7. Pull each interviewer's evidence

From {{ATS}} and {{NOTES_TOOL}}: each interviewer's written notes and scores for their owned criteria, with timestamps. Scores entered after a group discussion are tagged (group).

Step 8. Score against the anchors

1 to 4 per criterion, against the anchors in the scoring method. Every score cites the answer or example that earned it. A hypothetical answer caps at 2. No evidence, no score: mark it not assessed and assign the question to the next interviewer. Where two interviewers scored the same criterion two or more points apart, flag the gap first.

Step 9. Rank, flag and post

Weighted total per candidate, plus any must-have scored 1, which fails regardless of the total. Post the summary and stop.

[Role] / scorecard v[n] / agreed by [panel] on [date]
Seat motion: deal [size] | cycle [length] | buyer [level] | pipeline [source] | [new business/expansion]
Criteria (rank, weight): 1. [ ] [w] 2. [ ] [w] ... | Must-haves: [ ]
Candidate | weighted score | must-have fails | not assessed | top evidence
[A]       | [ ]            | [ ]             | [ ]          | "[quote]"
[B]       | [ ]            | [ ]             | [ ]          | "[quote]"
Scorer gaps: [criterion, interviewers, scores] or none
Decision: Advance [names] to [next round]? Yes / No
          Fill the not-assessed gaps before deciding? Yes / No

Step 10. Write back, after the decision

Only after a human answers:

• Log each candidate's scores and cited evidence as a note in {{ATS}}, with the scorecard version.
• Move stages only for the names the human said Yes to.
• Read every write back. Not found means retry once, then post a visible warning: "ATS write failed, manual entry needed".

Write failures never block the summary.

Batch mode

Score mode handles a whole shortlist in one run. Rank only candidates scored on the same scorecard version. A candidate with not-assessed criteria is shown with those gaps, never ranked as if a 0 were a real score. Totals within five points are a tie, broken on the top-ranked criterion.