In short
A workable startup scorecard has four criteria, each with a written definition of what good evidence looks like, scored independently by each interviewer before anyone discusses the candidate. It exists to stop the loudest opinion in the debrief deciding the hire, and it takes about thirty minutes to write per role.
Key facts
- Criteria per role
- Four, agreed before the first interview
- Scoring
- Independently, before the debrief
- Decision rule
- Written down in advance
- Setup time
- About 30 minutes per role
Why unstructured interviews fail small teams
Without structure, interviews drift towards whoever the panel found most enjoyable to talk to. That is a weak predictor at the best of times and a dangerous one when a single hire is a meaningful share of your team.
Structure does not mean bureaucracy. It means everyone assesses the same things, writes down evidence, and commits to a score before hearing anyone else's view.
Build the scorecard in four moves
Pick four criteria
Typically: core technical depth, problem ownership, collaboration, and one criterion specific to this role's hard part.
Define the evidence
For each criterion write one sentence describing what a strong answer contains. This is the part teams skip and the part that does the work.
Assign criteria to stages
Each interviewer owns one or two criteria. Nobody assesses everything, so nobody assesses anything shallowly.
Agree the decision rule
For example: no hire if any criterion scores below the bar, regardless of the average. Write it down before you meet candidates.
A four-point scale that avoids fence-sitting
An even number of points forces a direction.
| Score | Meaning | Evidence required |
|---|---|---|
| 1 | Below the bar | Specific gap, described |
| 2 | Mixed | Something works, something does not — name both |
| 3 | Meets the bar | Concrete example from their own work |
| 4 | Clearly above | Example plus depth under follow-up questioning |
Rules that keep it honest
- Submit scores before the debrief, not during it.
- Every score cites evidence from the interview, not an impression.
- The same questions go to every candidate for the role.
- Feedback goes to the candidate quickly, whichever way the decision goes.
- If two interviewers disagree sharply, that is a signal to test again — not to average.
Common questions
- Is a take-home exercise worth it?
- Only if it is short, paid or clearly time-boxed, and reviewed against the same criteria for every candidate. Strong candidates with options will decline anything long or vague.
- How many interview stages should a startup run?
- Usually two or three. More than that and you are testing patience rather than capability, and you will lose candidates who are in other processes.
- Should the founder be in every interview?
- For the first handful of hires, usually yes — but on a defined criterion, with a score like everyone else, not as a veto held in reserve.
Turn this into a brief
Answer six short sections and you have a technical hiring brief you can share internally or send straight to us.
Open the hiring brief builderAny percentages or figures shown in this article are illustrative examples used to explain the model. They are not quoted rates, market benchmarks or salary data.