Unstructured interviews predict early-career performance poorly, and they predict it worst exactly where you need them most: candidates with no work history give an unstructured interviewer nothing to anchor on except confidence and polish. A scorecard fixes the anchoring problem, and it does something else that matters at volume — it makes two interviewers comparable.
The template below is deliberately short. Scorecards fail when they have fourteen competencies and nobody fills them in.
The five competencies worth scoring
More than five and interviewers stop completing the form; fewer and you cannot distinguish candidates. These five cover what actually varies at intern and fresher level.
| Competency | What you are measuring | What it is not |
|---|---|---|
| Problem solving | How they approach an unfamiliar problem and choose between approaches. | Whether they got the right answer. |
| Craft in the role | Depth in the thing the job needs, and awareness of their own limits. | How many technologies they can name. |
| Communication | Whether a listener can follow them, and whether they adapt to who they are talking to. | Fluency in English, or confidence. |
| Ownership | Whether they own outcomes, including the parts that went wrong. | Enthusiasm. |
| Coachability | Whether they use feedback within the interview itself. | Agreeableness. |
The scorecard — copy this
INTERVIEW SCORECARD Candidate: ____________________ Role: ____________________ Interviewer: __________________ Round: ___________________ Date: __________ Score each competency 1–4 using the anchors. Score BEFORE the debrief. Evidence is mandatory — a score with no evidence is an impression. 1. PROBLEM SOLVING Score: [ ] 1 — Waits to be told the approach 2 — Finds an approach with prompting 3 — Chooses an approach and explains the trade-off 4 — Reframes the problem into a better one Evidence: ______________________________________________ 2. CRAFT IN THE ROLE Score: [ ] 1 — Recites definitions 2 — Applies what they were taught 3 — Applies it to an unfamiliar case 4 — Knows the limits of what they know and says so Evidence: ______________________________________________ 3. COMMUNICATION Score: [ ] 1 — Hard to follow 2 — Clear with effort from the listener 3 — Clear and structured unprompted 4 — Adapts depth to the listener Evidence: ______________________________________________ 4. OWNERSHIP Score: [ ] 1 — Attributes outcomes to others 2 — Owns the task 3 — Owns the outcome including the failures 4 — Fixes the cause, not just the instance Evidence: ______________________________________________ 5. COACHABILITY Score: [ ] 1 — Defends the first answer 2 — Accepts feedback 3 — Uses feedback within the interview 4 — Seeks the disconfirming view Evidence: ______________________________________________ OVERALL RECOMMENDATION (circle one) Strong hire | Hire | No hire | Strong no hire The one thing that most influenced this recommendation: _______________________________________________________ What I would want the next interviewer to probe: _______________________________________________________
How to write good anchors
- Describe behaviour, not quality
- "Explains the trade-off unprompted" is scoreable. "Good communication" is not — every interviewer means something different by it.
- Make level 3 the bar, not level 4
- If 4 is your hiring bar you will hire almost nobody and your interviewers will inflate. 3 should mean "yes, hire".
- Make level 1 real, not insulting
- Level 1 should describe something you have actually seen a candidate do, so interviewers use it instead of defaulting to 2.
- Keep anchors identical across interviewers
- The entire value is comparability. An interviewer who rewrites the anchors has opted out of the system.
The debrief rule
Every interviewer submits their scores and their recommendation before the debrief starts. If scores are discussed first, the most senior or most confident voice sets the anchor and everyone else converges on it — the effect is well documented and it is strong. Read the independent scores aloud, then discuss the disagreements, which is where the information actually is. A unanimous panel that talked first has told you very little.
Rejecting well
- Decide the rejection wording before the process starts, so nobody improvises under time pressure.
- Reply to everyone who reached a screening call, within a week. Silence is what candidates talk about publicly.
- Offer one line of specific feedback to anyone who reached a final round — it costs a minute and it is remembered.
- Never say "we will keep your CV on file" unless you actually will.
Frequently asked
Why score before the debrief?+
Because groups anchor. If the first person to speak is senior or confident, the rest converge on their view, and the panel produces one opinion held by four people rather than four independent readings. Independent scores first, then discuss the disagreements — the disagreements are the information.
How many competencies should a scorecard have?+
Five is the practical maximum for early-career hiring. Longer scorecards get filled in less honestly and less often; the fourteen-competency form is always the one interviewers complete from memory afterwards.
Should we score candidates on "culture fit"?+
Not under that name. "Culture fit" reliably becomes similarity to the interviewer, which narrows teams without improving them. If there is a genuine behavioural requirement — collaborating across teams, handling ambiguity — name it as a competency with behavioural anchors and score that.
Do scorecards slow the process down?+
They slow the debrief and speed everything else: fewer re-interviews, faster decisions because the evidence is written down, and defensible rejections. At campus volume they are the difference between deciding and re-litigating.
Can we share scorecards with candidates?+
Share the substance, not the form. A candidate who reached a final round benefits from one specific sentence about what was strong and what was not. Sending the raw scorecard invites a debate about the number rather than a conversation about the gap.