The short answer
A structured interview scorecard gives every candidate the same questions, in the same order, rated on the same anchored 1-to-5 scale. Build it from the role’s first-year outcomes: pick five or six competencies, write two questions for each, and define what a 1, a 3 and a 5 answer sounds like. Add one or two knockout criteria for the CV screen. Each interviewer owns specific competencies and writes their scores, with one line of evidence each, before the debrief.
Why unstructured interviews fail
In an unstructured interview, each interviewer asks their own questions and forms an overall impression. Four interviewers produce four impressions of four different conversations, and there is nothing to compare. The candidate who presents best usually wins, which is a measure of presentation, not of the job.
A structured interview asks every candidate the same questions and rates each answer on an anchored scale, where each score point is defined by an observable behaviour. Two interviewers hearing the same answer then give the same score, and candidates can be compared on evidence.
The parts of a scorecard
| Part | What it is | Where it comes from |
|---|---|---|
| Knockout criteria | One or two must-haves checked on the CV | The outcome the role cannot deliver without them |
| Competencies | Five or six skills the role needs | The role’s first-year outcomes |
| Questions | Two per competency, about past behaviour | Situations the role will face |
| Anchors | What a 1, 3 and 5 answer sounds like | The difference between weak and strong evidence |
| Owner | One interviewer per competency | Who is best placed to judge it |
| Evidence line | One sentence for every score | What the candidate actually said |
How to build one in six steps
1. Start from the outcomes, not the job title
Write what the hire must achieve in the first year, each with a baseline and a date. The competencies follow from the outcomes.
2. Set one or two knockout criteria
A knockout is a requirement the CV can confirm and the role cannot do without. Record, for every rejected CV, which knockout it failed.
3. Choose five or six competencies
More than six and interviewers stop assessing each one properly. Leave out anything that cannot be tested in an interview, such as “strategic leader”.
4. Write two questions for each
Ask about what the candidate did, not what they would do: “Tell me about a quarter when retention fell. What did you change?”
5. Write the anchors
Define a 1, a 3 and a 5 in terms of what the answer contains. A 1 describes general effort; a 5 names the metric, the cause and the result.
6. Assign each competency to one interviewer
So every competency is assessed in depth once, rather than lightly by everyone.
Worked example: Orrin
The scenario
A $287,000 mis-hire, and four interviewers who all liked him
Orrin sells revenue-analytics software, with $42 million of annual recurring revenue and 240 staff. In January 2026 it hired a Head of Customer Success on a “10+ years in SaaS, strategic leader” description. He left seven months later. By the time a replacement starts, the mis-hire will have cost $287,000, net revenue retention has fallen from 108% to 97%, and 3 of the 9 customer success managers have resigned.
Four people interviewed the last hire. No two asked the same question, all four rated him highly on impressions, and there was nothing to compare. The new scorecard starts from the role’s outcomes: net revenue retention from 97% to 105% by Q4 2027, and customer success attrition from 33% to under 15% a year.
Two knockouts screen the CVs: has led a customer success team of 8 or more, and has owned a retention number with a before and after. 14 CVs came in; 5 passed both. Then six competencies, two questions each. Three of them:
| Competency and owner | Question | 1 | 3 | 5 |
|---|---|---|---|---|
| Retention diagnosis (COO) | “Tell me about a quarter when retention fell. What did you change?” | Describes the team’s effort | Names the metric and one cause | Names the metric, the cause and the result |
| Team leadership through change (Head of Talent) | “Tell me about a team you took over after people had left.” | Describes hiring replacements | Describes what changed for the people who stayed | Shows attrition before and after, with the changes made |
| Renewal negotiation (CFO) | “A top account asks for 20% off to renew. What do you do?” | Offers a discount | Trades terms before price | Tests the reason, trades terms, and knows the walk-away |
Scoring and the debrief
Scores are written independently, before the debrief. Once a senior person states a view in the room, other interviewers’ scores move towards it. At Orrin, the CEO already preferred Candidate A. Written before the debrief, the panel’s scores averaged 3.4 for A and 4.2 for B, putting B ahead on five of six competencies. The CEO’s preference rested on presentation, which the scorecard does not measure.
A spread of 2 points between interviewers on the same competency is a question, not something to average. Two interviewers scored A’s team leadership 2 and 4; the debrief found one had asked about a team of 3 and the other about a team of 12. References were then asked to confirm specific figures: B’s claim of cutting churn from 14% to 8% was confirmed by a former CFO, and A’s retention result could not be confirmed by either referee.
Try it: a shortened Prompt 04
This is a shortened version of Prompt 04 from the HR & Talent pack, set up to run on its own. Paste it into ChatGPT, Claude or Gemini and fill in the brackets.
You are a former head of talent who has designed structured hiring for leadership roles in growth-stage companies. An interview only predicts performance when every candidate gets the same questions and every answer is scored against defined behaviours, so you build the scorecard from the role's outcomes, never from a list of desirable traits. THE ROLE: - Title, team and who it reports to: [details] - What the hire must achieve in the first year: [outcomes, each with a baseline and a date] - Why the role is open: [new, backfill, what went wrong last time] THE PROCESS: - The interviewers and their roles: [names or titles] - Number of CVs expected, if known: [figure] Produce: 1. THE KNOCKOUTS. One or two must-haves the CV can confirm, each tied to an outcome. 2. THE COMPETENCIES. Five or six, each tied to an outcome, with one interviewer as owner. 3. THE QUESTIONS AND ANCHORS. Two past-behaviour questions per competency, and what a 1, a 3 and a 5 answer contains. 4. THE SCORING RULE. How scores are recorded before the debrief, with one line of evidence each, and what to do with a 2-point spread. Leave out any requirement that cannot be tested in an interview, and say which of my inputs you removed and why.
The full Prompt 04 has eight parts rather than four, and it reads the Talent Brief built by prompts 01 to 03, so the scorecard uses the outcomes and requirements already written into the job description.
Checks before the first interview
- Every competency ties to a first-year outcome.
- Every question asks about past behaviour.
- Every competency has anchors at 1, 3 and 5.
- Each competency has one owner on the panel.
- Scores are written before the debrief, with one line of evidence each.
Questions
What is an interview scorecard?
A form that lists the competencies a role needs, the questions for each, and a defined scale for rating the answers, so every interviewer assesses every candidate on the same evidence.
What is an anchored rating scale?
A scale where each score is defined by what the answer contains, rather than by a word such as “good”. A 5 might mean “names the metric, the cause and the result”; a 1, “describes the team’s effort”.
How many questions should a structured interview have?
Two per competency, for five or six competencies: ten to twelve in total, split across the panel so each interviewer covers their own competencies in depth.
Should interviewers discuss candidates before scoring?
No. Scores and evidence are written independently first. Discussion before scoring moves everyone towards the most senior person’s view.
Written by a former Gartner Managing Partner and investment banking SVP
The guides and the prompt templates on this site come from a career spent building these documents: board decks, forecasts, business cases and hiring decisions, as Managing Partner at Gartner, SVP in investment banking and Country Manager at international subsidiaries. The worked examples are published in full on each product page. Browse the Prompt Library.
Orrin and the candidates described are fictional, and the figures are illustrative.
Employment law differs by country and state; nothing here is legal advice. Output from any AI tool should be reviewed before use.