An interview scorecard template is a one-page sheet that lists four to six job-related competencies, rates each on a 1-5 scale against written behavioural anchors, and makes the interviewer note the evidence behind every score. Fill it in before you discuss the candidate with anyone and gut feel stops running your hiring. Structured interviews are the strongest predictor of job performance among common selection methods, with a validity of .42 (Sackett et al., 2022, cited by Pin), yet only two-thirds of employers said they delivered a structured interview process in Talent Board's 2024 research (via ERE). The rest are guessing. Here is the sheet.
- Structured interviews have the highest predictive validity of common selection methods at .42 (Sackett et al., 2022), yet only two-thirds of employers ran a structured interview process in Talent Board's 2024 research.
- A working scorecard has four to six job-related competencies, a 1-5 scale with written behavioural anchors, and a mandatory evidence line for every score.
- Scores go in before the debrief and within 24 hours; Metaview's data on 811,298 scorecards shows nearly one in twenty arrives after a week and stalls the decision.
- Evidence notes are personal data under GDPR and the UAE PDPL, so write only job-related facts a candidate could read back.
Why does gut feel fail even with experienced interviewers?
Because experience makes you confident, not consistent. The same manager rates a chatty candidate higher at 9am than at 5pm, and two managers leave one interview with two different hires in mind. A 2006 meta-analysis by Aamodt and colleagues, summarised by Pin, found bias effects of d=.59 in unstructured interviews against d=.23 in structured ones. Not zero, so no scorecard eliminates bias, but far smaller.
Greenhouse's March 2026 benchmark of 6,000+ North American companies puts applications per role at 244 in 2025, up 111 per cent from roughly 115 in 2022, while recruiters per organisation fell 55.6 per cent. Each interview now carries more of the decision. And leaders admit it: 71 per cent of the 1,600 C-suite and senior HR leaders at large organisations in Korn Ferry's 2026 Global Talent Analytics Survey said they rely on instinct over insight.
What goes on a one-page interview scorecard?
Four things, on one page, so it gets filled in during the interview rather than reconstructed from memory afterwards.
- Competencies: four to six, all job-related, lifted from the role definition. Compono's scorecard guide (updated 19-10-2026) quotes two 2026 guides that land on three to six and four to six, and warns that more than six turns scores into noise.
- A 1-5 scale with words attached: Compono's anchors run 1 Strong No Hire, 2 No Hire, 3 Mixed, 4 Hire, 5 Strong Hire, each with a behavioural description.
- A mandatory evidence field: per competency, one or two lines on what the candidate actually said or did. No evidence, no score.
- An overall recommendation: given last, so it follows the evidence rather than driving it.
Use the identical sheet for every candidate for the role. The questions that feed it are in our guide to interview questions for frontline staff; the scorecard is where the answers land.

How do you write anchors two interviewers read the same way?
Write them before the first interview, in plain language, for this job. 'Good communicator' is not an anchor. For a shop floor role, the competency 'handles a difficult customer' might read: 1, blames the customer or freezes; 3, stays polite but escalates everything; 5, describes a real case, names what the customer needed and what they did about it. Each level is observable behaviour, so the interviewer matches what they heard to a line on the page instead of turning a feeling into a number.
Two rules keep the sheet legal as well as useful. Every competency must be a real requirement of the job and applied identically to every candidate. And nothing on it may touch a protected characteristic. Read the questions you cannot ask, and what to ask instead before writing the anchors.
How do you take evidence notes without a GDPR problem?
Evidence notes are personal data. In Croatia and the rest of the EU they fall under GDPR, and a candidate can ask to see them. The UAE's PDPL works the same way. So write nothing you would not read aloud to the person it describes.
That constraint is a quality control, not a burden. 'Seemed nervous' goes. 'Took 20 seconds on the stock-count question, then gave a correct method' stays. Paraphrase what was said, tie it to a competency, and skip appearance, family and background. Set a retention period for interview notes and stick to it. Our GDPR guide for recruiters covers what you may keep and for how long.
How do you run a debrief the loudest voice can't win?
Scores go in before anyone speaks. Metaview's 2026 AI & Hiring Alignment Report surveyed 505 recruiting leaders and hiring managers at companies with 200 or more staff across North America and EMEA, bigger shops than most of our readers, but the finding travels: 58 per cent wished they could work around their counterpart, and when nobody captures the signal in the room, the debrief goes to whoever holds the strongest opinion. 'Didn't click' is a feeling. The fix is to tie every read to a rubric line and a specific moment in the conversation.
Then set a deadline. Across 811,298 scorecards Metaview analysed (29-07-2026), the median arrived about two hours after the interview and 74.7 per cent within 24 hours, but nearly one in twenty took over a week, and the decision waits for the slowest one. Give interviewers 24 hours, then meet. Discuss only the competencies where scores differ by two or more, read the evidence aloud, and let the stronger evidence win, not the stronger voice. If you're weighing how many rounds to run, one scored round beats three vague ones.
The numbers here are illustrative, drawn from Qwiza demo campaigns and pilot targets, and the names are invented. Zelena Kosara, a small grocery chain in Zagreb, interviews two candidates, Ana and Marko, for a shop assistant role. Both completed the screening quiz: Ana scored 78, Marko 64, with a pasted-text flag on one of Marko's answers. Two interviewers, Ivana and Petar, rate five competencies on the same 1-5 sheet, independently, and submit within the hour.
At the debrief only one competency differs by two points: 'handles a difficult customer', where Ivana gave Ana a 4 and Petar a 2. Petar's evidence line reads 'generic answer'; Ivana's reads 'described a refund dispute, named what the customer wanted, offered the exchange herself'. Petar had cut the follow-up short. They settle on 4. For Marko, the pasted-text flag became a question, not a verdict: asked to walk through that answer, he explained it clearly, so the flag was closed with a note.
Final sheets: Ana 21 of 25, Marko 17 of 25, quiz and interview pointing the same way. Ana gets the offer; Marko goes into the talent pool with a note recommending him for the next opening. The whole decision took one 15-minute meeting, and both interviewers could show exactly why.
How does the scorecard connect to the screening quiz?
Carry the same competencies through the whole funnel. If the screening quiz tested stock counting and customer handling, the interview sheet rates stock counting and customer handling, so the two scores sit side by side and a gap becomes a question, not a mystery. In Qwiza, every applicant reaches the interview with an explained quiz score built from employer-graded answers and engagement signals such as time per question. Integrity flags from the quiz, tab switching or pasted text, are prompts to probe in the room, not proof of anything: ask the candidate to walk you through the answer they gave. Nobody is auto-rejected. The scorecard is where a human makes the call, with both scores on the same page.
Score the quiz, then the interview
Qwiza gives every applicant an explained quiz score before the interview and schedules the interview from the same pipeline. Your first campaign is free, and the live demo is at demo.qwiza.ai.
Try the live demoFrequently asked questions
Should the scorecard use a 1-5 or a 1-10 scale?
Use 1-5. The point of the scale is not precision, it is agreement: two interviewers must be able to see the same behaviour and land on the same number. You can write a clear behavioural anchor for five levels; nobody can explain the difference between a 6 and a 7. Compono's 2026 template uses 1 Strong No Hire, 2 No Hire, 3 Mixed, 4 Hire, 5 Strong Hire, each with a description of what that level looks like in the room. Keep the anchors specific to the job and rewrite them for each role rather than reusing one generic set.
Do I need a scorecard if I am the only interviewer?
Yes, for two reasons. First, you are not consistent with yourself: the candidate you saw on a busy Friday afternoon gets a harder hearing than the one you saw on a quiet Tuesday morning, and the sheet corrects for that. Second, the evidence notes protect you. If a rejected candidate asks why, or requests their data under GDPR, a job-related note tied to a competency is a defensible answer; a memory of 'not the right fit' is not. It also lets you compare candidates seen two weeks apart on the same terms.
Should the quiz score or the interview score count for more?
Neither wins by default. Treat them as two readings of the same competencies taken in different conditions: the quiz under time pressure with nobody watching, the interview in conversation with someone who is. When they agree you have a confident signal. When they disagree you have a question, and the interview is where you ask it. A strong quiz and a weak interview might mean nerves; a weak quiz and a strong interview might mean a rushed attempt on a phone. Either way a human reads both, writes down the evidence, and decides. Nothing is auto-rejected on a score.


