To evaluate a candidate well, decide in advance which skills and behaviors the job needs, gather evidence on each one through tests, structured interviews, work samples and references, score that evidence on a fixed scale, and compare candidates on the totals rather than on overall impressions. The aim is a decision you can explain: why this person, on this evidence, is the best fit for this role. This guide covers the criteria to use, a weighted candidate evaluation matrix, how to weigh different sources of evidence, and how to compare finalists without bias.
Key takeaways
- Evaluate candidates against criteria agreed before the first interview, not against each other or a gut feeling.
- Combine evidence from several sources: skills tests, structured interviews, work samples and references.
- A weighted evaluation matrix turns that evidence into a comparable score for each finalist.
- Score independently, then discuss, and write down the reason for the final decision.
Set the evaluation criteria first
Most poor hiring decisions start with fuzzy criteria. If the team has not agreed what good looks like, each interviewer quietly uses their own standard, and the loudest opinion wins. Agree the criteria with the hiring manager before the role is advertised.
- Pick four to six criteria that predict success in the role, such as problem solving, customer communication or Excel skills.
- Write one sentence describing what strong performance looks like for each.
- Decide which criteria are must-haves, where a low score rules a candidate out.
- Give each criterion a weight that reflects how much it matters in the job.
Gather evidence from more than one source
| Source | Best for evaluating | Limits |
|---|---|---|
| Resume and application | Relevant experience, required qualifications | Shows presentation more than ability |
| Skills or aptitude test | Ability with core skills, reasoning | Must match the job to be useful |
| Recorded or structured interview | Communication, behavior, judgment | Needs a rubric to be consistent |
| Work sample | Quality on a realistic task | Takes time to set and review |
| References | Past behavior and reliability | Often positive; ask specific questions |
Map each criterion to at least one source, and important criteria to two. If problem solving matters most, check it in a test and in an interview question, so one strong or weak moment does not decide the result.
Use a weighted candidate evaluation matrix
A candidate evaluation matrix lists the criteria down the side and the candidates across the top. Each candidate gets a score of 1 to 5 on each criterion, which is multiplied by the weight and added up.
| Criterion | Weight | Candidate A | Candidate B | Candidate C |
|---|---|---|---|---|
| Customer communication | 30% | 4 | 5 | 3 |
| Problem solving | 25% | 5 | 3 | 4 |
| Product knowledge | 15% | 3 | 4 | 2 |
| Organization | 15% | 4 | 3 | 4 |
| Team collaboration | 15% | 4 | 4 | 5 |
| Weighted total (out of 5) | 100% | 4.10 | 3.90 | 3.55 |
The matrix does not make the decision for you, but it makes the reasoning visible. If the team wants to hire Candidate B despite a lower total, it should be able to say which evidence the matrix is missing.
Write a rating scale everyone uses the same way
| Score | Meaning |
|---|---|
| 5 | Clear, specific evidence of strong performance, beyond what the role needs |
| 4 | Good evidence; meets the role's needs well |
| 3 | Some evidence; meets the basic need with gaps |
| 2 | Weak or vague evidence; significant gaps |
| 1 | No evidence, or evidence of the opposite |
Add role-specific examples to each level where you can. For customer communication, a 5 might be "explains a complex issue simply and checks the customer understood"; a 2 might be "uses jargon and does not check understanding".
How to evaluate interview answers
- 1
Listen for specifics
Strong answers describe a real situation, what the candidate did themselves and what happened. Vague answers stay general.
- 2
Separate content from delivery
A confident speaker is not always the stronger candidate. Score what was said against the criteria.
- 3
Score straight away
Rate each answer during or right after the interview, before memories blur and impressions take over.
- 4
Note the evidence
Write a short line with each score, quoting or summarizing what earned it.
Avoid the common evaluation biases
| Bias | What it looks like | Counter |
|---|---|---|
| Halo effect | One strong trait makes everything else look good | Score each criterion separately |
| First impression | The first few minutes decide the rating | Score answers, not the opening |
| Similarity bias | Rating people like you higher | Use a rubric and a diverse panel |
| Contrast effect | Judging a candidate against the previous one | Score against the criteria, not each other |
| Recency | Remembering the last candidate best | Score immediately after each interview |
| Confirmation bias | Looking for evidence to support an early view | Ask the same questions of everyone |
Compare finalists and make the decision
- 1
Each evaluator scores independently
Collect everyone's scores before the decision meeting.
- 2
Check must-haves first
Remove anyone who scored below the bar on a must-have criterion.
- 3
Compare weighted totals
Look at the totals and at where scores differ most between evaluators.
- 4
Discuss the differences, not the totals
Where evaluators disagree, share the evidence behind each score.
- 5
Record the decision and the reason
Write down who was chosen and why, in terms of the criteria.
For the first round, a screening tool such as springhire can provide part of the evidence: a skills test score and recorded answers scored against your rubric, ranked across all applicants, so the matrix starts from consistent data.
Green flags and red flags
| Green flags | Red flags |
|---|---|
| Specific examples with their own role clear | Only talks about what "we" did |
| Asks thoughtful questions about the work | No questions, or only about perks |
| Admits mistakes and what they learned | Blames others for every failure |
| Consistent across test, interview and references | Big gaps between what they say and what tests show |
| Prepared: understands the role and company | Has not read the job description |
