How to Compare Finalists Fairly Before You Decide

To compare finalists fairly, line up the same evidence for each person against the criteria you agreed at the start, have each interviewer score independently, discuss where scores differ, check your own bias, weigh the criteria that matter most for the role, and write down why you chose one person. The decision is still a human judgement. The method makes sure it is based on the same information about everyone.

Why finalist decisions go wrong

By the final round, the candidates have been seen at different times, by different people, with different questions. The last person interviewed is freshest in memory. The one who is most like the interviewer feels easiest. A confident talker outweighs a quieter person with stronger evidence. Without a method, the loudest voice in the debrief wins. A structured comparison does not remove these pulls, but it makes them visible. A quick check on your own process: were all finalists given the same stages, the same core questions and a similar amount of time? If not, note the differences now and take them into account, rather than treating the uneven results as comparable.

Go back to the agreed criteria and line up the same evidence

Go back to the agreed criteria. Start from the intake brief and the scorecard: the must-haves, the outcomes the role needs, and the weight each matters. If criteria were never agreed, agree them now, before looking at who wins. Changing the criteria after you know the candidates is how bias gets in. Put the same evidence side by side. Make a simple grid with candidates in columns and criteria in rows. For each cell, fill in what you actually saw: a project they described, a task result, a reference comment, the scorecard rating and a note. Where one candidate has no evidence on a criterion, mark it as not tested rather than guessing. If one candidate was asked different questions, mark that clearly so that the gap is not read as weakness. When the discussion starts, ask the least senior person to speak first, and the hiring manager to speak last. This stops the most senior view anchoring everyone else. Keep the conversation tied to the evidence on the grid and ask people to say which criterion they are speaking about. Include the reference and background check results in the same grid where they were run, and mark the ones still outstanding. A finalist with a pending check is not comparable to one whose checks are done, and the team should know that before the vote. Use the same wording for the same rating, so that a three means the same thing in every row.

Collect independent scores first

Before the debrief, ask each interviewer to submit their ratings and notes without seeing others. Then compare. Differences are the useful part of the discussion: someone rated one finalist low on a criterion where two others rated high. Ask each person to say what they heard. Sometimes one person saw something the others missed, and sometimes one person is applying a different standard. If an interviewer could not attend or only saw part of the process, mark their input as partial rather than letting a missing score count as a low one. Fairness depends on comparing like with like.

Weigh the criteria, not the headcount

Not every criterion matters equally. A finalist with the best average may be weaker on the one thing the role cannot do without. Decide in advance which criteria are essential, and treat a serious gap there differently from a gap on a nice-to-have. Be careful with simple totals. They can hide a candidate with one serious weakness or undervalue one with a rare strength.

Look at risk, and test your gut

Look for what is risky, and what can be learned. Ask for each finalist: what is the biggest risk if we hire them, and can we reduce it with training, support or a better structure? A skill gap that can be trained is different from a values or reliability gap. Check references and background checks, where they apply, so that you are comparing verified information. Test your gut. If you feel strongly about a person, say why. If the reason is job-related evidence, add it to the grid. If it is "good energy" or "like me", question it. Ask whether you would rate the other finalist differently if their background, accent, age or gender were different. Make sure nothing unrelated to the job, in many places protected by law, is influencing the decision.

Decide and record why

Choose, and write two or three lines on why, in terms of the criteria. Record why the other finalists were not chosen, so you can give honest feedback and consider them for future roles. Name a second choice in case the first declines. Laws on documenting decisions and on candidates' rights to see data differ by country, so check local rules. After the decision, tell the finalists who were not chosen promptly and kindly, and offer feedback based on the grid where your policy allows. Finalists are your best pool for the next search, so a respectful close is also practical. Keep the grid with the record of the hire for a reasonable period, within the retention rules where you hire. When someone later asks why a person was chosen, or why another was not, the grid and your two or three lines are the answer, and they protect everyone involved.

Common mistakes to avoid

Several mistakes recur at this stage. Choosing the person you would enjoy having a drink with confuses likeability with capability. Giving extra weight to the most recent interview favours whoever came last. Letting one dramatic story outweigh a pattern of consistent evidence is another trap. Changing the criteria halfway to fit a favourite is hard to spot in yourself, so compare against the written list. Asking for further rounds because you are unsure, without saying what you hope to learn, wastes the candidates' time. Finally, avoid deciding by compromise, picking the candidate nobody objects to. Sometimes the person with one clear strength and a manageable gap is a better hire than someone average on everything.

In Rezumz

In Rezumz you can rank and compare candidates with the supporting evidence for each requirement, see every interviewer's scores and recommendation in the Debrief and decision view, and see where interviewers disagree. You can then record advance, hold, reject or hire with a line on why. Rezumz does not pick the winner, and the score is a guide for your discussion. People decide.