How to compare interview candidates without letting the best storyteller win

The way to compare interview candidates is against the job requirements first, and against each other second. Skip that order and the best storyteller in the room becomes the standard, even when the story says little about the work. A good comparison shows, for each requirement, the evidence, how it was assessed and what is still unknown.

The decision still takes judgment. The worksheet makes that judgment visible enough to question.

1. Finish your notes alone before the meeting

Each assessor should complete their own evidence notes and ratings before reading anyone else's conclusions. CIPD recommends this order because early discussion can sway other interviewers.

Use the same role requirements and rating anchors. Paraphrase the relevant answer or behavior and say which anchor it supports. Hold off on "strong hire" until the requirements underneath it have been assessed.

OPM's structured-interview guide describes the same thing: independent ratings, then discussion of the evidence behind differences. It is a useful reference, and you don't need to copy a federal hiring process.

2. Build a comparison that shows missing evidence

Here is a made-up example comparing two candidates for an implementation role. The scale is 1 for evidence below the task requirement, 2 for evidence meeting it and 3 for stronger evidence against predefined behavioral anchors. N/A means the task was not adequately assessed. These values are not validated predictors of job performance.

Don't add up the ratings. Candidate A totals five from two completed ratings and Candidate B totals six from three, so the sums mostly count how many criteria got assessed. A's missing item is not a zero either. (The clock ran out; Candidate A did not.)

3. Discuss one requirement at a time

Ask each assessor for their evidence on a requirement before their recommendation, so one overall impression can't fill every cell.

Useful prompts:

• Which action in the answer supports that rating?
• Did we apply the same anchor to both candidates?
• Are we crediting the candidate's own work or the outcome of a large team?
• Did a standard probe resolve the ambiguity, or did we leave it unexplored?
• Is the disagreement about the answer, the meaning of the requirement, or evidence we failed to gather?

Each answer to that last question has a different fix: correct the notes, clarify the scorecard or run a comparable follow-up. A vote fixes none of them.

4. Fill gaps before you call a winner

In the example, assess Candidate A's handoff ability using the established process before concluding that B is stronger overall. If the follow-up adds a new task or criterion, consider whether other candidates need the same chance.

If evidence stays missing and you must decide anyway, name the unresolved item and who accepted the uncertainty. Never rewrite N/A as a confident rating to make the paperwork look complete.

Keep essential requirements visible. A high score elsewhere should not quietly offset a requirement the team set as necessary. And don't invent a new essential criterion during the debrief because it favors one person.

5. Write the decision in work terms

Use a short record with four fields:

Requirement: The task or capability being decided.

Evidence: The relevant assessment result and its source.

Uncertainty: Missing information, differing contexts or a limit of the method.

Disposition: Proceed, gather a defined additional item, or do not proceed under the established criterion.

Keep it to information relevant to the role. A familiar conversational style, a shared hobby or an interviewer's unexplained feeling is not a criterion.

A structured comparison won't remove bias or guarantee the right hire. It lets another reviewer see which work mattered, what evidence you gathered and why you decided what you did, so start your next debrief with the first requirement on the scorecard.