Most small business owners believe they need instinct or experience to evaluate candidates fairly. They need something simpler: written criteria applied the same way to every person. A scorecard doesn’t replace judgment. It makes judgment consistent.
Without a scorecard, interviews produce impressions. One candidate was articulate and confident. Another seemed quieter but answered more specifically. These impressions don’t compare with each other, and they don’t hold up six months after the hire.
A scorecard applied consistently turns impressions into evidence. It forces each interview to collect the same signals. It makes the decision auditable: by you, and by yourself in six months when you need to know what you were thinking.
The Candidate Comparison Scorecard has five fields. It is the natural next step after How to Write a Job Description That Filters In, Not Just Out has brought the right candidates to the table. It takes fifteen minutes to build before the first interview and changes what every interview produces.
Why Hiring Decisions Fall Apart at the End
The obvious failure: you interview three people, like two of them, and can’t decide between them. Both seemed right in the room. Both said the things you needed to hear. You make the offer based on who you liked better, which is mostly a function of who interviewed last.
The less visible cost is what this does to your ability to learn. If the decision was based on impression, you have no record of what you evaluated. When the hire doesn’t work out, you can’t identify which signal you missed. The next hire starts from the same process and the same risk.
The deepest problem is that unscored interviews produce inconsistent evaluations within the same search. If two people interview each candidate, they walk away having tested different things. Nobody can combine their assessments because they weren’t measuring the same criteria.
A structured scorecard eliminates the comparison problem. It does not eliminate judgment; it gives your judgment something to work from.
The Candidate Comparison Scorecard
The scorecard has five fields. Each field is filled out for every candidate, in the same order, during or immediately after the interview.
Field 1 records what the candidate showed on each key criterion
Take the role’s two or three non-negotiable traits from the success definition you built in How to Define What Success Looks Like Before You Post a Single Job. For each one, write a one-sentence note on what the candidate showed: a specific example, a direct answer, or the absence of either.
Do not score yet. Just record. “Described building a process from scratch at her previous role, gave the specific steps” is evidence. “Seemed like she’d be good at it” is an impression and doesn’t belong here.
Field 2 rates each criterion on a simple scale
After recording the evidence, rate each criterion on a 1–3 scale. One means the criterion was not demonstrated. Two means it was demonstrated but with gaps. Three means it was demonstrated clearly and specifically.
A three-point scale is sufficient. A ten-point scale sounds precise and produces false precision. Most interviewers cannot meaningfully distinguish a 6 from a 7. Everyone can distinguish among “not demonstrated,” “partial,” and “clear.”
Before: After interviews, you debate which candidate “felt” stronger. After: Each candidate has a criterion-by-criterion evidence record and a 1–3 rating for each. The comparison is on paper.
Field 3 notes the specific concern for each candidate
Every candidate has at least one thing that could go wrong. Field 3 names it explicitly before the decision is made.
For the finalist you’re leaning toward, what is the scenario in which this hire fails? Name it specifically. “She’s done this in a large company but never for a six-person team” is a specific concern. “Just a feeling” is not.
Naming the concern before the decision is made is not pessimism. It is the discipline that keeps you from hiring the impression and discovering the gap six weeks later.
Field 4 captures the evidence-supported recommendation
After recording the criteria ratings and the concern, write one sentence: what you recommend and what evidence supports it.
“Recommend Candidate B: demonstrated system-building in two examples, acknowledged the ambiguity concern directly, rated 3 on both non-negotiable criteria” is a recommendation. “Candidate B just seemed right” is not.
The recommendation sentence is what you will want to read six months from now if the hire works. It is also what you will want to read if it doesn’t.
Field 5 records the final decision and the rationale
After comparing all candidates, record who was hired and the one or two reasons. Not a paragraph. One sentence.
This is the last entry in the scorecard for this search. It becomes the first entry for the next one.
How the Conductor Surfaces Your Candidate Evaluation
You’re down to two candidates and it’s time to decide. Both scored well on the criteria. Both have concerns. The Conductor is the right tool for this moment.
The Living Library is Kiluma’s knowledge layer: accumulated role notes, scorecard records, and past hiring data organized in one place. The Conductor, Kiluma’s context-aware AI, reads from that record to help you think through a specific decision rather than working from a generic hiring framework.
Ask it: “Based on the scorecard records and what we know about this role, which candidate is the stronger fit?” It reads the filled scorecard fields, the concerns recorded for each candidate, and the role’s success criteria. It returns a structured comparison: where each candidate scored, where the gaps are, and which concern is more manageable given the role’s actual requirements.
That comparison is not a recommendation to hand the decision to the Conductor. The decision is yours. The Conductor makes the decision from an evidence base rather than from the impression left by whoever interviewed most recently.
Build the Scorecard Before the First Interview
Before your next hire, write the scorecard. Take the role’s non-negotiable traits from the success definition and create one field for each. Add a concern field and a recommendation field.
That is the whole document. It can be a simple table or a document with five sections. The format matters less than filling it out consistently for every candidate.
Set a rule: no hiring decision without a complete scorecard for every candidate in the final round. The rule takes the gut-feel collapse off the table before it can happen.
When the Decision Has Something to Stand On
The owner who makes a hiring decision from a completed scorecard knows what they decided and why. Six months later, whether the hire worked or not, they have a record that informs the next one. Try Kiluma free for 14 days at kiluma.ai.