Reviews and Feedback

Why It Matters

Structured reviews reduce bias, improve consistency, and give your team a shared framework for evaluating candidates. Instead of informal opinions, Kit provides blind voting and weighted scoring to surface the best candidates.

How Reviews Work

Reviews happen on stages that require reviewer input — typically Team Review stages, but also Code Assignment and Portfolio Upload stages when reviewers are assigned.

The flow:

1. A candidate reaches a review stage
2. Assigned reviewers receive a notification (email and in-app)
3. Each reviewer evaluates the candidate independently
4. Reviewers submit their recommendation, scores, and comments
5. When enough reviews are in, the candidate advances (or is rejected)

Blind Voting

Kit uses blind voting by default. When you open a review, you can’t see what other reviewers have submitted until you submit your own. This prevents anchoring bias — where early reviews influence later ones.

Once you submit your review, all other completed reviews become visible. This encourages honest, independent evaluation while still allowing the team to discuss and compare notes afterward.

Submitting a Review

When it’s your turn to review a candidate, you’ll see the application in your My Reviews section. Click into it to open the review form.

Recommendation

Choose one recommendation that best reflects your assessment:

Recommendation	Value	Meaning
Strong No	0	Serious concerns — should not advance
No	1	Does not meet the bar for this role
Neutral	2	No strong opinion either way
Yes	3	Meets expectations — should advance
Strong Yes	4	Exceptional candidate — definitely advance

Scoring Criteria

If the stage has scoring criteria configured, you’ll also rate the candidate on each criterion. Each criterion has:

• Name — What you’re evaluating (e.g. “Code Quality”, “Communication”)
• Scale — Maximum score (default: 5)
• Weight — How much this criterion matters relative to others (1-100)

Rate each criterion on its scale. Kit calculates a weighted total based on the weights you’ve configured.

Comments

Add free-form comments to explain your recommendation. Comments are visible to other reviewers after you submit and provide context that raw scores can’t capture.

Scoring Criteria

Scoring criteria turn subjective evaluations into comparable data. Define them on the stage configuration.

How Scoring Works

Each criterion has a name, a weight, and a maximum scale value. When a reviewer scores a candidate:

1. The raw score (e.g. 4 out of 5) is normalized relative to the scale
2. The normalized score is multiplied by the criterion’s weight
3. All weighted scores are summed to produce a weighted total

Example: A stage has two criteria — “Code Quality” (weight: 3, scale: 5) and “Communication” (weight: 1, scale: 5). A reviewer gives scores of 4 and 3 respectively. The weighted total is (4/5 * 3) + (3/5 * 1) = 2.4 + 0.6 = 3.0 out of 4.0 maximum.

This means higher-weighted criteria have more influence on the overall score, reflecting your team’s priorities.

Voting Configuration

Voting settings control how reviews translate into pipeline decisions. Configure these on the stage:

Setting	Default	Description
Threshold	2	Minimum number of positive votes (Yes or Strong Yes) needed to advance
Require All Reviewers	No	Wait for every assigned reviewer to submit before evaluating
Veto Auto-Rejects	No	A single Strong No automatically rejects the candidate

How the Threshold Works

The threshold counts positive recommendations (Yes = 3 points, Strong Yes = 4 points). When the sum of positive recommendations meets or exceeds the threshold, the stage is considered complete.

For example, with a threshold of 2: two “Yes” votes (3 + 3 = 6) would exceed the threshold and advance the candidate. One “Yes” vote alone (3) would also exceed a threshold of 2.

Require All Reviewers

When enabled, Kit waits for every assigned reviewer to submit their review before checking the threshold. This prevents premature advancement when only a subset of the panel has weighed in.

Veto Auto-Rejects

When enabled, a single Strong No recommendation automatically rejects the candidate — regardless of other votes. Use this when any team member should have the power to block a candidate (e.g. for culture or values concerns).

Auto-Advancement

When a voting threshold is configured, Kit can advance candidates automatically. Here’s the sequence:

1. A reviewer submits their review
2. Kit checks if all required reviewers have submitted (if require_all_reviewers is enabled)
3. Kit checks if any review is a Strong No (if veto_auto_rejects is enabled)
4. Kit sums the positive recommendations and compares against the threshold
5. If the threshold is met, the candidate automatically advances to the next stage

This eliminates the manual step of clicking Advance after every review round. Your team submits their reviews and the pipeline keeps moving.

Reviewer Notifications

Kit keeps reviewers informed at every step:

Event	Notification
Candidate reaches your review stage	Email + in-app notification
All reviews submitted on a stage	Email summary to stage lead
Candidate auto-advanced by voting	Slack notification (if connected)

In-app notifications appear in real-time via WebSocket, so you’ll see new review assignments immediately without refreshing the page.

Quick Checklist

• Assign reviewers to every stage that requires evaluation
• Define scoring criteria with appropriate weights for your priorities
• Set voting thresholds that match your team size and standards
• Enable veto auto-rejects for stages where any concern is a dealbreaker
• Use “Require All Reviewers” for important stages to ensure full panel input
• Submit reviews promptly — blind voting means others are waiting on you
• Add comments to explain your recommendation beyond the scores