Confidence score: measuring conviction on a Key Result

The confidence score is a subjective indicator, updated weekly by the owner of a Key Result, expressing their conviction in hitting the target. It complements objective progress measurement.

Its value-add: capturing weak signals before the numbers tell the story. A team may be tracking apparent good progress on a KR while sensing that upcoming structural blockers will compromise the outcome. The confidence score makes that intuition visible and discussable.

How the confidence score works

The most common scale is 1 to 10. Some teams prefer 1 to 5 for simplicity. A third popular variant uses a color code: green (strong confidence), amber (watch out), red (alert).

What moves a confidence score up or down

The confidence score isn't a vague feeling: it rests on identifiable factors. Surfacing them explicitly in team reviews turns the indicator into a real steering tool.

Confidence score vs objective progress

The two signals can diverge, and that divergence is precisely what makes the confidence score useful.

Example 1: low progress, high confidence. KR "Move Day-7 activation rate from 32% to 55%". Mid-cycle, the metric is stuck at 36%. But the team has just finished a major onboarding flow rebuild that ships next week; internal A/B tests show a 2x effect on activation. Confidence score: 8. The numbers will catch up, it's just a lag effect.

Example 2: good progress, dropping confidence. KR "Sign 30 new scale-up customers". Mid-cycle, the team has 18 signings, at 60% pace. Looks good on paper. But 14 of those 18 came from a pre-existing pipeline, and new lead flow has dried up. The owner drops the confidence score to 5. Without that signal, the team would discover the problem at end of cycle, too late.

• Low progress, high confidence: lag effect. Worth watching, often recoverable.
• Good progress, low confidence: early wins aren't reproducible. Priority signal to address.
• The two aligned: normal situation, no surprise.

How to steer with the confidence score

The confidence score is the primary tool of steering rituals:

• Weekly check-in: every owner updates their score. Any 2+ point drop triggers a discussion.
• Monthly review: analyze the score trajectory over the month, adjust Initiatives.
• Mid-cycle review: at the midpoint, KRs whose score is stuck below 5 should be requalified (lower target, more resources, or deliberate abandon).

Common confidence score mistakes

• Default optimism. Many owners start every KR at 8 or 9. Recalibrate: at mid-cycle, a healthy average sits around 6-7.
• Score that never moves. If the score is the same every week, nobody's actually using it. Force the update.
• Tied to compensation. Toxic. The confidence becomes political and stops being honest.
• No automatic alert threshold. A score dropping below 5 should trigger a review, not wait for end of cycle.

Serendly insight: the confidence score is a conversation trigger

The score itself barely matters. What matters is the conversation it triggers: "why did your score drop this week?" or "what would move you from 6 to 8?".

An organization that steers by confidence score builds a culture of anticipation and transparency. An organization that only watches objective progress reacts only at end of cycle.

Make the confidence score part of your rituals

The confidence score has value only if it's updated regularly and discussed in team. Let's talk about how we structure those rituals.

Request a demo