Calibrate stretch with a target score

Aim where hitting ~70% is a strong result, so comfort means you aimed too low.

Why it works

Deliberately setting Key Results so full completion is unlikely keeps the goal in the difficult-but-committed zone that drives the highest effort. Scoring against a target below 100% removes the perverse incentive to lowball, because partial achievement on a hard goal is read as success rather than failure.

How to do it

1. Set stretch Key Results where roughly 70% achievement is the expected good outcome.
2. Decide in advance that scoring 100% means the goal was too easy.
3. Keep a few "committed" (must-hit) goals separate from stretch ones to avoid all-or-nothing risk.

Evidence

Consistent with goal-setting findings that harder goals raise performance up to the limit of ability — but the specific "0.7 is good" scoring convention is a practitioner heuristic from companies using OKRs, not a measured threshold. (anecdotal)

Stretch scoring backfires where reliability matters; some goals (safety, compliance) must be committed 100% goals, not stretch goals.

Sources

• Goal-setting theory (difficulty drives performance) underlies the practice; the 70% target itself is organizational convention

Common mistake

Treating every OKR as a stretch goal, including the ones that genuinely must hit 100%, which normalizes missing things that cannot be missed.