Goal-Based OKR for Product
Goal-based OKRs for product set Objectives as customer or business outcomes (not features) and Key Results as measurable metrics that prove the outcome moved. Marty Cagan distinguishes 'goal-based OKRs' from 'feature-list OKRs': goal-based OKRs leave HOW to the team. Itamar Gilad's GIST framework places OKRs above ideas and steps — OKRs constrain the search space, ideas explore it, steps execute. The discipline: a Key Result is a NUMBER WITH A BASELINE AND A TARGET. 'Improve onboarding' is not a Key Result. 'Increase activation rate from 32% to 50% by end of Q3' is. Most product OKR failure modes trace back to KRs that aren't actually measurable, or Os that are disguised feature lists.
The Trap
The trap is the feature-disguised-as-KR. 'Launch v2 of the dashboard' is a feature, not a Key Result, no matter how it's labeled. The team will ship v2, declare the KR done, and have moved zero customer outcomes. The other trap: too many OKRs. Cagan's rule of thumb is 1 Objective with 2–4 Key Results per team per quarter. Teams with 4 Objectives and 12 KRs are not focused; they're hedging. Hedging in OKRs is the same as having no OKRs — you'll hit some, miss others, and call the quarter 'mixed.'
What to Do
One Objective per team per quarter. Each Key Result has format: 'metric, baseline, target, by date.' Reject any KR that doesn't fit this format. Hold a mid-quarter checkpoint at week 6 — review whether KRs are tracking and adjust hypotheses (not targets). At quarter end, score 0–1 (Google scale): 0.7 is a healthy target hit. Anything above 0.9 means you sandbagged; below 0.4 means you misunderstood the problem.
Formula
In Practice
Marty Cagan repeatedly cites Google's original product OKR practice as the canonical example: small teams owned one Objective per quarter (e.g., 'Improve search result relevance for ambiguous queries'), with 3 Key Results that were measurable (e.g., 'Increase CTR on top result from X% to Y% for queries flagged as ambiguous'). The HOW was left to the team. Scoring 0.7 was the explicit target. This format produced the discipline of choice and prevented the 'we shipped 5 features and called it a success' anti-pattern. Source: Marty Cagan, EMPOWERED, and John Doerr, Measure What Matters.
Pro Tips
- 01
Cagan: 'If your team has more than 4 Key Results, they have no Key Results. They have a to-do list with aspirations attached.'
- 02
Forbid features in Objectives entirely. The Objective is the customer outcome; the team gets to discover the features. If executives are writing features into Objectives, OKRs are being used as project management.
- 03
Measure 'KR malformation rate' — what percent of submitted KRs lack a baseline + target + date. Above 30% means the org doesn't actually understand the format and the next quarter's OKRs will be theater.
Myth vs Reality
Myth
“OKRs replace the roadmap”
Reality
OKRs and roadmaps answer different questions. OKRs say what outcomes the team commits to. Roadmaps say what the team is exploring or building. A team can have an OKR to improve activation and a roadmap that lists the bets being explored to do so. They coexist.
Myth
“Hitting all your OKRs is good”
Reality
Hitting all your OKRs at 1.0 means you sandbagged. Google's design intent was for ambitious targets where 0.7 is success. A team that consistently hits 1.0 should raise targets next quarter. A team that consistently scores 0.3 should narrow Objectives and improve discovery.
Try it
Run the numbers.
Pressure-test the concept against your own knowledge — answer the challenge or try the live scenario.
Knowledge Check
A team submits these Q3 Key Results: 'Launch redesigned billing flow,' 'Ship 3 new integrations,' 'Improve customer satisfaction.' Which of these is a properly formed Key Result?
Industry benchmarks
Is your number good?
Calibrate against real-world tiers. Use these ranges as targets — not absolutes.
Healthy Quarterly OKR Score (Google Scale)
Product teams using Google-style OKRsSandbagged Target
> 0.9
Strong Hit
0.7 – 0.9
Acceptable
0.5 – 0.7
Misunderstood Problem
< 0.4
Source: John Doerr, Measure What Matters; Marty Cagan, EMPOWERED
Real-world cases
Companies that lived this.
Verified narratives with the numbers that prove (or break) the concept.
Google (early 2000s)
2002–2010
Google's product OKR practice, introduced by John Doerr based on Andy Grove's iMRI from Intel, became the canonical model. Each team owned one Objective per quarter with 2–4 Key Results, scored 0–1 at quarter end. The 0.7 target was explicit cultural doctrine — anything consistently at 1.0 meant the team wasn't aiming high enough. The format produced the focus that allowed Search, Ads, and Maps teams to operate semi-autonomously while still driving toward company outcomes.
Objectives per Team per Quarter
1
Key Results per Objective
2–4
Target Score
0.7 (not 1.0)
OKRs work when the format is enforced ruthlessly — one objective, measurable KRs, 0.7 as the target. Companies that water down any of these get the OKR overhead without the focus benefit.
Hypothetical: A Series C SaaS
2024
A Series C SaaS adopted OKRs but allowed each team to set 3 Objectives with 5 Key Results each — 15 KRs per team per quarter. Of the 60 KRs across 4 product teams in Q1, 38 were features ('ship X'), 14 were vague ('improve Y'), and 8 were properly formed. At quarter end, the team scored ~70% completion on the feature KRs, so the OKRs were declared 'mostly successful.' The product's actual outcome metrics didn't move. The team concluded OKRs were a useful planning tool. They were actually a project tracker with quarterly cosmetics.
Objectives per Team
3 (vs. recommended 1)
KRs per Team
15 (vs. recommended ≤ 4)
% of KRs Properly Formed
13%
Outcome Metrics Moved
Negligible
Volume of OKRs is inversely correlated with focus. 15 KRs per team is a project list with OKR labels. The format is meaningless without enforcement of the constraints.
Decision scenario
Cleaning Up the OKR Process
You're a new VP of Product. Last quarter, the org submitted 60 Key Results across 4 product teams. Less than 20% were properly formed. Most were features. Q2 planning starts in 3 weeks. The CFO is skeptical of 'wasting time on OKR ceremony.'
Q1 KRs Submitted
60
Properly Formed KRs
~13%
Avg Score Reported
0.72 (mostly features shipped)
Outcome Metrics Moved
Negligible
Decision 1
Your VP of Engineering wants to abandon OKRs entirely. Your CEO is sympathetic to OKRs but doesn't want to micromanage. The Head of Sales is indifferent. You have 3 weeks before Q2 planning.
Abandon OKRs and replace with a quarterly roadmap commitment — simpler and matches what teams are already doingReveal
Keep OKRs but enforce hard constraints: 1 Objective per team, max 3 KRs, must have baseline + target + date. Reject submissions that fail the format. Hold a 'malformed KR' clinic in the planning week.✓ OptimalReveal
Related concepts
Keep connecting.
The concepts that orbit this one — each one sharpens the others.
Beyond the concept
Turn Goal-Based OKR for Product into a live operating decision.
Use this concept as the framing layer, then move into a diagnostic if it maps directly to a current bottleneck.
Typical response time: 24h · No retainer required
Turn Goal-Based OKR for Product into a live operating decision.
Use Goal-Based OKR for Product as the framing layer, then move into diagnostics or advisory if this maps directly to a current business bottleneck.