OKRs

How to Write Good OKRs: Method and 12 Examples

Learn how to write OKRs that drive outcomes: a 5-step method, formula templates, 12 bad-to-good examples by function and the wording pitfalls to avoid.

By the RoadmapHero team · Updated on June 26, 2026 · 7 min read

Knowing how to write OKRs is a craft, and like most crafts it's learned faster from examples than from theory. A well-written OKR does three things at once: it states an ambition worth caring about, it defines success with numbers nobody can argue with, and it leaves the team free to choose the path. This article gives you the method, the formulas, and twelve real-world examples — each shown in its common broken form, then rewritten. If you're new to the framework itself, start with our complete OKR guide for product teams.

Before writing: anchor in strategy

The most common OKR-writing mistake happens before a single word is typed: starting from what the team plans to build instead of what the business needs to change. Before the drafting session, gather three inputs:

• The company Objectives for the period — your OKRs must plausibly serve at least one of them.

• Your product strategy — which segment, which differentiation, which pillar this quarter advances (see our product strategy guide).

• Evidence — the 3–5 metrics or customer signals that hurt the most right now, with actual numbers.

A team that walks into the workshop with these three inputs writes OKRs in 90 minutes. A team without them debates wording for two weeks.

The 5-step writing method

1. Name the change. Complete the sentence: "By the end of the quarter, the most important thing that will be different is…". No metrics yet — just the change.

2. Turn it into an Objective. One sentence, qualitative, inspiring, no numbers. Write it in the language your team actually speaks, not corporate dialect.

3. Ask "how would we know?". Brainstorm every observable signal that would prove the change happened. This list is your KR raw material.

4. Pick 2–4 measurable signals. For each, write baseline → target. Kill any signal you can't measure within the first two weeks of the quarter.

5. Stress-test. For each KR ask: "Could we hit this number while the Objective fails?" (gaming test) and "Could the Objective succeed while this number stays flat?" (relevance test). Rewrite until both answers are no.

Formula templates that work

Two templates cover 90% of well-formed OKRs. For Objectives, John Doerr's classic frame:

We will [qualitative ambition] as measured by [this set of Key Results].

For Key Results, the movement formula:

[Verb] [metric] from [baseline] to [target] by [date].

Examples: "Increase week-4 retention from 22% to 30% by March 31", "Reduce median support first-response time from 9h to 2h". Milestone KRs ("Launch in 2 new markets with ≥100 paying customers each") are acceptable when the milestone embeds an outcome — the "≥100 paying customers" part is what makes it a Key Result rather than a task.

12 real examples (bad → good)

Product team

Growth team

Engineering team

Customer success team

Wording pitfalls to avoid

• Output dressed as outcome. "Launch X" with no metric attached. If the verb is ship, launch, build, migrate or release, look for the change the work should cause and measure that instead.

• Vanity metrics. Page views, followers, downloads, registered users. They go up and to the right regardless of value. Prefer activation, retention, conversion, revenue-adjacent measures.

• No baseline. "Reach 50% activation" is meaningless without knowing you're at 47% (trivial) or 12% (heroic). Always write from → to.

• Compound KRs. "Increase activation to 50% and reduce churn to 2%" is two KRs stapled together — impossible to score cleanly. Split them.

• Metrics nobody owns. If no one can pull the number weekly, the KR will silently die by week 3.

• Hedged language. "Explore", "investigate", "work towards" — these words guarantee a 1.0 score and zero change. Commit to a number or don't write the KR.

Run a team OKR-writing workshop

The 90-minute format that works for most product teams (5–9 people, one facilitator):

1. Context (10 min). PM presents company Objectives, strategy reminder, and the evidence pack (metrics, top customer feedback themes).

2. Silent drafting (15 min). Everyone writes 1–2 candidate Objectives on stickies, alone. Silence prevents anchoring on the loudest voice.

3. Cluster and vote (15 min). Group similar Objectives, dot-vote, keep the top 1–3.

4. KR brainstorm (25 min). For each retained Objective, list every measurable signal, then apply the gaming and relevance tests from the method above.

5. Baseline check (15 min). For each candidate KR: do we know the current value? Who pulls it weekly? Kill or fix the unmeasurable ones.

6. Wrap-up (10 min). Assign one owner per KR (an owner tracks the number — the whole team moves it) and book the weekly check-in slot before leaving the room.

Expect the output to be 80% right. The alignment pass with leadership and the first two check-ins will finish the job.

Tracking OKRs in RoadmapHero

Writing good OKRs is half the work; keeping them alive is the other half. In RoadmapHero's Vision & OKR module, each Key Result carries its baseline, target and current value, and roadmap initiatives link directly to the KR they serve — so the connection between goals and delivery survives the quarter (a discipline we detail in aligning your roadmap with OKRs). The AI progression analysis then does the tedious part of the weekly check-in for you: it reads each KR's trajectory and the linked tickets, and produces an actionable summary per Key Result — what moved, what's blocked, where confidence is dropping. Free trial, transparent euro pricing, no commitment.

Related reading: choose outcome metrics with our product KPIs guide, and see how outcomes reshape planning in outcome-based vs feature roadmaps.