Practice6 minOKR Management

Good vs Bad Key Result: 10 Real Examples That Show the Difference

The difference between an OKR that works and an OKR that turns into a task list isn't methodology, isn't tooling, isn't team discipline.

It's the Key Results.

A good KR drives action and measures real progress.

A bad KR confuses the team, demotivates when the quarter ends, and turns the whole OKR into corporate theater.

There's only one rule: a Key Result measures consequence, not activity.

The core rule that separates good from bad

Activity is the HOW. Outcome is the WHAT. OKR cares about the WHAT. Activities live in the action plan, in weekly tasks, in the day-to-day. Not in the OKR.

When the Key Result becomes a task list, the OKR loses its superpower: forcing the company to look at what actually changed.

A team that completes every planned task and celebrates at the end of the quarter without checking the numbers isn't using OKR. It's using a to-do list with a quarterly calendar.

10 examples side by side

#BAD (activity)GOOD (consequence)
1Launch the new websiteRaise website conversion from 1.5% to 4%
2Post 30 times on InstagramGenerate 500 qualified leads via social
3Hire 2 salespeopleGrow qualified pipeline from $50k to $200k
4Create onboarding programReduce 30-day churn from 40% to 10%
5Implement CRMReduce average follow-up time from 72h to 24h
6Run satisfaction surveyRaise NPS from 25 to 50
7Redesign sales proposalRaise close rate from 15% to 30%
8Train the support teamReduce complaints from 20/month to 5/month
9Create company blogReach 10,000 monthly organic visits
10Renegotiate supplier contractsReduce delivery cost from 45% to 30% of revenue

The pattern that emerges from the list

Notice it?

The bad side describes what you're going to do.

The good side describes what's going to change.

Running a survey doesn't guarantee NPS rises. Implementing CRM doesn't guarantee follow-up improves. Launching a website doesn't guarantee conversion grows. Activity can happen in full and the result can fail to show up.

That's the difference that protects an OKR from becoming decorative. When the KR measures consequence, the team stays forced to ask "did the number move?" throughout the cycle. When the KR measures activity, checking off tasks is enough to feel like the goal was met.

To understand how to build the full OKR from well-written KRs, read Bad vs Good OKR: 8 examples side by side.

The Key Result test in one question

i
The question that decides everything

*If I complete every planned activity and that number doesn't change, is the OKR met?*

If the answer is no: good Key Result. The number rules, no matter how many tasks you finished.

If the answer is yes: bad KR. It's measuring activity disguised as outcome. Rewrite it, replacing the deliverable with the number that actually shows business progress.

Three specific errors that kill a KR

Error 1: Action verb at the start

A KR that starts with "Launch," "Implement," "Create," "Deliver," "Run" is almost always a disguised task. A good KR starts with a quantifiable noun: "Conversion," "Rate," "Time," "Volume," "Margin," "Number."

Error 2: KR mixing different metrics

"Raise conversion and reduce churn" in the same KR creates ambiguity. If conversion goes up but churn doesn't move, was the KR met? Each metric deserves its own line.

Error 3: KR with no baseline

"Reach $50k MRR" without saying where you're starting from doesn't measure real effort. A well-written KR has starting number (where you are today), target number (where you want to land), and timeline (by when).

!
The minimum KR structure

[Indicator] + [starting value] → [target value] by [deadline].

Example: Website conversion: from 1.5% to 4% by 6/30.

To avoid the most common error in building full OKRs, read When OKR Fails: 5 common traps.

The provocation I want to leave you with

Pull this quarter's Key Results, if your company has them.

For each one:

  • Does it start with an action verb or a metric?
  • Does it have starting value, target value, and deadline?
  • If I complete all the tasks, can the KR still fail?

If at least one answer comes back wrong, you have well-formatted tasks, not Key Results.

Rewriting bad Key Results mid-quarter takes 30 minutes and saves the entire cycle. Living with bad KRs costs 90 days of invisible work and team frustration at the end.

Structure: objective + 2 to 3 measurable KRs + weekly check-in. Right in your browser, no account needed.

Create Key Results With Numbers

OKR doesn't require complex methodology. It requires Key Results written as consequence, not as activity. That single rule, applied with discipline, is the difference between methodology that works and a fancy task list with quarterly branding.

Keep reading about OKR

Frequently asked questions

What's the exact difference between a Key Result and a task?

A task measures activity: what you did. A Key Result measures consequence: what changed in the world because of what you did. Launching a new website is a task. Raising website conversion from 1.5% to 4% is a Key Result. The task contributes to the result, but it isn't the result.

How do I know if my Key Result is well written?

Run this test: if I complete every planned activity and the number doesn't move, is the KR met? If the answer is no, the KR is well written. If the answer is yes, the KR is measuring activity disguised as outcome and needs to be rewritten.

Does every Key Result have to be a number?

Yes. A KR without a number is a wish in good formatting. The number can be absolute (reach $200k pipeline), percentage (raise conversion from 1.5% to 4%), or interval (keep NPS above 50). It always has to be quantifiable and independently verifiable.

How many Key Results per objective is ideal?

Two to three. One alone makes the objective fragile to a single indicator. Four or more dilutes focus. The practical rule: each Key Result covers a different dimension of the same objective (volume, efficiency, quality).

Can a Key Result combine multiple metrics?

No. Mixing metrics in one KR creates ambiguity: one goes up, another falls, was the KR met? Each metric deserves its own line. If two metrics are connected but measure different things, they're two separate KRs, not one combined.