Examples of Good and Bad OKRs

If you get your OKRs right, they are so much more powerful. Word them wrong, and they just won't have the impact you want.

Why are OKRs so hard to write?

Objectives and Key Results are very simple on the surface. An aspirational, qualitative objective that describes the outcome you want to achieve. And specific, measurable key results that indicate when you've actually achieved it.

So why are they so hard to write?

Firstly because they're subjective. One person's "good" might not match another person's idea of "good", so even if your OKRs are actually well written, it might not seem like it because everyone's still debating them.

Secondly because they're often trying to describe something that is unknown. You may not know exactly what outcome you want, and you may not know exactly what it looks like in advance, and so even well written OKRs could leave people feeling uncertain.

Thirdly because the difference between outcomes and outputs is not immediately obvious to everyone, and takes some practice to get right.

Fourthly because they're written collaboratively, so everyone might have a slightly different idea of what the right outcome is, so you end up going round in circles because you can't agree and it feels difficult. (although this is actually OK and collaboration is encouraged! see below).

Lastly because you're limited by what you can measure, so sometimes you can't set a target on what you really want simply because you couldn't measure it if you tried.

How can well written OKRs help you?

Good OKRs are motivational. They create an energy and excitement towards achieving a tangible outcome that is valuable for customers and the business.

They drive alignment across teams and departments, because everyone was involved in setting them and buying into the goals, and agreeing what is on the table and most importantly what is not!

You'll have clarity on outcomes that you're trying to achieve so it's obvious when you actually achieve them, and you can be certain whether what you've done is working or not, and whether you need to iterate or pivot.

Good OKRs leave flexibility for changing the actual solution, as long as the outcome is still acheived.

When the KRs are right you can see clear progress against them, which is motivational and creates focus to keep making more progress.

Lastly, OKRs are all about learning: even if you don't succeed, you will have tried to achieve something clear and specific and valuable, and if you haven't, you will have at least learned why not.

In other words - it's extremely valuable to practice writing OKRs and improve them!

Sidenote: after you've written them, you can use an OKR template to help when you're presenting your OKRs.

Common mistakes when writing OKRs

• Too many Os: try not to have more than 3, they should capture the biggest and most important and most valuable things you can work on in a quarter.

• Too many KRs: try not to have more than 3 per objective. You don't need to have every single metric, just the key ones that are the strongest measures of success. Perhaps 1 guardrail metric if there's a big risk of something going wrong. You can still watch other metrics of course, but they just don't feature in your KRs.

• Too long or short wording: you need enough explanation that the message and intent of the OKR is properly communicated, but not so much that it turns into paragraphs of text. Save that for a one-pager or other documentation.

• Outputs not outcomes: don't describe things that you'll deliver, describe things that will happen as a result of that delivery.

• Not describing value that's clear to everyone: if people read your objective and think, "so what?", then the value is not clear. It's got to be obvious why achieving this objective is valuable to either customers or the business.

• Not specific enough KRs: if they're not precise, they're not useful.

• Unclear targets on KRs: part of being precise is having a specific target. You could phrase it as "increase metric X from Y to Z", that way you give context about the size of the increase and the current value.

• Prescribed solutions: OKRs shouldn't list actual features (very rarely that's OK), otherwise you're committed to that thing but you might learn along the way that building something else is better.

• Too many deliverables, not enough measures: don't just list things that you'll do or build because what if you do all that work and nothing happens? List metrics that will move when you take action.

• KRs are not obviously linked to the objective, or not understandable: most of the time your OKR needs to be readable by people outside your team, so make sure they understand what you're measuring.

• KRs can't be measured: if you can't measure a KR because you don't have the data or the impact is too lagging to be seen in-quarter, pick something else.

Examples of bad OKRs

Here's an example of an e-commerce shop selling shoes...

• The objective is too short and not inspiring. Re-phrase around a customer problem that we're solving which will persuade people to buy more shoes.
• The first KR needs a specific target.
• The second KR is a deliverable, which is not a terrible thing to have, but it's over-prescribing the exact discount campaign to run.
• The third KR is just a statement, it's not an obvious clear metric that people would understand.

Here's an example for a SaaS product...

• The objective is way too long! Also, it's really a statement of insight, rather than an outcome we're trying to achieve.
• The first KR is a specific target, but lacks context about what the current level of registrations is, and since the objective is vague it is not clear how this KR contributes to the goal.
• The second KR is a specific deliverable, but the outcome of the message is that more people give their information, so measure that.
• THe third KR is unmeasurable (how do we know that customers have given all their information?) and unachievable (customers giving every single piece of information they have is unlikely).

Examples of good OKRs

Let's revisit the example of an e-commerce shop selling shoes...

• The objective is qualitative and describes a clear customer benefit: people can't buy shoes if they can't find what they want
• The first KR is specific, measurable and has a target. If people find what they want, in theory we should sell more.
• The second KR is specific, measurable and has a target. If people find what they want in search, they'll view more product detail pages.
• The third KR is the same. If we stop people from ending up with no results in search, they'll have less reason to leave the site.

And let's revisit the example for a SaaS product...

• The objective is describing concrete business value, and is also an example of how to focus around a solution theme without over-specifying exactly how to go about capturing the user profiles.
• The first KR is specific and measurable with an achievable target that is related to the objective.
• The second KR has a precise number, and is relevant to the objective.
• The third KR is a useful guardrail metric to avoid any unintended side-effects
• The fourth KR describes the ultimate goal: more effective campaigns. The first 3 KRs reflect the success of the data capture itself.

Lastly, here's a new example for a gaming app...

• The objective is qualitative and aspirational, centered on customer value (excitement!) which translates to business value (retention).
• The first KR is reasonable and achievable, it specifies users who are still active on day 7 rather than 100% of all app installs because that would be unrealistic
• The second KR confirms that those users are still getting the excitement of a win because they come back at the normal rate, to avoid cheating the first metric by giving away meaningless wins
• The third KR focuses on ensuring that the early win translates into a longer term retention effect

What do you think? Could you improve these OKRs even further? Give it a try!

That's all folks, you know how to do it now

You've now got a solid understanding of what makes an OKR good or bad, and the examples of good OKRs should help you to craft your own.

Try out a free OKR dashboard for tracking your OKRs and visualising progress through the quarter.