OKRs are just dreams without grounded key results

You may be struggling with objectives and key results (OKRs) when getting started. The main reason is that you need to reframe the way that leaders provide direction to those on their teams. It is going from very specific work output to the customer outcomes they want.

After setting great objectives you need to measure progress towards those objectives with key results (KRs). They must be quantitative and time-bound. They are ideally leading indicators of progress and not following business impacting metrics even though the business impact will be top of mind for a team’s leadership.

In this article, I will take you through the stages that teams go through to create great KRs for objectives. Before we write the KRs we should have great objectives first rather than creating an objective that fits KRs you already have. If you don’t already have them check out this previous article on how to write great ones and the stages of evolution your team will go through.

The evolution I’ve usually seen in teams first getting started with KRs goes like this:

  1. As initiatives
  2. As following, business-impacting metrics
  3. As an outcome

As initiatives

Just like in the previous article about writing great objectives, you and your team will start writing OKRs primarily as initiatives for both the objectives and then the KRs. This is because when you can’t put them into the objective they should surely go into the KR, right? Wrong!

Let’s say that a video streaming service’s recommendation team had the following objective:

Members are excited to watch content they have been talking about

The team might take this inspirational, end-state outcome and try to just put another initiative as a KR:

Recommendation engine shipped to production

This is a common trap to fall into but you shouldn’t. The point of a KR is progress towards the ambitious objective, not to track the output of the team. I’m looking to find the right balance between output tracking (e.g. how many features the team shipped or bugs the team fixed) and moving closer to leading indicators that the team is doing the right thing.

As following, business-impacting metrics

The step after initiatives will be to go too far past an outcome and go straight for the business impact metrics. You skipped right over the outcomes to business impact if you look at the progression that Josh Seiden has developed in his “Outcomes Over Outputs:”

How to measure progress with outcomes table
How to measure progress with outcomes by Josh Seiden

For example, the video recommendations team may try to use the following KR:

Member retention is increased by 5%

While this is a great goal to have it will be hard to directly tie the work that the team does to this and it may take longer than a quarter to show that result. I’ve found that when you try to use business impacting metrics the distance between their ability to make change is too far away.

This post by Curtis Stanier has a great point to make about how to observe changes you make vs. the end goal, which is usually a business-impacting change.

Measuring Impact — picking the right metrics in product
Measuring Impact — Picking the right metrics in Product by Curtis Stanier

You may want to directly impact the share price but you have much more capability to change things like CTR or CVR. Ideally, you push towards something that is customer-impacting between these near-term metrics and the long-term metrics.

As an outcome

Now you are trying to dial in the KR to be between output and impact, you will generally find some type of metric that is an outcome that progresses towards the objective.

In the video recommendation’s team example, you may see them adopt the following KR:

Increase starters by 10% for members that have the modified “recommended for you” panel

Ideally, this is a set of KRs that you are tracking together. In all of the KRs that I’ve helped people build, I generally start from a set of these three adopted for the objective:

  1. Progress KR that leads to an outcome
  2. Counter-KR to avoid adverse problems if the first metric is a runaway success
  3. Experience KR to make sure that people enjoy the experience

In our example, the team may expand the single KR for the video recommendations group to be the following:

Increase starters by 10% for members that have the modified “recommended for you” panel

Keep completers ratio the same as non-modified “recommended for you” panel

No increase in complaints to customer support about modified “recommendation for you” panel

The first KR is looking for a direct change that leads to more watching with the second KR balancing against bad effects (in this case starting more but completing less). Finally, the last metric is to make sure there aren’t any detected issues with customer experience.

Who sets KRs?

I’ve found that OKRs are an interface point between leaders of an organization and the teams that are doing the actual work. For this reason, I’ve recommended a pattern where the objective is co-authored between the leaders and team first with signoff by the leaders it is the right objective to go after.

It is then up to the team to draft the KRs. This is because the team will have the best understanding of what progress will look like towards an objective.

If the leaders disagree this is an important discrepancy to discuss. It may show that the team will not be able to make progress towards a leader’s objective.

Next article: writing great initiatives

Just kidding. Honestly, this is the least important part of OKRs. The initiatives are a possible set at the beginning of the OKR cycle and could be totally different by the end.

Between you and me, the only reason why I include initiatives in the OKR framework is to make people stuck on the output side of things feel more comfortable. Once you get used to setting the full OKR it ends up being way less important.

It is almost that most traditional roadmaps are an output-driven list of initiatives… Hmm, interesting!

Whether you make progress or don’t

As you have seen in this article, writing great KRs isn’t just copying and pasting the KPIs you already track in a scorecard. You may pull metrics from there but I’d argue that if you are doing something truly aspirational with your objectives the likelihood of you already tracking a metric of progress towards it is very low.

After you set these OKRs you may find that you aren’t able to make progress. This is an important signal that you shouldn’t ignore. OKRs are used as a feedback mechanism to tell whether you are achieving what is considered to be most important by the strategy of the organization.

If you are not able to make progress towards an objective through the key results it could be that:

  1. The team doesn’t have as much agency over the progress measure as you thought
  2. Leadership has set an objective that isn’t currently possible (can’t make progress) or valuable (won’t make progress)
  3. Leadership and team haven’t found the right matching objective and key results

The key of key results, so to speak, is to make the best bet on what would progress. If you find you can’t make that progress you should adjust and try again.

I hope that you can use these lessons to create great objectives and key results for your team.