.png)
When people first get to grips with OKRs, there is often confusion about how key results differ from key performance indicators, if at all. It’s likely that your management reporting has always contained a list of key performance indicators to capture (with varying degrees of success) the performance of your business. Now that you’re looking at OKRs, how do KPIs fit into this equation? Do you just abandon them? Out with the old, in with the new? And if not, how can you leverage your KPIs to help you craft good key results? Should you even be doing that?
If you google this question (or leverage a language model of your choice), there are lots of blog posts that address this question - not all of them get it right in my view (more on that later). One of the better ones I have come across refers to OKRs as “KPIs with soul” which I thought is quite poetic… But poetry aside, applying theoretical knowledge in the real world can be surprisingly difficult - it’s always messier than the theory suggests - so I wanted to work through an example in this blog post and explain how I would go about journeying from KPI to (O)KR.
KPIs vs Key Results
First things first: what is the difference between KPIs and KRs? Is there one? The answer is, of course, it depends.
Let’s start with defining what we mean by key performance indicator. Again, leveraging the interweb, the most agreeable definition I’ve found is a “set of quantifiable measurements used to gauge a company's overall long-term performance”. It’s effectively a set of data points you look at to give you an indication of how well you’re doing across a range of tangible metrics. This could be obvious things like revenue, profit, market share but it could also be more subtle: employee satisfaction, customer churn, environmental impact.
A key result on the other hand is “a measurable outcome required to achieve [an] objective”. The key differentiator is thus the concrete link to a strategic objective. It is an indicator of success within the specific context set by the objective. The context gives it soul 🙂
The other difference is that a key result almost always implies a target and an associated direction of travel - it’s a measure of change. This isn’t necessarily the case for a key performance indicator, which you may just be monitoring to assess the overall long-term health of your organisation: “Improve client retention by 20%” vs “Keep employee attrition below 10%”.
The way I remember this is as follows: If you are actively looking to change a metric, it’s a candidate for a key result. If you are monitoring a metric for on-going performance without an active change initiative behind it, it is more likely to be a key performance indicator , which can become an excellent “health metric” (i.e. nothing needed for as long as it stays where it is, but if there’s a sudden deterioration you probably want to have a closer look!).
Let me give you a brief example: Let’s say you’re launching a new product and you expect it to have a major impact in the market. You set out to sell at least 100 subscriptions in the first quarter post launch. A likely key result will be the number of subscriptions sold, starting at 0 pre-launch and aiming for 100 by the end of the period. You may have other, related key results for your objective of making a noticeable impact on the market: active users, frequency with which users are logging in, time spent using your product etc.
At the same time, launch excitement notwithstanding, you are conscious that you don’t want the team to burn out. It’s a long road from product launch to mature, established, successful game changer. So you may want to keep an eye on employee engagement, satisfaction, attrition and related performance indicators. For as long as they remain stable, there is nothing at all to do here. Just keep checking in. But should your attrition spike, you better respond - or you may be left with a product in the market and no team to care for it. (At that stage, you may decide that you need an objective in its own right to attend to your culture, your workplace culture and health and now your key performance indicator becomes a viable candidate for a key result).
Lastly, you may have a number of KPIs that you’d like to improve but that aren’t obviously related to a strategic objective. This is where you have to practise saying “not yet”. Don’t fall into the trap of trying to improve everything at the same time! OKRs are supposed to act as a forcing function that funnels your attention to where it is most impactful. The “also important but less so” initiatives must come later.
From KPIs to OKRs: the journey
Let’s dig a little deeper and use a concrete example. Let’s imagine you’re a landlord with lots of properties and lots of tenants. Things occasionally break and while you have been keeping more or less on top of repairs, you want to do better. You consider your tenants as customers of your business and you want to provide a good service to your customers. That is in your interest too, because if tenants find it easier to report issues, they are more likely to do exactly that - report issues when they’re first spotted. This allows you to react more quickly, attend to small repairs as they arise and prevent potentially larger, more expensive repairs further down the line. If tenants have a good experience, they are more likely to stay and they’re more likely to communicate with you which leaves everyone happier as a result.
Knowing this, you set out to craft an appropriate OKR for your business. Maybe something like “We want to provide excellent customer service to our tenants when handling repairs” or “We want to make it as easy as possible for our tenants to live in well-maintained homes”. The exact wording does not matter for the purpose of this example, as long as the intent is clear.
Now the trickier question - how do we know we’re winning? How will we measure our success? Let’s imagine you are already tracking the following metrics related to repairs performance:
Time to complete repairs
Proportion of repairs fixed first time
Disrepair claims
Health & Safety compliance
Regulatory compliance
When you look at your previous status reports, you notice that most of them look strong, although you feel that you could do better on “time to complete repairs” (all numbers are entirely fictitious):
Time to complete repairs - 8 days
Proportion of repairs fixed first time - 99.7%
Disrepair claims - 0
Health & Safety compliance - 100%
Housing Quality Standard compliance - 100%
This is a perfect example of existing KPIs only painting half of the proverbial picture. While we are tracking time to complete the job and the quality of the repair and as a result the state of our homes, there is no mention at all about the human experience.
We don’t know how many times the tenant had to call, text or email to get a response from us (in fact we don’t know if we support multi-channel reporting in the first place!).
We don’t know how many appointments we have missed as part of carrying out the repairs. How many times did the tenant have to wait at home only for nobody to show up? How many times were tenants late for school pick-up because our staff showed up late? (Just think of the number of hours we have collectively wasted waiting for a broadband engineer to turn up!)
We don’t know how courteous our tradesmen were during the repairs process, or whether they left a right mess after leaving.
We don’t know how well we kept the tenant informed along the way - maybe a repair took longer because a part needed to be ordered. Did we let the tenant know? Did we explain the situation? Did they feel at ease and understood what was happening to their home?
I’m sure there are a number of other angles that you can think of. The point of this exercise is less about completeness and more about showing how existing KPIs may make useful key results but may also prove to be insufficient on their own.
Once you have reviewed your existing KPIs, ask yourself: “What do I not know yet that is vital for delivering a successful outcome?”. Theoretically, your key results taken together should necessarily predict a successful outcome. That means if you specify the right set of key results and you hit all the associated targets, you should be guaranteed to have achieved your objective. In reality, it’s rarely that simple - but it’s a useful guiding principle nevertheless.
Putting it all together
Here is an example “repairs” OKR that may result from our exploration above.
Provide as delightful a repairs experience as possible
as measured by
95% of repairs are completed within 5 days
100% of tenant repair reports are captured on the first attempt
100% of tenants were kept informed about the status of their repair
100% of tenants felt treated with respect both for their time, their home and their person
Now we can also be explicit that there is no tolerance for any of the other repair KPIs to deteriorate while we work on improving the service experience. That means we expect there to be no compliance breaches or disrepair claims and our first-time-fix rate should stay above 99.5% (or whatever threshold feels appropriate before active intervention would be required). This turns the KPIs that do not form key results for your objective into health metrics on the wider performance of your repairs service.
If you repeat this process for all of your OKRs, you will end up with a nice dashboard of strategic objectives, associated key results and supporting health metrics - perfect as a foundation for your new status reports (please let these replace your old ones because “OKRs is how we do things, not something we do.”)
Note on “The How”
Note how none of the above goes into the how of achieving these measures. This is deliberately left up to the organisation to innovate on - maybe getting a perfect repair report experience requires more people at the end of the phone, it could mean longer opening hours, or weekend calls, or it may mean additional ways of reporting repairs such as online, WhatsApp (TikTok!?) or at the other end of the spectrum drop-in sessions that allow for face-to-face interactions… there will be a myriad of ways you can deliver a successful outcome. We do not want to dictate what that looks like at the beginning, we just want to agree what good looks like and then leverage the collective wisdom that already exists in our organisation to find the best solutions for our context.
Check out our Services page if you are thinking about implementing OKRs, or are already on your way and would like some help!
Comments