Lifting focus by employing OKRs

Using a goal driven framework in a result driven organization

Authors: Marco Gerritsen & Tim Valkenburg

Introduction

We introduced objectives and key results (OKRs) nearing Christmas in 2018 wanting a fresh launch in 2019. Having read Andy Grove’s work on High Output Management (Random House Usa Inc, 1995) and further desk research, there was an opportunity of bringing direction by employing objectives in the midst of a variety of methods such as Agile Scrum, Objectives Goals Strategies and Measures (OGSM) and Prince2 project management. Our problem? In the years prior to OKRs we merely described what we did. Now, having achieved several key results quarter after quarter, we aim to do what we set out to do as formulated in our OKRs. This paper aims to give a brief description on our experiences, and we provide three steps that might get you started.

The stage

Dutch Railway, principal railway operator on the Core Network in the Netherlands, has broken quite a few records as of late. Providing a safe commute for 1.3 million passengers with 4600 trips powered by renewable energy daily, it has improved its on time performance year on (boasting a 92.9% on time performance the first half of 2019). Working for its commercial and IT departments, we struggle daily with what Per Lundquist coins “a conspiracy to not keep us focused with so much going on” (breakout session, computer futures’ OKR Forum, October 3rd 2019). This was really the case for us being a program with 11 Scrum teams and three project teams having several dependencies within as well as outside the organization e.g. other operators and a clearing house. We struggled with wanting to attract more customers during off-peak hours, large scale renewal of legacy systems, policy changes due to the introduction of (European) legislation and incidents all of which drew our attention. But to summarize this: it really got us unfocused.

What we did: three steps

First order of business was to derive program objectives. We took a lot of inspiration and direction from our corporate strategy and set out to get three objectives. Together with everyone involved with the products and projects, we settled for five objectives. These were really qualitative statements. And to be frank: they were open to interpretation as well. But that’s fine and quoting Felipe Castro (computer futures’ OKR Forum, October 3rd 2019): “Your objective can be vague, your key results can’t”. The objectives did however appeal to the imagination and gained a lot of attraction, because they were just five and they set out a daring perspective for 2019. And we made sure we delivered them in such a way our colleagues would be able to talk about them at home if they wanted to.

Second, after setting the objectives, we asked product owners, stakeholders from both the commercial department as well as finance, control, legal and so on, to attend a meeting to validate, tweak and set the right tone for the objectives. After validating, we asked our senior management to choose three objectives out of these five to focus on the upcoming quarter. This is a crucial step to its success, in two ways: one, it ensured management buy-in on how we put strategy to execution. And second, it gave us focus. Doing five things is just too much. Three things to do seems to be the optimum.

Third, getting the key results in place was a bit more challenging. As with other large companies, we struggle to get measurable outcomes. Meaning we could not always relate a success with customers to a specific feature or service. The product owners set out to get measurable results that would correlate heavily with the three objectives. Remember, senior management already steered the program in a direction by picking three objectives where we would deliver key results the upcoming quarter. We used the first weekly product owner alignment session to challenge each other’s key results. “Looks great, how do you measure this key result?”, “you have a big dependency there, can we help to tackle this?” and so on. Making this work is getting all the product owners and key influencers from other departments at this table as well. If they need to do something to get a key result done, they should be there. In the process, you are setting a shared key result. That’s just cool. But more importantly, it associates the other team (of the other department or organization) nicely to your timetable this quarter, without the use of your own legal department and long legalese to set out the terms. The process of getting a shared key result is something we learned the hard way in the first half of 2019.

Results

Until OKRs were introduced, we usually set out to do something, but would get distracted halfway with something urgent. Then we would really accomplish a little of both. This is also because we were used to talk to each other in terms of years, not quarters. In neither case would we ask ourselves: “Is this bringing us closer to achieving the strategy?” Now, we have a quarterly focus based on objectives. Of the new big requests and features, we would simply say: “Save it for when we have an open dialogue about the key results for the upcoming quarter.” Worst case, you had to wait a few weeks. Just to be sure: we did allot time to solve incidents as quickly as possible and ensured policy changes were addressed (i.e. GDPR etc.).

We left the top-down organization in place and created a semipermeable OKR-membrane at the program level: based on objectives, the membrane allowed new developments in the market to get through but kept out others that did not add to the objectives. Thus, creating focus for 11 Scrum Teams and three project teams. You can start small.

Key takeaways

1. Get buy-in for the OKR way of working from at least one senior executive, manager or lead
2. Allot yourselves some time to get the objectives right: they should be inspiring and easy to tie-in to corporate strategy (We want to go to Mars via the moon!)
3. When describing key results, let the product owner and team members do it — and challenge them in an open session with all the product owners present. This way you can tackle interdependencies quickly and address a common timetable between departments (and organizations even!)
4. Try to start simple and keep it simple in the first year. Aim for inspirational objectives and measurable key results and refrain from adding all kinds of extra’s, such as the weighing of certain key results. OKRs really work well with Scrum. For example, you can derive Sprint Goals from key results.
5. Really take advantage of the quarterly focus. By allowing yourself to look forward to this quarter, you can really shield teams from distraction during the quarter. Thus, allowing teams and yourself time to set a course for a new or sustained direction the upcoming quarter: what is happening that needs attention? Do we have all the qualified personnel to tackle new developments? Etc. Key results that are achieved by teams add to fun and engagement. As we are rather result oriented, this is a side effect, but a most welcome one. Pride to be part of a company doesn’t come easy.