Wong Liang Zan

Wong Liang Zan

© 2020

Lessons learned from implementing OKRs

Objectives, Key Result or OKR for short, is something I’ve came across from blogs and people. As I understood how OKRs help a company focus, I became intrigued. Staying focused is much easier said than done. When presented with choices, we often choose tasks that we like doing, the easiest or the most urgent, rather than the right thing that has the most impact. It is hard enough to do it on your own, let alone an organisation. In OKRs, I saw a system that can keep an organisation laser focused as it scales. I didn’t have a chance to implement it; until I co-founded Propine. While there’re plenty of articles about what OKRs is, there aren’t much written from those who’ve practiced it. After two quarters of practicing OKRs, does it work?

Writing OKRs

How do you implement something that you’ve only read about? I’ve not worked in a company that practiced OKRs, neither did my co-workers. We don’t know if we’re practicing it the right way. We don’t know if it’s right for an early stage startup with less than 10 people. Nonetheless, we thought the long-term benefits was well worth trying. That we have not done it before did not deter us. We will do it first and correct things out as we go along. So we started writing our first quarterly and annual OKRs for the company. Writing the objectives were hard. It has to be high level, yet not so abstract that you can’t grasp it. It has to be ambitious, and yet not feel unachievable. It has to point the company in the right direction, in the face of unknowns. The Key Results(KR) were no easier to formulate. How do you design a measurable result such that achieving them means certain achievement of the objective? How do you estimate how long something would take to complete when there’re external factors we don’t control and unknowns that will arise? Are the goals realistic? Can we commit to them? We don’t have answers to them. After multiple revisions, we withered it down to 4 annual OKRs, and 5 first quarter OKRs.

Adopting OKRs

Our next step is to get everyone in the company to understand and adopt OKRs. To our delight, there were slides available. We set up introductory meetings explaining what OKRs were. We showed examples of OKRs and how it aligned everyone’s efforts towards achieving a common goal. Then we presented the company’s OKR for the year and the quarter, and we asked everyone to write their personal OKRs, and make sure they align to the larger objectives. People were open to trying OKRs since many illustrious companies were practicing it. We were genuinely curious what effects it could have. A week passed. With eager anticipation, we read those first drafts in a team meeting, and it was different degrees of terrible. Though everyone get the idea, designing and writing the goals isn’t easy. To improve them, we set up feedback sessions where everyone gave suggestions on everyone’s OKRs. We did one-on-one sessions to refine their OKRs. It took about 2-3 such sessions that we deem everyone’s OKRs satisfactory. We kept the OKRs on an internally available OneDrive folder where everybody can access. By then, each person had spent close to a day in meetings related to OKRs, not counting the time spent writing them. Though the time spent is not trivial, I see this as a necessary planning activity. Without which we’d be operating like a headless chicken.

Practicing OKRs

Now that we completed easiest part of adopting OKRs, we move to the harder part - practicing them. It’s natural that people feel highly motivated after setting a goal. Once the initial enthusiasm dies down, it is difficult to stay disciplined and focus on working towards those goals.

To prevent that, we adopted the process of Monday Commitments, Friday Wins from Radical Focus. Everyone would come together on Monday morning, and present what they intend to do for the week. Everyone would need to explain how those items align to the OKRs and what is their current confidence level in achieving the OKRs. It is meant to be short, no longer than 3 minutes long per person. More importantly it is supposed to be public - everyone in the company knows that you’ve promised to do them. It is brilliant in the way that it borrowed the psychology principle of commitment and consistency from Robert Cialdini’s Influence. Psychology research showed that public commitments helps to improve the probability of achieving them. At the end of Fridays, we come together and review what was done against what was promised. It is a time to celebrate work done well. When things didn’t go well, we seek to reflect what could we have done better rather than fall into a culture of blaming and excuses.

What worked and what didn’t

Gradually we realised what worked and what didn’t. The first problem was goals becoming irrelevant. Changes are common at a startup and as we learn more about the problem and as new opportunities arise, our goals change. It felt odd as we have committed to something that doesn’t align to a changed strategy. Yet, if we change the objectives, it felt like cheating. When is it right to change and when is it not?

The second problem is poor estimation of key results. Some of our estimates for key results began to look impossible as we gone deeper into the problem. That is expected since we didn’t understand the problem well at the start. Not being able to meet the key results was morale draining. Some were worried that it would affect their performance appraisal although we explicitly stated that OKRs are not tied to performance. Metrics makes it easy to compare and naturally people feel bad if they can’t meet them.

The third problem was we realised there is a lot of time overhead to implement OKRs. We need to hold meetings to teach, multiple meetings to refine them and finally one to rubber stamp them every quarter. We need to check and measure against it every week. When new hires join, it is back to square one. Like any habit, it takes discipline and persistence to do it quarter after quarter. At Q3, we didn’t manage to set our OKRs. Setting OKRs is time consuming and isn’t urgent. Investors won’t ask for it. Customers couldn’t care less about it. We ourselves won’t clamour for it. It is always tempting to push it to tomorrow. Tomorrow became next week, and next week became next month and it never gets done. If the founders don’t find the discipline to do it quarter after quarter, it won’t become an organisational habit.

So what worked?

OKRs are great for setting the context. It helps everyone understand why they are working on this task. It helps them understand how others contributes towards the larger picture. Understanding the context allows everyone to make correct decisions. We always have more things that we want to do than not. One example is the blockchain workshops that we conducted. We used to do workshops monthly in 2018. Workshops are great for marketing and community outreach. But it doesn’t align to our OKRs. It had no measurable impact on our sales figure. Hence I asked for it to stop so we can focus on other tasks. Looking back, it was the right decision. OKR is like a compass. It helps us to make choices, helps us decide what not to do, and do the things that matters. Once everyone in the company start making more correct decisions, like a boat with everyone rowing in the same direction, a company will charge towards it goals.

OKRs force you to plan. It force you to think things through because you need to write it down. The very act of writing clarifies your thinking. It forces you to weed out the fluff. It forces you to communicate to the company. That is why people procrastinate because that is hard. If you don’t, you end up passively reacting and not actively driving the outcome.

Back to the original question, do OKRs work? How do you measure OKR’s impact? Ideally we would like a control group where the control group startups don’t follow OKRs and another group which does. Then we track them over time, and we measure both groups and count how many exited and calculate the 95% confidence interval. But that is not possible in practice. Every startup is different and you can’t in good faith believe that such a study would prove anything conclusive. Given that there is no way to measure OKR’s impact, do they work? Rather than measuring, why don’t we frame it differently? Does setting the context help the company? Does long range planning help the company? If you know the answers, you’ll know what to do.