Some time ago I explained how we replaced roadmaps for OKRs at Locaweb. In that article I also talked about some lessons learned. Recently we made a retrospective of the last quarter and guess what happened? We learned some new interesting things, which I will share here!
One of the points that drew attention during our OKR retrospective was the lack of clarity in defining success and failure criteria for each KR. In an attempt to create an equal analysis pattern for all KRs we defined each KR with a target value, a success criteria defined as the attainment of 80% or more this target value and a failure criteria defined as the achievement of 40 % or less of the target value.
The problem is that each KR has its own context and dynamics. While for a certain KR the achievement of 80% of its target value could be considered a good result, to another KR the same achievement of 80% could be considered not good and to be considered good, reaching 98% of the target amount would be required.
Because of this, each KR will now have three values, the target value, the value above (or below) which the KR will be considered OK and the value below (or above) which the KR shall be considered not OK.
When defining KRs, it is common to come across monthly measurement metrics, for instance, churn, new sales, revenue, margin, support contacts and so on. Since these metrics are monthly, during a quarter we have only two opportunities to evaluate whether what we are doing is having an effect in moving the needle. Two opportunities is not enough, if we miss the first one, we only have one more.
For this reason we’ve decided split monthly metrics in weekly values. The idea is to do this very simply by dividing the monthly amount by 4. For example, if the goal is 2,000 new sales per month, we should have at least 500 new sales every week. Thus, if a given week we have 380 sales, we know that we are short and that we should compensate the following weeks. Another example can be a KR to achieve a churn 2% per month or less. This churn rate of 2% per month is equivalent to approximately 0.5% churn per week. Thus, if a given week we have 0.3% churn we will be fine, but if we have 0.6% churn we will need to compensate somehow in the following weeks.
With weekly KRs we can follow easily and frequently whether we are in the right path and make appropriate actions to maintain or improve the KR without having to wait the month end to see how close we are.
The majority of the metrics that we are used to can be called lag measures or historical measure. They represent the result of something that already happened, but do not show what was done to make it happen. For something to happen it is important that we know what are the lead measures, i.e., what needs to be done. For example, when a person loses 10 pounds in three months, the lag measure is the 10 pounds lost in three months, but what were the lead measures taken to achieve this result? Probably she restricted her food intake to a given number of calories per day, restrained from eating some specific food categories and exercised for at least certain amount of minutes a certain number of times per week. These are the lead measures that led to the described weight loss lag measure of losing 10 pounds in three months. Another good example, now that we are watching the Olympics, are the sports results. For example, US is leading the Olympic games in terms of number of medals. This is the lag measure. And what were the lead measures to achieve this result? All policies related to sport incentive in the US.
KRs usually are lag measures. NPS, Average Call Duration (ACD), new sales, revenue, cost, churn, number of contacts, etc. All these metrics are lag measures. But what are the lead measures? What should we do to change their values? What does influence the final result? It is very important to know these influencers in order to understand where to work to move the metric. At Locaweb we mapped the influencer for NPS and ACD. This work consisted of some brainstorming sessions where we try to identify everything that influences the NPS or ACD. Then, for each item, we evaluated whether it was already measured, if the measured value seemed to be OK and its impact in the lag measure, using a simple scale such as Small, Medium or Large. We got a good guide of what should be our focus in order to improve NPS and ACD.
Another tool that can help is the concept of visible management. This concept is widespread in agile methodologies, with the teams having a board which allows the constant observation of how are the key metrics and, in some cases, its historical evolution. The use of visible management originated from the production system used in Japanese factories known as kanban. Incidentally, the Japanese word Kanban literally means sign or score.
The idea here is to create a visible board that shows what are the OKRs and how they are progressing. It mades it easier for the whole team to follow the progress of each of its metrics.
However, only having a visible dashboard is not enough. It needs to be constantly updated to be useful. Hence the fifth lesson learned.
This is another tool that we are borrowing from agile methodologies. As explained above, besides creating its visible dashboard, the team needs to update this dashboard frequently or else the dashboard will become another office decoration piece. The suggestion is to update it weekly since we discussed earlier how important it is to update KRs weekly.
A standup meeting no longer than 15 minutes to update the KRs where each team member tells what she did last week and what she intend to do this week to improve each KR. This looks very simple and it is very simple indeed, but it helps to improve communication and team commitment.
We have learned a lot from the use of OKRs. As we are now measuring more stuff, this helps us better understand how things are moving – or not moving – and identify areas for improvement. We are confident that we will continue to learn a lot and will continue to share as we learn.
Have you already implemented OKRs in your company? Please share your lessons learned so everybody can improve faster!
In 2015 I wrote a book on Software Product Management in Portuguese. In the beginning of 2016, Paulo Caroli talked to me about how he enjoyed the book and how this book could be useful to people in the software industry not only in Brazil but anywhere in the world. For this reason, we decided to create an English version of my book.
The book is organized in 5 sections:
This book is suitable for anyone working with software. Even companies that do not have software as its core business use software in their day to day and often have developed some software that interfaces with its customers such as a website or a mobile application. It is important for these companies to understand the software product management role and responsibilities, so they can better manage this software and increase its chances of success.
We are working on the translation but as we progress we are already releasing the content. If you want to see the work in progress, please visit the book page at LeanPub. Still in beta but already with valuable content. Feedback is always welcome!