Do you truly understand the business that you operate?
Operations is one of the most complicated and intertwined functions within a company. It is the core of whatever a company is doing, thus the main driver of customer experience (and with that, eventually a significant driver of revenues). For operations heavy companies, it usually includes the majority of employees and costs. To make operations leaders' life even more challenging, companies often allocate temporarily or permanently random teams - that do not naturally fit to other functions - to operations. As such we could also call an operations executive as the Chief Everything Officer - it's probably not a coincidence that they are often the natural successors of CEOs at companies.
Leading such an essential but difficult area, operations executives must understand the business in its details. As I mentioned in the previous article (see below), one core exercise in that shall be the development of a KPI tree for the venture.
What's a KPI tree? It's a powerful method to break down high level objectives or KPIs into their drivers to receive a thorough understanding what are the potmeters that managers can fine tune to improve their main business indicators.
Let's start with an example that I quickly drafted for an imaginary online retailer so that it's easier to explain how to build a KPI tree. Due to the limited resolution of your laptop or mobile screen, I created 3 snapshots of the KPI tree.
Deep-dive into COGS (cost of goods sold) and the drivers below:
Deep-dive into logistics costs and the drivers below:
The first thing that you need for drawing a KPI tree is a paper (or tablet) and a pen. Even if I use sophisticated analytics tools, I love to do this exercise on paper as it allows for brainstorming and faster iterations.
First of all, I recommend that you always start from profit. Even if you have an operations related objective, such as maximising the utilisation of some equipment or minimising the cost of the workforce, it's still a good idea to start with profit in order to get a complete understanding of the business, and so that you could understand how your area of responsibility fits into the overall big picture. Also, starting from profit forces you to also think through the revenues section of the tree - do not underestimate operations' impact on that.
Secondly, you should aim for exclusiveness and exhaustiveness. We called this MECE (Mutually Exclusive Collectively Exhaustive) at McKinsey. You neither want to miss an important driver of your KPI (exhaustiveness) - a good practice for that is to always include the "other" category -, nor wish to double count something (exclusiveness) and thus put too much importance on it. Always challenge yourself after drawing a tree: have I missed something important, have I segmented the drivers without having any overlap?
You should also go as detailed as possible. This may be counterintuitive for a C-level executive, but that's the only way to understand what levers drive your main KPIs. At the end of the day you want to attach actions to the respective drivers. If you went detailed enough, it will feel really simple to come up with new ideas how to improve the relevant drivers. To demonstrate this, I added some improvement lever ideas into my KPI tree example in purple.
While the KPI tree is quite standard until a certain level of depth, there can be different ways to segment the drivers beyond a certain point. For instance, you can segment number of customers based on geography, loyalty or age. You may choose to create the actually most relevant segmentation, but you may also create different versions of the tree.
Once you are satisfied with what you built, I suggest that you populate it with numbers. (Note that I haven't done this, given I illustrated the tree with an imaginary company.) I recommend not only adding actuals, but also historic numbers to spot trends. That helps you grasp where you need to focus on. Furthermore, you may want to find benchmarks for the different drivers on the tree to understand where you can have the biggest levers to improve the overall KPIs. For that exercise, of course you should have already moved from paper to Excel or spreadsheet.
Once you decided what are the main drivers to focus on, you should conduct two follow-up actions: you ought to ask your analytics team to set up dashboards to continuously monitor the relevant levers, as well as create meaningful objectives for your respective managers with the help of OKRs (objectives & key results). I'll write a separate article on what are OKRs and how to use them in operations teams.
What happens if your company grows large and it's difficult to monitor all the KPIs within the tree on a daily basis? That is when operations leaders usually focus only on the top layers and look only at aggregated numbers, failing to identify quickly enough negative trends on lower levels. In a later article I'll explain how to avoid that (teaser: by setting up automated monitors & alerting).
If you haven't developed a KPI tree yet, or if you have done it too long ago (that was me), you should grab a pen & paper as soon as possible and do it again. It will not only provide you a better understanding of your business, but also greater level of confidence and thus peace of mind.