[EN] 3 principles of good KPIs that works for me

There are three principles that guide me when setting good KPIs:


1. Does KPI help me to make decisions?

The main task of each KPI is to support the decision-making process. The basic question is whether we need a given KPI at all? How will it affect business? How many customers will you be able to acquire thanks to it? How much money can you save? How many accidents at work you can avoid? We have to realize that KPIs cost money. Time is needed to calculate KPIs, put in presentations, print and post on the board, review with the team, review with management ... The time of the people who calculate it and the time of the people who review it. And these people often have a very high cost of a man-hour. It is not a big investment at one time, but if we multiply it by the number of KPIs and the number of reports and reviews during the year, we can get a really large amount of money. Therefore, you should always consider whether a given KPI is needed at all, and if the investment worth it? If it will only be a decoration on a board or PPT slide and we will not make any decisions or any actions thanks to it ... then most likely it not worth it.


2. KPI measures the process, not the people

I think the prevailing paradigm that KPIs should measure the quality of our (people) work is very detrimental to the organization, at it causes two trends:


  • Correcting the results. If something is not our fault, we should exclude it from the calculation. KPI will look nicer and more honestly present the results and quality of our work. This way, we may not see the problems our client sees (whether internal or external). And what's worse - we may not take any action for this reason. Because everything is ok. Or maybe not?

  • Measuring only those KPIs that we have influence on, not measuring those that we have no influence on. Very similar effect to the previous one. Avoiding measuring indicators, on which we have no direct influence, just because we will look bad, not necessarily because of our fault. Here, we can also miss situations where we need to take the necessary action (we personally or we as an organization). This most often applies to KPIs that relate to inter-division gray zones or external factors such as the customer.

Both situations distort our picture of the whole. If we treat KPI not as an evaluation of our work, but of the efficiency of the processes we manage, we would not be tempted to avoid or hide problems that need to be dealt with.

Going further, the key is to, first of all, do not measure what we do but what our processes give to the organization. What products these procesess generates, which are used by other processes in the company or are delivered to the customer. And these products should be measured in terms of timeliness and quality. And this is really the most important thing because the functioning of the rest of the organization and the satisfaction of our customers depend on whether we deliver our products on time and in the right quality or not.


3. Endless goals and trend

It is a practice all over the world to define a goal for each KPI. 95% OTD, max $ 2m of poor quality costs, no more than 5 complaints etc. This also causes some pathologies in the form of failing to act and failing to respond to abnormalities as long as we are on target.

Or, on the other hand, if our results are well below the target and we are (in example) tasked with going from 60% OTD to 95%, the target may seem unrealistic and we may feel paralyzed. Because what big projects would have to be implemented to achieve this target. What could get us paralized and do nothing or spending month on design this 'big project'.


An alternative that I have been using in recent years is defining an infinite or final goal for KPIs. If OTD, then only 100%, if the costs of poor quality, then only 0, if the number of complaints, 0 as well, if the budget, then only better than planned(not “as planned” but should be better! How much? As much as possible!). This seems paradoxically even more difficult than the goals described earlier. But if we define them as goals to be pursued (not achieved) and our task is not to achieve 100% OTD, but to take continuous action to get closer to that goal. Every day, every week, every month. When planning activities for a given year or quarter, we must answer the question: what can we do during this time to get closer to our goal?

With this approach, we can get fantastic results that are driven by people's attitudes. People who are not limited by the finite goals they sprint towards. But people who know the direction we are going to take and who take it into account in all our daily decisions. People who focus not on arbitrarily defined levels of metrics to be achieved, but on the constant pursuit of excellence, on continuous improvement. That is what is so valued in the Lean Manufacturing approach.

At the same time, it is important to observe the trend - are we really going in the right direction or should we do something different than before?


Thanks to these three principles, I am able to provide such effects as:

1. A small number of KPIs (and in some areas no such KPIs)

2. KPIs that are relevant to the process and actually show how much it is

3. Evolutionary goals that support Lean / Kaizen philosophies on the way to perfection.