With Don Reinertsen to Flow: Put an End to KPI Illusionists and Metrics Magicians
What do illusionists, sleight-of-hand artists, and many controllers have in common? They lead our attention, our focus, down a wrong track. The trick or the company figures are explained to us verbally or non-verbally, but this is actually a feigned logic or a false relation. The magician surprises us with the trick. He has managed to make us believe how the game works, while this pretended logic is only show and the real magic.
What do illusionists, sleight-of-hand artists, and many controllers have in common? They lead our attention, our focus, down a wrong track. The trick or the company figures are explained to us verbally or non-verbally, but this is actually a feigned logic or a false relation. The magician surprises us with the trick. He has managed to make us believe how the game works, while this pretended logic is only show and the real magic. The link that the magician creates between his actions and the result does not exist and is the real illusion. Companies encounter the same problem with their metrics. After all, as just described, magic is not only performed on stage. Organizations fall for illusions of this kind every day. They use metrics and KPI’s to optimize their system, but in doing so, they fall for metrics whose actual interaction has not been sufficiently recognized. Often metrics are defined based on gut feeling or corporate history without looking at what they actually say and how they interact.
As an example of such a bogus metric, the number of Scrum Masters per employee in the organization or how many employees are trained in agility is often measured to determine the agilization of a company. However, this only provides limited information about agilization or the actual productivity of the team being managed. This is comparable to trying to derive the current speed of a car from its fuel consumption. Instead, it would make sense to measure how long a team needs on average for a user story or a new feature. Here, there is a direct interaction between metrics and the desired value, the agility of the organization.
What Don Reinertsen advises for flow is not to follow metrics for metrics’ sake. For this, it is important to understand that metrics for an organization are like the sensors of an airplane, they generate a situation picture of the system based on which the system can then be optimized. An organization can be optimized from two sides.
- system-internally: to create the best possible workflow / internal outcome.
- system-externally: to generate the highest possible impact on the market / optimization of the outcome on the market or from the customer’s point of view.
Accordingly, it is first necessary to decide which side of the coin is to be optimized in order to then select the appropriate metrics. In this way, it can be determined where the need for optimization lies and how it can be managed. For example, to manage workflow, information is needed on:
- Lead time of a single artifact.
- How many artifacts are in the system?
- How many of them are being processed in parallel?
These are first clues for useful metrics to create a situation picture in the first place. Subsequently, additional metrics can be added to further sharpen the pilot’s picture of its system. But just as every organization generates impact in a different way, the best metrics cannot be generic. Different impacts lead to different outputs. That’s why the best metrics are individual, direct, and objective. When working with them, the most important thing is that the statement about their interaction has been fully understood. Further, for the metric to remain meaningful, it is important that only those who have understood the metric work with it. And last but not least, metrics should always be seen in relative terms and never in absolute terms.
For example, a lead time of two days from one team should not be so easily compared and judged against a lead time of two months from another team. Where one team’s lead time refers to the construction of a steering wheel, the other team’s lead time refers to the construction of a multi-story house. Accordingly, with any metric, it is always the relationship that is most important. Metrics must always be considered astutely and with logical understanding. This is the only way to end the illusion.
Until the best metrics are found for one’s own system, experimentation is allowed. In doing so, one should start and act with an open mind. Metrics are never set in stone and can always be corrected if they do not provide the desired information precisely enough. After all, that is the ultimate goal of metrics: to provide precise and reliable information that clearly and unambiguously paints a picture of the current situation. Only in this way, on the basis of such clean data, the pilot can face the storm.