Archive | Sin categoría RSS for this section

Lean thinking: Common sense and (apparent) contradictions

Lean is based on common sense, clear principles  and simple rules. And, at the same time, its conclusions are counter-intuitive to many. In fact almost everybody exposed to Lean has experienced that “when has this become so complicated!” moment when the pieces simply don’t fit together in your head. I see this very often when helping people with Lean. Some questions and apparent contradictions are very common. These are some of the most popular:

Popular contradiction #1: Splitting a process in parts and improving each part must necessarily improve the process.

Our common sense shouts out loud that this must be true. Well, it is WRONG, let’s see why. Imagine we are the owners of a pizzeria. This is our process:

Pizza 1.jpg

Our process has a bottleneck (oven), which determines the process capacity. There are also different constraints (oven, dough, ingredients) which vary depending on the demand. Any improvement action that does not improve the bottleneck speed will not improve the process capacity. If we split the process in parts (e.g. 1. Dough, 2. Ingredients, 3. Oven, 4. Waiter) and improve each part separately, we will be wasting time and money in useless improvement actions, as far as they not improve the oven performance (the bottleneck). For example, we might be working hard making the waiter service more effective, but the effect on the overall process capacity will be zero.

Pizza 2

Please note that this example assumes “speed = improvement”, it is a simplification to make it easy to understand. Of course there are other improvement factors like health, quality or cost that could drive meaningful improvement too.

Conclusion #1: Partial process analysis can promote useless improvement actions. ALWAYS understand the bottleneck and analyze the process globally.

After this example you may think: “Ok, I can waste time and money but at least we can’t do any harm”. WRONG again. Let’s see this example: imagine a kayak (the process) with different people (the functions) working together to make the kayak move forward. Improving a function can’t do any harm, right?

Canoa 2

CREDIT: Image created by the Lean Enterprise Institute.

This example shows that improving a function without considering the full process can make things worse! This situation happens all the time in real life.

Conclusion #2: Improving a function without considering the effect in the full process can worsen the global performance. Balancing the process is key.

The combination of a global process analysis (understand the bottleneck and constraints) and balancing the process is a winning strategy.

Popular Contradiction #2: Metrics always drive improvement.

Yes, we have always heard that what is not measured is not improved. Let’s measure everything! How can this be wrong? Well, it is. If you measure the wrong thing you can foster an incorrect behaviour.

Let’s see the previous kayak example. Improving one part of the process makes the global process work worse. Why does this happen? Why does only a part of the process improve? This happens in most cases because organizations use functional metrics (vs. global metrics) to define success. If the success of each paddler is measured by the number of strokes (functional metric), no matter the impact on the kayak speed (process metric), they will focus in their individual performance and work hard to make it bigger. Metrics are sadly promoting the wrong behaviour.

Conclusion #3: Metrics can foster wrong behaviours. Think carefully what metrics you’d like to use. Metrics must be meaningful from the customer point of view and measure global performance.

Using functional metrics to understand global performance is like trying to know a pizza temperature measuring just the temperature of the tomato sauce.

Popular Contradiction #3: Splitting in functions increases improvement.

This is a variation of contradiction #1 and, again, it is WRONG. Splitting a process in functions typically creates the wrong idea that improving each function performance must necessarily improve the process performance. It looks right but it is not true.

Let’s see an example. Sport teams have different functions working together. In soccer you have the goalkeeper (function #1: save goals), defenders (function #2: defend own goal), midfielders (function #3: connect defense and attack) and strikers (function #4: score goals).

  • A functional approach would take all functionally similar players from all teams and make them train together. They will learn from each other and grow faster. When all players come back from the functional training to their teams to play a match, the team will benefit from the functional improvement.
  •  A process approach would make all functionally different players from the same team train together. The will have a common coach to synchronize functions and make the process (soccer) flow.


What works best in reality? The process approach. There are thousands of teams with great “functional” players who, for different reasons, can’t play as a team. They rarely end up as champions.

Conclusion #4: Focusing in functions creates “functional sub optimization”, this is, using time and money in actions or projects that improve a function of the process and, at the same time, worsens the process globally (e.g. kayak example).

Functional Sub Optimization

Please note: Functions are very important because they take care of the knowledge (TRUE) but the process must go first and functions second. If the example above has made you think that Lean hates functions, forget it. This is also WRONG (learn more here)

Popular Contradiction #4: Splitting a process in functions improves process ownership. This is a very popular idea, but it is WRONG. Different functions owning their area of expertise looks like a good idea, but it does not work as expected in most cases. One process owner with enough cross-functional resources improves ownership and fosters improvement almost always.

Let’s see an example: think in a park and then think in your home. The park has its ownership / responsibility shared among multiple users with different expectations. Your home has one /two owners who are responsible of everything related with maintenance and “performance”. What is typically best used / maintained: a park or your home?


There are also many business examples. Let’s think in a hospital, what is best?


Conclusion #5: One process owner with cross-functional resources improves ownership because the grey zones of responsibility are eliminated.

Final conclusion: Lean rules and principles are based on common sense. Its apparent contradictions come typically from our brain trying to think the way is has always thought. Give Lean a chance!


The Toyota Way by Jeffrey Liker [free Audiobook]


Voltaire, détail du visage (château de Ferney)

Voltaire, détail du visage (château de Ferney)

Voltaire once said “Le mieux est l’ennemi du bien“, perfect is the enemy of good.

If you think, it is not a bad description of how PDCA works. Try a solution that makes sense, sounds reasonable or can help you learn. It does not matter if it is not perfect, just try it and see what happens. If you wait for the perfect solution before you start doing, you’ll probably wait forever.