Complex vs. complicated

Complicated

When we are talking about complicated systems it is not immediately clear how the cause and the effect correlate to each other. We need to analyze the system or need to gain some knowledge to understand how the cause and the effect are correlated to each other. 

Usually we find a solution by using the following pattern: 

  Observe - Analyse - React

Think of a manual that describes how to build something. Or about an expert that knows how to interact with the system. 

Most of the technical systems are complicated for someone who is not an expert. An ordinary person will not know how a computer exactly works. An engineer can understand the system and predict how long it will take to execute a task. He can also estimate how long it will take to build a software or a computer.

Products are complicated if they can be estimated via knowledge or analytical thinking. Building a house is complicated but not complex.

Complex

Complex systems cannot be analyzed. Knowledge and manuals does not help to predict the outcome of complex systems.

Usually we find a solution by using the following pattern:

  Try - Observe - React

When we talk about try we do not mean try and error. Before we try we set up a hypothesis and test this hypothesis for validity. In agile product development we also talk about "Build - Measure - Learn". This means that the impact of implementing a feature cannot be clearly analyzed upfront. It may lead to increased sales or it may lead to more bugs and a complex system that is hard to handle.  After the new feature is implemented we can analyze the sales report and will get more clarity if the feature is worth the effort. 

The validation of our hypothesis is not complex any more. It may be complicated - as it is only possible with specific knowledge - but it can be analyzed and measured. 

Products are complex if the end product cannot be fixed upfront or if the product will change over time. We also talk about complex products if the customers needs are not fully predictable or if the customer group is very huge but not very homogen. 

Chaotic 

Systems are chaotic if the cause and effect chain is not present at all. Chaotic systems cannot be analyzed. It is also not possible to create a hypothesis and test if the hypothesis is correct. 

Usually we try to find solutions by using the following pattern:

  Do something - Observe - React

Meteorological systems, stock exchange and air turbulences are typical chaotic systems. 

Why is this important

If we create a product we need to think what kind of product we are building. Based on this knowledge we need to use the right pattern. Most of the time we make the failure that we are building complex products but try to do this with the pattern 'Observe - Analyze - React'. This will rarely lead to a successful product.