Eliminating the complexity of customization
In the previous blog post, I argued that when CPQ systems are treated only as sales efficiency tools, they fail to address the underlying complexity of customization.
The configure-to-order paradigm attempts to systematize product customization by pre-defining customizable parameters of the products. The challenge is that those parameters have overwhelming interdependencies, and those interdependencies interact and combine in ways that are hard to comprehend. When encoding those interdependencies as rules in the CPQ system, mistakes are easily made. And most CPQ systems provide no overview of what the combined indirect consequences of all the rules are. Millions of combinations cannot be tested. The CPQ is merely automating the mess of manual customization. The complexity remains.
So what can companies do about it?
In mass production, the key to success is that we can model the products and the production processes abstractly with a high degree of accuracy. We can do this by having (i) sufficiently intuitive representational correspondence between the model (the design) and the real world, and (ii) tools that can process the models in various ways, to enable us to validate the correspondence, e.g. 3D visualization and process simulation.
Although mass production models are complicated, they are accurate enough to make mass production sufficiently predictable, and therefore manageable, with expertise and good tools. That is what makes mass production immensely efficient.
In the configure-to-order paradigm, this modeling lesson has not been taken seriously. The focus is on conventional ease-of-use of entering each rule, neglecting the challenge of interdependencies that is unique for configuration logic. The result is an automated mess.
With a modeling approach, the configuration logic can be documented concisely in a way that has (i) an intuitive representational correspondence between the model and the real variability and interdependencies in the products, and (ii) tools that can process the model, to overview its consequences and enable validation. Tools can also use the model for multiple purposes to support different business processes. We call this a smart product definition and smart navigation of that product definition.
(The difference between the traditional rules coding approach and the modeling approach is similar to the difference between 2D CAD and 3D CAD. A 2D drawing of mechanics is easy to enter in 2D CAD on a PC line by line, but impossible to validate and use for multiple purposes.)
The configuration logic is a core asset of any company that wants to provide customizable products efficiently. It has implications for every decision in the company related to products. Thousands of decisions every day depend on that asset. It deserves much greater attention and care than currently is the case.
Product management can be taken to a whole new level when configuration logic is modeled in a way that is accessible to product management, with tools that support decision making before detailed engineering begins. The range of customizability can be optimized for profitability, instead of being driven by ad-hoc decisions about new variants without any visibility of the consequences.
There is an unavoidable complexity in product management and engineering due to the unpredictable business environment. However, companies that understand how to manage configuration logic can eliminate the complexity that is due to customization, to combine the efficiency of mass production with customization and fulfill the promise of mass customization and the configure-to-order paradigm.
Companies that succeed to manage configuration logic will experience a magnitude of improvements in efficiency and effectiveness throughout their business processes. It is the holy grail of B2B manufacturing.
And this is not only a matter of profit. Society needs customization to avoid waste. And we need a new green generation of products, which requires transparent processes of design, production, maintenance, and recycling to get improvements out to the market quickly.