Most products today really have digital genes – data that just like genes define how to build, manufacture, sell, use and maybe even update real life “children” of your digital parents of these products.
Typical practice today is to have this digital product-DNA defined as a haphazard combination of definitions and interpretations across different and isolated systems in different categories such as PLM, ERP, CRM, CPQ and after sales systems.
The digital product DNA captures the product knowledge and IP so it is arguably the most valuable asset to a company. The way product modeling is carried out today is a bottleneck in business process, a source of errors and often handled by proxies to the real subject matter experts.
Jeff Immelt, CEO of GE
Tomorrow, we will need to act as software companies developing our digital product-DNA against a professional software platform and governed by employees with software development skills.
We have discussed how configuration helps you work with a product promise as opposed to standard products (see www.hoegeye.com/services) and how the cost of this approach is governed by the following simple equation:
Cost (promise) = Risk (error) x Effort (mitigate)
Configurators eliminate errors and they also tend to steal the spotlight. When companies evaluate CPQ-systems (Configure-Price-Quote systems) they tend to focus almost entirely on the end applications (the configurator) although it is the quality of the product models that determine the quality and usability of the configurators.
Designing or changing the UI and deploying configurators to new platforms and users are small challenges compared to establishing the necessary product modeling foundation.
Product modeling is much more important than the end configurator itself.
In fact, product modeling is the most important part of working with configurable products and it is becoming even more important as products grow more feature rich and software enabled.
The feature-flux – the pace with which features that describe or control products are introduced – is increasing and partly as a result of this overwhelming complexity we may be on the brink of seeing artificial intelligence supplement or replace both product learning (rules) and the traditional configurator enabled applications.
Both of these macro trends further drive the need for new product modeling platforms and strategies.
Product Modeling has been more or less at a standstill for the last 25 years. The overview below describes the three types of product modeling at play or coming into play today.
SOLVE BY EXPERT -VALIDATE BY SOFTWARE
The earliest types of configurators allowed experts to lead you to a final configuration. This approach is problematic because the experts get entangled in deeper and deeper if-then-else nestings and struggle to control state-variables.
This type of modeling is called “imperative”. It is still in use today and is the typical approach for do-it-yourself configuration but it is widely recognized to be unsustainable because it becomes too difficult to maintain the configurators. This approach is called a Validator because in order to make sure that the end configuration is valid it forces the end-user to follow a pre-determined sequence of questions with little or no guidance.
Excel-based configurators, custom software and many BOM-configuration applications in ERP-systems typically fall into this category.
MODEL BY EXPERT - SOLVE BY SOFTWARE
This type of configurator was introduced in the late 80’s and is still the predominant standard.
The idea is that experts only has to state the “what” and then the computer figures out the “how”. The “what” is expressed as a declaration of logical statements that constrain allowed product configurations.
This is why the cores of these configurators are known as “constraint engines” that solve the puzzle of satisfying all the constraints.
You will find a more detailed discussion of different variants of these configuration technologies in the book: “Knowledge-Based Configuration” published in 2014 by Felfernig, Hotz, Bagley and Tiihonen.
The one-man-one-model approach of the last 25 years is now breaking down under pressure from increasing feature-flux.
This means that “Product Modeling” is changing from a discipline mastered by (and limited to) a few individuals in an organization into an overarching concept called “The Modeling Enterprise”.
The Modeling Enterprise is a platform for continuous modeling with contributions from users across different business functions. The Modeling Enterprise is able to develop, track and update product modeling data throughout the product lifecycle and produce solvers that guarantee correct product configurations for specific business scenarios at specific points in time.
It is no coincidence that “The Modeling Enterprise and IoT” was the main theme of the Configuration Lifecycle Management Summit that took place in the fall 2016 in Marco Island, FL (CLM Summit).
The need for improved product modeling platforms and strategies is urgent and in the not so distant future the modeling enterprise will also be supporting artificial intelligence that gradually builds and learns product knowledge to provide an entirely new form of user guidance.