Welcome to a new year of my musings on #BPM and I am very grateful to have gathered a stable following of my newsletter. Once again I am committed to publish a number of articles this year on the wonderful topic of business process management and all of its facets. I hope you all have enjoyed a bit of time off over the holidays and the new year, I know I have. I barely touched my work equipment and it was kind of refreshing (this has been different over the first 25 years of my career, then again, the older you get, the wiser you become typically).
Anyway, let’s get on with it in 2026…
Over the course of the last couple of months I witnessed something interesting. Amidst the avalanche of AI news I have been seeing more and more news items that stress the importance of process documentation (or process modeling if you insist, or process management if you want to take a slightly broader view). There were personal articles on this, but also studies done by consultancy firms with the same message: if you want your enterprise AI to be successful, you need a sturdy foundation in the form of a digital twin of your organization.
Let’s be honest here, a #DTO is not something new. I’ve been talking about this (and not just me by the way) back in 2018-2020. Granted, the technology was not at the same level as it is now and thus a modern DTO can comprise out of many more facts compared to the first DTOs, but the concept is exactly the same. Why is this important then all of sudden?
The root cause for this lies in #AI and its applications. The sheer speed and breadth with which AI is finding its way through your organization is astonishing and it is incredibly difficult to keep up from a governance (process & data governance that is) perspective. We all know, deep in our hearts, that you need to govern something as powerful as AI otherwise it will wreak more havoc than anything else. To make a terrible analogy, you do not put a black bear in the same cage as some furry rabbits without decent safety protections for the little bunny hoppers, do you?
With AI it is even more important to have decent governance, because of the GIGO factor. If you put garbage into AI, you will not magically get useful things out of AI. On the contrary, AI loves to just put one and one together and come up ‘window’ unless you tell it specifically what you expect and within what boundaries it needs to stay. So, translating this to your enterprise context means that you need to have a pretty darn good idea of how your organization is supposed to work (documented, orchestrated and automated processes) and how it actually works (mined processes) before you let an enterprise AI loose on your operational data.
Let’s find a better analogy here. Imagine I have two mountains. One made out of rocks (a traditional mountain so to speak) and one made out of sand (a Sahara dessert mountain if you will). The rocks depict a sturdy DTO (documented, mined, automated, the whole nine yards). The sand represents a more loosely connected organization. AI can be seen as a blanket of snow that is poured out over the mountain. In both cases it will lead to an avalanche ultimately, the big difference will be that the rock mountain will still be standing and will be able to guide the avalanche in order to get something useful out of it (ski slopes for example), while the sand mountain will simply be leveled completely.
In other words, if you have a good foundation in the form of a Digital Twin of your Organization (DTO), you will be able to guide the application of AI to a much higher degree compared to when you do not have this foundation. One of the cornerstones of this DTO foundation are your documented processes (among other things like mined processes for a required reality check) and this explains (at least partly) why there is a ‘sudden’ revival for process management going on right now. To put it more popularly: we’ve entered the FO phase of FAFO and we’re coming to the conclusion that we’re missing something and that something is process documentation (as a representation for aligned, harmonized and standardized (where applicable) processes).
What do you think about this? Please comment or repost with your own thoughts!
See you next time…
Ciao, Caspar
