When Digital Twins Don't Keep Up

The buzzword of Digital Twin has been around for a while. In strategic and operational planning a digital map of all organisational processes comes handy. It allows strategists to play around and test scenarios with real time data whilst seeing the impact of changes to the bottom line.

In theory that sounds great but such a Digital Twin is often built on several shaky assumptions:

1.) Accuracy

While surveys, reports, standard operating procedures, data warehouses, and log files are great sources for process mining the processes are often not fully digitised and therefore not harmonised across silos. Rework or tacit knowledge might be captured or not. But most of all, these processes depend on people. People come and go. Two people might not perform the same process the same way. For all these reasons the processes in our Digital Twin cannot be accurate.

2.) Timeliness

If data is not captured in real time it might be outdated by the time it is captured in the digital twin. Depending on the time lag such distortions can lead to completely misleading input which in turn leads to the wrong output and hence to errors in strategic planning and operat.

3.) Completeness

As long as processes are run in silos and are not drawn out across the whole organisation they can never show the complete picture. The Digital Twin is at best fragmented. And these fragments are pieced together by assumptions derived from surveys. As long as there is no common understanding across silos of organisation wide processes, there will be no complete Digital Twin.

4.) Relevance

Obviously, where timeliness, accuracy and completeness are not guaranteed scenario testing cannot deliver accurate and reliable results greatly reducing the relevance of working with scenarios.

5.) Real Impact

Let's assume for a second that the assumptions on which our Digital Twin is built were reflecting reality despite all the shortcomings in the above mentioned assumptions. Implementing the strategy derived from the scenarios modelled in that Digital Twin might require another Change Management process. How will we then guarantee that the results will be executed instead of being summarised in a report for the shelf? Only by making the findings repeatedly executable will there be real and lasting impact.

It is actually possible. A properly designed and implemented process layer will achieve process excellence, given that:

a.) Such a layer communicates with the relevant silo systems

b.) Standard Operating Procedures are implemented into digitally guided processes

c.) Workflows come with detective controls, an escalation engine, and documentation

Once processes are digitised they can serve as an accurate and timely single source of truth. At that point changes in the Digital Twin can translate directly into process changes and these processes ultimately impact the bottom line as expected from scenario testing.

Governance, business process continuity, and cost savings are nice side effects stemming from such an effort.