The last few decades, “going digital” has been the norm for many organizations. We often use the phrase digital transformation (DT) to deliberate, significant transformation of the organization with a major impact on its digital capabilities. DT-initiatives tend to also have an impact on products and services as well as interaction with (potential) customers.
The rise of artificial intelligence (AI) only strengthens this trend: it seems that the arrival of generative AI (GenAI) systems in general, and Large Language Models (LLMs) in particular have fueled the drive of organizations to jump on the digital bandwagon. The question is: is that justified? In this short blog, I will argue that it might be justified under the condition that the organization has their data management in order. To see what that is the case, consider the following observations:
DT: know where you are. DT is about making significant changes about the organization. If you don’t know what your current organization looks like – sometimes in great detail – then how will you know with any degree of confidence what the effect of interventions will be? Data about the current situation can come in the form of models, reports, process handbooks, etc.
DT: know where you are going. Winnie the Pooh famously said, “if you don’t know where you are going, then any road will get you there.” To avoid initiatives taking off in different directions, it is a good idea to have a sense of shared direction. This data tends to come in the form of strategy documents, architectures, and policies.
AI is data-hungry. AI systems come in many shapes and forms. The currently popular GenAI-systems are just the tip of the iceberg. A common aspect about these systems is that it tends to take much data to train them. Also, having it accomplish tasks may require even more data.
We could go on and on with this list. It seems safe to conclude that data is an important ingredient for DT-initiatives. This raises a new set of questions that focus on the availability and quality of (this) data. For example:
- Do we have enough data available?
- Do we know what the data means and where it is located (metadata)?
- Can we access the data when we need to (security, integration, interoperability)?
- Is the data a correct representation of the world under consideration? Is it complete? In the format that we want? (quality)
The conclusion that I’ve drawn is: data management helps to take care of data, which is a key ingredient for successful digital transformation initiatives. I wonder what your thoughts are on this conclusion. If you’d like to continue the discussion/share your thoughts: feel free to reach out.