"AI transformation" is really the wrong word

"AI transformation" is really the wrong word. It is always a transformation of the company, and AI is an occasion, an enabler, an accelerator.

For me, transformation begins with the customer:

• What is the customer's request?
• What is the end-to-end process? From the first request all the way to fully fulfilling that request to the customer's satisfaction?

In most organizations, this end-to-end process contains cumbersome steps that grew historically, breaks between media, and too many organizational units handling it through handovers.

That is why transformation always also means digitalization, simplification, streamlining.

You need not only knowledge about AI, but really a great many other kinds of knowledge, such as: what is the customer journey, what UI/UX do we need, what does end-to-end responsibility mean, what are agile ways of working, how do you work with hypotheses, what is Lean, and which modern leadership structures do we need.

Only once that is clear are you able to define lean, effective end-to-end processes pointing toward the customer.

You need a frame, a north star that shows you the direction.

After that, you pick a process that contributes to creating value for the customer and that has also become somewhat complicated over the years.

In many companies there are processes in which cumbersome manual entries have to be made along the way, or in which only a few employees know what to do in which special case. These are all potentials that can now be unlocked.

Or there are dozens of managers, experts, and committees whose permission you have to ask before changing anything. That is an obstacle too.

As soon as it is clear what the desired outcome of a customer-facing process is, you need clear responsibilities. Who implements and decides? Who supports and helps? You will not get far with a rigid organizational structure. You need teams with decision-making authority and leadership teams that enable and improve the system.

For implementation, you proceed in small steps: What are the hypotheses, and with which experiments can you check whether you are on the right track in improving the process and using AI?

You also need kill criteria: when it is clear that something has not paid off and should be discarded.

It is a learning journey. What you learn with one process gets scaled to other processes. The organization needs a learning rhythm.

Throughout, the customer has to be at the center, and AI is merely a means to an end. AI gets deployed in the places where it fits.

The moment you start this way, you automatically begin a transformation, because you are rethinking your processes from scratch.

What matters is that the first process must be productive right away, from the very beginning.

It is equally important NOT to do a great many things. Leave them out. Lean thinking. And do the important things RIGHT instead.

That is why, in my opinion, you need an AI consultant who has exactly these strengths. AI knowledge is quickly acquired. But not everyone has the experience of helping companies streamline processes and organizational structures.

Write to me: which questions remain open for you after reading this page?