Why Your Digital Transformation and AI Initiatives are Going To Fail

I hope that at least 84% of you are sitting down.

I am not hear to tell you what you want to hear, or sell you a whiz bang solution which solves the problem for you. In my next article I will talk about what you can do about it, but for THIS article, I am going to tell you why you are not going to get the promised value out of that Digital Transformation, and AI project.

Last year Forbes had an article (12 Reasons Your Digital Transformation Will Fail) which, besides estimated the 84% failure rate, really dissected the problem, but, though I think it was good, solid article, it discussed 12 symptoms, and not root causes for failure. Since then the AI star has risen, and I am seeing the same mistake all over again.

So …. without any extraneous preamble …. what are the root causes?

We are trying to solve the wrong problem

In the Harry Potter books, the main characters have magic wands which allow them to, among other things, clean up messes without all the work. Though this is great in fiction, it ain’t real.

Though technology can do many marvelous things, it’s “realty” is based on a foundation of process and data. If the foundation is broken, the results will be as well.

Let us consider a simple example.

Robotic Process Automation (RPA) is one of the technologies in a Digital Transformation. Still, in the majority of implementations we use it to go around a questionable business process or to “fix” bad data. Why is this a bad idea? Because instead of fixing the real problem (i.e. the questionable business process and the bad data) we have put a technical bandaid on it, which will never hold long-term.

In actuality, you have just increased your technical debt, which will increase the time needed to support your RPA bandaid and reduce the time for using your RPA developers for innovation and getting the real value out of the technology.

We think that Digital Transformation is about technology

You have heard this one before, probably from your Organizational Change Management team, and they are right, but it isn’t all about fear of change.

It is about unlearning bad practices

In the pre-paperless office was even thought about, we ran on paper. When Department A needed Department B to do something, they would fill out a form (or talk to someone who owned it in Department B), Department B would get the work done (which someone would check was done right, and, if it was cross-functional, probably pass to Department C (with a form they created). Though we killed lots of trees, the handoffs were fairly clean, and Dr. Deming and many others, helped us become more efficient.

In the post-ERP/Core Application days, however, unless the technology required it, people spend time interpreting what it means. Some fields may be non-standardardized, blank, or just plain incorrect, and instead of going back and fixing, some companies just fix it, during close.

Does accountability really exist?

In the old world, the department would be responsible for the form, and what was passed down to the next department. With mostly hierarchical structures this was enforced by the VP/directors/managers with increasing responsibility/accountability for their scope of control.

Another Forbes article (Your Accountability Problem Is A Leadership Problem: How To Unlock Next Level Performance) talks about this problem in more depth.

In the modern more matrixed world, accountability has become fuzzy in many companies. Many Shared Services disconnect the function or operational unit, and isolating them from these shared processes. Where there used to be one leader responsible, now their may be five or six, with plenty of blame to go around.

For a Digital Transformation, this accountability is hugely important, especially for the Process and Data Governance. Separating the Process Governance and Data Governance adds in another layer of possible misalignment, again, limiting the possible success.

Is it about technology at all?

Technology is about enablement, so Digital Transformation does have a Technological component, but most failures have absolutely nothing to do with the technology, but with the people and processes.

Summing up

I don’t want to sound gloom and doom! It is not hopeless, not at all, but understand you can’t just slap technology and make a Digital Transformation. You also can’t take your bad data and have AI really move the needle. It is garbage in, garbage out. Most of what we talked about was Structured Data … data in the applications … just imagine your documents and your other Unstructured Data.

It isn’t quick or easy, but you CAN make it work. In next week’s article, I’ll discuss ways you can attack the elephant and make tactical gains while doing the work needed to get the real value out of a Digital Transformation and AI.