Why do automation projects fail?

Automation is not an option. It provides a huge opportunity in terms of time to cash, efficiency and customer experience. And with the COVID-19 pandemic, automation only became even more important in order to quickly respond to the changing context.

However, when you do a little market survey, you get a mixed picture. People have quite some frustrations about their automation projects, because they don't deliver the results they expected. Some have invested in huge projects involving 50 to 100 persons but see little or no returns yet. Others claim automation still has to prove itself, that we are only at "partial automation" while the real value is in end-to-end automation, which is still a long way ahead.

So what's going wrong?

That's the question I have been asking myself. End-to-end automation is not a dream or illusion, it's real and up for grabs. So why do so many automation projects not succeed in delivering up to the expectations? Because the basics of automation are not in place. We're thinking too far ahead and want to go immediately for the most complex use cases like network slicing and closed loop automation, while we forget to build the foundation for successful end-to-end automation.

It is important to tackle these issues once and for all. A few weeks ago I gave a presentation at the Layer123 World Congress about this topic to share my vision about the key issues and how to resolve it. I received a lot of positive feedback and confirmation, so I decided to share our lessons learnt also through a series of articles.

In the next few weeks I will write an article for each key issue that we have identified. For the ones who can't wait for the separate articles, you can find a video of my presentation below.