Why companies fail at IT automation
A lot of companies are stuck in a “transformation process”. This blog post describes some of the reasons.
A couple of years ago I started a job at a small IT company. By that time I already had 5 years of experience from IT Automation projects at the university where we ran clusters with up to 100 nodes and the only manual work that had to be done from time to time was changing some failed hard drives, network components or entire machines. Things that could not be automated were documented. Documentation was an integral part of the process. No one would have ever thought of doing “documentation later”.
But that company I started with was full of manual work. It managed hundreds of servers and literally there was no documentation at all. I thought that would be a great opportunity to help them automate their business and use that as an opportunity to learn how to automate heterogeneous environments which of course is different from automating a cluster that has not much variation between nodes.
Some colleagues agreed that we had to go forward and professionalize the workflows. Every time we made proposals to the management we heard something like “Automation and documentation is always a good idea”. By that time I did not understand what that actually meant. I thought it meant that our ideas had fallen on fertile soul.
When we actually started doing plans for automation the reply changed to “You can do that automation thing as a side project, it does not have to delay your daily work”. When we proposed to install Dokuwiki on an existing server the reply was something like: “You can do documentation in text files or e-mails. We do not want any additional software that adds complexity to our systems.”
At first glance that seemed to be the opposite of “it is a good idea”. But as I understood later it was exactly the same thing.
Automation is a big investment and it is extremely difficult to calculate the amortization time. That makes it difficult to compare it to other investments like developing new product features. But of course, if we can have it for free, we will take it.
I decided to take my efforts to another level and created a spreadsheet listing the cost of all the unnecessary work I had done in the past months. Once I sent that to my boss it seemed that it had changed something. Within the instant of a moment he was interested in automation and professionalizing our workflows. Well, at least for a couple of days then his interest was gone.
Yery soon I decided to leave that company and that was one of the best decisions in my professional life.
However it left a big question open: “If it is not money that drives companies to DevOps and automation, what is it?” I realized that the answer is deeply seated in the core beliefs of a human being. The majority of people focusses on prestige and money. They think in status and hierarchies. Earning a lot of money, having a PhD or a job title like “Head of Senior Lead Developers” or something like that is their definition of success.
On the other hand there are people who we call successful despite the fact they don’t have job titles, degrees or tons of money. Think of Nelson Mandela, Mother Theresa or Martin Luther King. Their definition of success was using their precious time to make the world a better place.
Think of Google and Amazon who have taken IT Automation to a completely new level with Kubernetes and AWS. They are not successful because they focus on money. They are successful because their vision is to do something meaningful and that involves doing it in the best possible and most efficient way.
I am very grateful for the experiences I had at that company even though it cost me more than 2 years of my life. But the lesson I learned is priceless.
If you can relate to that experience, do not try to change your environment. Leave it.
