Last week was the State Opening of Parliament in the UK. Before the event the Yeomen of the Guard search the cellars of the building using lanterns, looking for barrels of gunpowder. Why? In 1605 there was a plot to blow up the ceremony by placing barrels of gunpowder in the cellars underneath the House of Lords. 411 years later they’re still checking to make sure it doesn’t happen again.
I found this funny because my immediate thought was “So, it’s not just us then.”
At work we all build up processes and institutional habits over time, be that weeks or decades, and I’d wager that in most companies they’re still searching the cellars. Something happened once, caused a stir, and a check or process was put in place to try to ensure it never happened again. However as time passes it may be that problem can never arise again but the process still exists.
I can think of several examples from where I work, affecting multiple departments. For example, the group IT policy on website deployments dictates several days of testing on staging, then deployment over several days as each platform (database, web server, application code, etc) must be deployed separately, with lots of time separating them.
When we set up a new team, hosting on a new platform and with new deployment processes I asked why these restrictions existed. The answer was a series of ‘we had this problem once in a deployment’. Those deployments were all different technology stacks, different teams, different divisions even, but they contaminated the whole company. This stifled the introduction of new, better processes by trying to avoid issues that might never be able to occur on a particular project. Continuous integration & continuous deployment? No chance, totally against policy.
Because of this we set up our own independent hosting and support so that we could deployment how we wanted, when we wanted, like any development team should. Our ability to deploy quickly allows us to respond to a demanding business quicker than any other development team in the business and when you work in events that’s exactly how you need to be.
All this is to say: question everything. Every extra step in a process takes up precious time and focus. So optimising processes and making sure that you’re not searching the cellar pays back in efficiency, quality and morale.
Problems do occur and you do want to minimise them but where possible focus on fixing the underlying cause by improving processes, not multiplying then. If you are changing your processes make sure you’re doing it for the right reasons. If something happens once it could be bad luck, if it happens twice, then start thinking of solutions, not patches.