An interesting conversation came up a few weeks ago at a conference panel: there is definitely one strong misconception when it comes to Software Engineering, DevOps and transformations in general - if you want to start, you need a number of things which might seem necessary.
That’s why I strongly disagree. The reality is that these things might actually be overkill, useless or even damaging to your business. In order to start, you need a catalyst. And for the catalyst to be effective, you need to bring tangible value straight-away.
What does it take to trigger the butterfly effect?
Think about it, for any of the topics above: what do you need to get going? Anyone you ask will give you a different answer, and it usually relies on a technology, a certain flavour of a methodology or even worse a tool. Some will even throw in culture, for good measure. However I disagree. It’s not just that. These are all consequences.
Keep things simple. You want to start with DevOps? Don’t think about a tool, or any other substantial change that will bring you to a standstill. The business needs to keep going! Instead, start by introducing automation.
Automate something, and incrementally go from there. Scripting does wonders. Once the automation is in place it will reduce the amount of time a certain operation takes (compared to a manual interaction) and will also reduce the risk of errors. CI/CD is an excellent starting point, but it isn’t necessarily the only one. If you already have CI/CD in place but you want to go further, look into automating configurations, reducing manual maintenance actions, etc.
Long story short, the goal is to enable you and your team to re-invest that time you are saving into further automations and other (potentially bigger) changes. This is the butterfly effect of DevOps.