This post is from the CollabNet VersionOne blog and has not been updated since the original publish date.
Before You Automate
We live in an era that is all about automation – Cloud Automation, DevOps Orchestration, Test Automation, Agile, CI/CD, Containers, Microservices, Security, Data and so forth.
But what kind of assessment do you conduct before you automate? Do you just dive in? Where on that spectrum between proper assessment and diving in does your organization fit?
To have the best impact on the business and benefit from automation, you must assess before you automate. I rarely see people do this and I’m going to offer some suggestions. Keep reading.
The truth is automation still reflects efficiency because it reduces the time and costs of routine tasks, and it performs in a consistent way. Automation not only speeds up things, but very importantly, it reduces the risk of failure or omission of a process or processes. So, our industry impulse is to automate wherever we can. But we don’t always measure the value as we go.
Often, we only see the impact in the end when it is revealed through the financial analysis of CAPEX, OPEX and ROI. That’s too late in my opinion.
If you’re an engineer, sales engineer, sales or even a service provider you know that we’ve seen many companies understand the effects of more automation mainly as shown in that image. But there is another way that gives us more control and an ability to forecast the impact.
The big picture around automation is more complex, and by proper assessment through mapping, measuring and managing your value streams you can understand the value as you go and make strong decisions about your automation. You can even foresee the financial impact it will have on the traditional analysis described above, as well as other areas of impact before they occur such as potential service level implications.
But many of us realize it should be more proactive. But still, the motivation to automate an area is often a reaction to a bottle neck or an issue – automation is seen as the means to prevent a frustrating or recurring problem and many times it is adhoc with many tools and different teams automating independently of each other.
So, to be proactive, do you ask yourself, why should we automate? Have you analyzed why you Do or Don’t and what to automate first? Have questions such as these occurred to you:
- Which areas need my focus for automation?
- Are we spending more time Planning Automation or Planning our Business?
- What makes sense to automate? Can everything can be automated? Is that the best answer?
- Who should be included and what are the dependencies and relationships this automation needs to take into consideration?
- Is our approach to automation a short-term fix or lasting?
- What are the security implications?
You have probably thought about those questions and many others. Yet, I can tell you that despite those questions having been considered, in my 20+ years of experience I have rarely seen a team or organization conducting an actual proactive Assessment for Automation.
There are companies that have Automation Teams that are responsible for all automation and orchestration, but usually these teams act as an auditor. They get all the business perspectives (Target and Goal) and then they start screening processes and areas that can be improved through automation.
The best method I have seen used as an Automation Assessment is, as I mentioned earlier, Value Stream Management (VSM). It is the most proactive thorough approach.
It may sound strange if you’re unfamiliar with it, but Value Stream Mapping and Management is possibly the best way to understand the entire processes from Planning to Release.
VSM also gives a team and organization visibility through metrics and analysis areas that can be improved and automated.
As a DevOps coach I highly recommend you do an Assessment of your automation idea and check for the following:
- Can this be flexible – can it be changed based on my business demand?
- Is this Intelligent Automation or just Execution?
- How much impact will this Automation have when it runs?
- Will I have tools overlapping after the Automation?
- Will Operational Expense increase? (e.g., Will I need to hire people with specialized skills to operate tools and customizations?)
Like I said in the beginning, automation can be challenging but it has many benefits. If it is done in the right way, it can have financial benefits as well as reduce bottlenecks and frustrations for the actual teams using the automation.
So, keep automating, keep improving but keep the control of this in your hands!
Please share this with your community. See you in the next post!