This is part 2 of a series of installments of our white paper: How to create an automate-first culture without losing control. Missed part 1? You can read it here. We will be releasing the entire paper in installments on the blog - but if you want to read the full piece - you can download it here
Part 2: The different ‘flavors’ of automation
When it comes to automating ‘the right way’, a lot of the key considerations depend on the overarching goals for introducing automation, the process involved, your risk appetite, and the potential impact of failure or poor performance.
There are two flavors of automation which are outlined here:
- Fully automated - This involves minimal or no supervision, this process is highly robust and successful, with a good track record. It can be run with confidence.
- Human augmentation model - This is where humans need to be involved. Whether this is to run assurances at key breakpoints, or intervene at brittle stages for risky processes, adding knowledge and skill.
Neither of these options is better than the other. Which you use is completely dependent on the trust in the automation, and risk and significance of the processes being undertaken. Increasingly, we’re seeing organizations leveraging both of those models, leading to the question: when is it right to use each model? Some guidance on this is outlined as we continue, but it is also highly dependent on your organization’s existing automation landscape and preferences.
Don’t underestimate the value of people in an automate-first culture
As we can all identify in ourselves from time to time, humans are not very good at repetitive manual processes; we get bored, we disengage, and we experience ‘brain drain’ when tasked with doing the same thing over and over. All of these uniquely human traits can lead to error, which is why the introduction of machines and automation - whether as a replacement, a verifier, or a guiding source that can better inform a tired brain - is so beneficial. That said, you can’t put a price on human knowledge and experience, and this is often naturally codified within a particular person or team in an organization. This is precisely why documentation and knowledge sharing is so important to mitigate the risk of losing this crucial insight. As you introduce automation, it is crucial to prepare for failure and retain all relevant knowledge, in case there’s a time when your automation doesn’t work as expected.
How to avoid a loss of knowledge while optimizing for automation
With a hybrid approach becoming the norm, it’s so important to maintain any legacy knowledge needed of onsite tech processes. Here’s the panel’s advice for this challenge:
- Don’t rush. Automation benefits are huge, but it is also a force multiplier - if you go too fast and automate in the wrong way, you could bring more fragility to your processes.
- Risk mitigation is crucial. Review existing manual or semi-manual processes to understand the steps, but also take the time to find out WHY you have these in place and do some spring cleaning - it’s the perfect time to get rid of unnecessary/ unwanted processes.
- Always plan for failure. Keep a list and a clear understanding of the potential risks you face, and ask yourself ‘what happens if this ceases to work?’ This will keep you better prepared.
- Always have a plan B and practice it. This advice relates to the point above. Organizations that are successful at this have a plan B, and even C, built into their methodology. They not only know what to do when things go wrong, but they have confidence in execution because they’ve actually tested it.
The conversation moved on to explore four common threads to automation, as outlined by Marcus Wildsmith, CPO at Cutover. To find out how to bring automation into the light, and the reason why ‘clutch control’ is crucial to effective automation - make sure you don't miss the next post in this four-part series.
Why not read the next section (and the rest of the installments, including a must-read Q&A) now by downloading the full white paper here.
