Automation is an exciting ambition in the world of digital transformation. Everyone aspires to implement a lot of it. It gives you leverage, allowing you to do more with less. Who wouldn’t want that?
There are two main problems with automations though:
- For many scenarios, the cost of automation outweighs the benefit. Either the benefit is low (you’re automating something that doesn’t move the needle, especially for low-volume tasks), the cost of automation is high (people might calculate the up-front costs but often forget the maintenance costs and knock-on costs of digital automations), or both
- They don’t know the full complexity of what they’re automating, often building an automation that’s incorrect because it’s incomplete
Hence I suggest to follow the 5 Ds of automation rather than just building an automation because it seems like a good idea:
- Do (the task manually, usually many times)
- Document (codify what you’ve learnt)
- Demonstrate (that the documented approach is repeatable and reliable)
- Decide (whether it’s worth automating – do the benefits really outweigh the costs)
- Duplicate (build the automation that performs the task repeatably)
Note that the 5 Ds approach takes a narrower perspective on automation – of automating a point solution / problem / task / workflow. We can also expand the frame of view to cater for a more holistic application of automations.