Running the business · 4 min read
Which process should you automate first?
A ten-minute way to pick the one process worth changing first, and the ones to leave for later.
ZoolCoder
Start with one process, not a platform
Most teams know something is slow. The hard part is choosing where to start. A first project that tries to fix everything tends to fix nothing, so pick one process that happens often and that people already complain about.
Four signs a process is ready
A good first candidate usually shows all four:
- It happens every day or every week, not once a quarter.
- It follows the same steps each time.
- Someone moves information by hand between two places, such as from an email into a spreadsheet.
- Mistakes are costly, or slow to catch.
Score it in ten minutes
For each candidate, write down four numbers: how often it happens in a month, how many minutes it takes each time, how many people touch it, and what a mistake costs.
Frequency times minutes gives the hours it takes each month. A worked example: an invoice typed in 200 times a month at four minutes each is more than thirteen hours a month, before anyone fixes a typo.
What to leave for later
Some processes make poor first projects, even when they are painful:
- Processes whose rules change every month.
- Work that needs a fresh judgement every time.
- Anything whose steps nobody can write down yet.
Keep a person in control
The first version should prepare the work and let a person approve it. Once it has proved itself on real cases for a few weeks, widen what it does on its own.
What to bring to the first conversation
The process you picked, your four numbers, and a few real examples: twenty invoices, messages or forms is enough. That is all we need to say whether it is worth doing and what the first step is.