When a business decides to automate, it almost always aims at the biggest process: “let’s automate the whole sales area.” It is the fastest way to spend months and end up with a half-finished project.
It pays to start with something small that shows. A process that, once it stops being done by hand, frees up hours this very week.
The four questions
Before choosing, run every candidate through these questions:
- Does it repeat? If it happens ten times a day, every minute saved multiplies. If it happens once a month, it may not be worth it.
- Does it have clear rules? “If the order is above a certain amount, the manager approves it” is a rule. “It depends on the customer” is not one yet.
- Does the data already exist somewhere? In an email, a form, a spreadsheet. If it has to be made up every time, that needs sorting out first.
- Would anyone miss it? If someone feels that process is “their job”, talk to them first. They are often the one who knows the exceptions best.
The best first candidate answers “yes, yes, yes and no”.
Examples that usually work
- Orders arriving by email or WhatsApp that someone copies into a spreadsheet or the system.
- Reminders that depend on someone remembering: a payment about to fall due, tomorrow’s appointment, a missing document.
- The Monday report that someone builds by putting three files together.
They have something in common: they are tasks nobody chose to do, but they take up part of the day, every day.
What not to automate yet
A process that changes every week, or where every case is different, is not ready. Automating a mess only makes it faster. First you agree on the rules, even on a sheet of paper, and then you automate.
How to know it worked
Before you start, write down how long it takes today and how often someone gets it wrong. A month later, measure the same thing. If there is no difference, the problem was somewhere else, and that is useful information too.
If you are not sure which of your processes passes the test, write to us and we will go through it together.
