The usual argument for automation is time. Twenty minutes a day, five days a week, is roughly eighty hours a year. It is a fine argument and it is not the important one.
Manual work is optional work
A task a person does by hand happens when that person has time. On a quiet Tuesday it happens. On the busiest week of your year it does not — and the busiest week is exactly when following up with customers matters most.
So the real cost is not the eighty hours. It is that the work is least likely to happen precisely when it is most valuable, and nobody notices because there is no record of the thing that did not get done.
Automated work happens on your worst week. Manual work happens on your best one.
The second cost: it is invisible
When a person handles follow-ups from memory, there is no data. You cannot tell how many enquiries went unanswered, or how long the average response took, or whether responding faster would help. You are not just doing the work inefficiently — you are also losing the ability to learn anything from it.
What is usually worth automating
In our experience, the same handful of things, in roughly this order:
- Recording an enquiry. Not answering it — just making sure it exists somewhere and is assigned to someone.
- Reminders before an appointment or a stay. Cheap to set up, directly reduces no-shows.
- Review requests, sent at the right moment rather than when someone remembers.
- Internal alerts when something needs a human — a high-value enquiry, a failed payment, a broken integration.
- Reporting. Not because reports are exciting, but because a report that assembles itself is a report that actually gets read.
What is not
Conversations. Anything requiring judgement. Anything where the process itself is wrong — automating a bad process just produces bad outcomes faster and makes them harder to see.
And anything that only happens occasionally. A task done twice a year is not worth the maintenance cost of an automation nobody will remember exists.