Free · nothing to sign up for
What is that manual task really costing you?
Everyone knows the weekly job that eats an afternoon. Almost nobody has multiplied it out. Put your numbers in and you will have a figure you can take to a partner, and a rough idea of whether fixing it is worth anyone’s money.
Cost of leaving it alone
$12,480
That is what this task costs you every year at $42 an hour fully loaded, across 50 working weeks.
Recoverable
$8,736
per year
Hours back
210
per year
How fast it pays back
Most builds of this kind land somewhere in these ranges. Find your recoverable figure above and read across.
| If the build costs | It pays back in |
|---|---|
| $4,000 | 5 months |
| $8,000 | 11 months |
| $15,000 | 21 months |
| $30,000 | 41 months |
My rule of thumb: under twelve months is worth doing, twelve to twenty-four is worth a conversation, and beyond that the problem is probably not the right one to start with.
How the maths works
- Why multiply the hourly rate by 1.3?
- Because an employee costs more than their wage. Payroll taxes, benefits, equipment and the share of overhead they carry usually add 25 to 40 percent. Using the raw wage understates what the task really costs you, often badly.
- Why 50 weeks rather than 52?
- Holidays and slow weeks. It is a small adjustment that keeps the number defensible if you take it to a partner or an accountant.
- Is 70 percent realistic?
- For a well-chosen task, yes. Something almost always still needs a person: the odd exception, the judgement call, the customer who has to be phoned. Anyone promising 100 percent is selling you something.
- Where do the build cost ranges come from?
- Typical scope for this kind of work. A single workflow connecting two systems sits at the low end. Something reading messy documents, touching money, or spanning four systems sits at the high end. Your actual number needs a conversation, not a calculator.
The number is bigger than people expect
Six hours a week feels like nothing. It is roughly seven and a half working weeks a year, for one person. That is why the tasks worth automating are almost never the dramatic ones. They are the boring weekly jobs nobody thinks to add up, which is exactly why they survive for years.
If your figure came out large, the useful next step is not a proposal. It is working out whether that specific task is a good candidate, because some are and plenty are not. That is a short conversation about cost and where this actually helps.
Last reviewed .