When Does the Automation Pay for Itself?
A quote has two numbers on it, a setup fee and a monthly charge, and the saving it promises is measured in hours. Putting those on the same footing is a four-line calculation, but it is the one that decides whether the project gets approved. Enter the quote and the hours, and read off the month it breaks even.
Use the hours saved you can defend, not the ones on the brochure. If you have not measured the manual time yet, run the manual task cost calculator first and bring the figure here.
Automation Payback Period Calculator
Setup fee, monthly fee, measured hours. Read off the break-even month.
The quoted build fee plus your own internal time to document, test and train.
Default is the Starter plan on our pricing page. Enter any quote.
Measured manual hours, less the exception share and any review time.
Roughly 1.3x base pay for an Australian employee.
Hours saved at your hourly cost, minus the subscription. Negative means it never pays back on labour alone.
Labour hours only. Errors avoided, faster turnaround and capacity gains are not priced here. Net return figures can be negative when the setup cost has not yet been recovered.
How to Read the Result
Payback answers one question: how many months of net saving does it take to recover the setup cost. The 12 and 36 month figures show what is left after that, which is what a finance person will ask about next.
Monthly net saving
Hours saved multiplied by the hourly cost of the people who were doing the work, minus the monthly subscription. If this number is zero or negative, the automation never pays back on labour alone, and the case has to rest on something else: fewer errors, faster turnaround, or capacity you cannot hire.
Payback period
Setup cost divided by monthly net saving, rounded up to whole months. Under six months is a decision most owners make quickly. Six to eighteen months is normal for a build with real integration work. Beyond that, question the hours saved or look for a smaller first scope.
Net return at 12 and 36 months
Cumulative net saving minus the setup cost at each point. The 12 month figure is what the first year actually delivers. The 36 month figure matters because most automations run far longer than a year once they are stable, and the setup cost is paid only once.
The Formulas
Four lines, nothing hidden. Every figure on the page can be reproduced on paper from the inputs you entered.
Gross monthly saving
Monthly hours saved multiplied by the loaded hourly cost of the people who were doing the work.
Net monthly saving
Gross monthly saving minus the monthly subscription. This is the amount available each month to recover the setup cost.
Payback in months
One-off setup cost divided by the net monthly saving, rounded up. If the net saving is zero or negative the calculator shows 120, meaning it does not pay back within ten years.
Net return
Net monthly saving multiplied by 12 or 36, minus the setup cost. A negative 12 month figure with a positive 36 month figure means the project pays back in year two.
The Inputs People Get Wrong
The arithmetic is simple. The inputs are where a payback calculation becomes optimistic, and each one has a predictable direction of error.
Hours saved is a measurement, not an estimate
This is the input that swings the result most and the one most often taken from a vendor slide. Hours saved should come from timing the manual process, multiplied by real monthly volume, then reduced by the share the automation cannot handle. An automation that handles ninety per cent of cases saves ninety per cent of the hours, not all of them.
- Time the task with a stopwatch across several real instances
- Multiply by the actual monthly count from the source system
- Subtract the exception share that still needs a person
- Add back any new time spent reviewing the automation’s output
Hourly cost should be loaded, and it should be the right person
Base salary divided by hours understates the cost. Superannuation, leave, payroll tax and the overhead of a desk push the loaded cost to roughly 1.3 times base for an Australian employee. Also check who actually does the work. Manual tasks drift toward the most reliable person, who is often the most expensive one.
- Start from base hourly pay and multiply by about 1.3 for on-costs
- Use a contractor’s charge-out rate directly if that is who does it
- If the hours are spread across people, use a weighted average
- Do not use an owner’s notional rate unless the owner does the task
Setup cost includes your own time
The quote is the visible part. Internal time to document the process, clean the data, test the build and train the team is real cost that lands before any saving arrives. Adding it to the setup figure makes the payback honest. Leaving it out is how a six-month payback becomes nine in practice.
- Add internal hours for documentation, testing and training at loaded cost
- Include any licence upgrade needed for API access
- Include data clean-up if the automation depends on it
- A shadow-run period delays the saving by that many weeks
The subscription is not the only ongoing cost
Someone reviews the exception queue. Someone updates the workflow when a supplier changes their invoice layout or a system is upgraded. Those hours are small, but they are monthly, and they belong in the net saving. A realistic approach is to reduce hours saved by the review time rather than trying to price it separately.
- Budget a fixed weekly review slot and subtract it from hours saved
- Expect occasional maintenance when connected systems change
- Check whether the subscription price steps up with volume
- Ask whether support is included or billed by the hour
Next Steps
Manual Task Cost Calculator
Measure the hours and cost of the manual task before you enter them here.
Measure the task →How to Calculate Automation ROI
The longer guide to building a return case that survives a finance review.
Read the guide →AI Automation Cost in Australia
What setup and monthly figures typically look like, so you can sanity-check a quote.
See typical costs →Frequently Asked Questions
What is a good payback period for business automation?
There is no universal threshold, but the way owners actually decide is fairly consistent. Under six months rarely needs a business case beyond the number itself. Six to eighteen months is normal for a build that integrates several systems and is usually approved when the hours saved are well evidenced. Beyond eighteen months, the honest move is to reduce the first scope to the part of the process with the clearest saving, get that paid back, and extend from there.
Why does the calculator show 120 months?
Because the net monthly saving is zero or negative: the subscription costs as much as or more than the labour it saves. The calculator caps the display at 120 months rather than showing an infinite or negative figure. If you see 120, the project does not pay back on labour within ten years. That does not always mean it is a bad idea, but it means the case has to be made on error reduction, speed, compliance or capacity, none of which this calculator prices.
Should I include the hours saved by the automation running out of hours?
Only if those hours were previously being worked, or would have to be. If invoices currently wait until morning and the automation processes them overnight, the saving is turnaround time, not labour, and it should not be counted as hours saved. If the alternative was paying someone overtime or hiring a second person to cover the volume, then the avoided hours are real and belong in the figure.
What if the hours saved will grow as the automation is tuned?
Enter the hours you expect in month one and run the calculator again with the figure you expect once the exception rate has settled. The two results give you a range for payback rather than a single optimistic point. Most deployments handle a smaller share of cases at first while supplier formats, coding rules or routing logic are refined, so the first-month figure is the one to use for the approval decision.
Does this calculator account for errors avoided?
No. It prices labour hours only, because that is the figure you can measure with a stopwatch and defend without argument. Errors avoided are often a larger saving than labour, particularly for invoice processing and data entry, but they need their own inputs. The data entry error cost calculator and the invoice processing cost calculator on this site both put a dollar figure on errors that you can then add to the hours saved here, converted at your hourly cost.
Can I use this for a subscription with no setup cost?
Yes. Set the setup cost to zero and the payback period shows zero months, meaning the automation is cash-positive from the first month as long as the net monthly saving is positive. The 12 and 36 month figures then show the cumulative net return. This is the common case for off-the-shelf tools with no integration work, and the useful comparison is the net monthly saving against alternatives, rather than payback.
Want a Real Quote for the Setup Cost Field?
Tell us the process and the hours you measured. We will come back with a setup and monthly figure you can put straight into this calculator, and we will say plainly if the payback looks too long to bother.