Cash That Stops Is Counted In Full.
External spend and wages that genuinely stop leaving the business convert to profit, less the running cost of the build. That net figure is the only one used for payback.
Test whether an automation build pays back, on your own figures. You get two answers side by side: the cash payback, and the time the build returns. Nothing is pre-filled from benchmarks, and the two are never added together.
Echelon AI Solutions built this ROI calculator so you can test whether an automation pays back on your own figures. Enter the hours a task takes, what that time costs, any spend the build would stop, and the build and running costs from a quote, and it shows payback and return with nothing assumed for you.
External spend and wages that genuinely stop leaving the business convert to profit, less the running cost of the build. That net figure is the only one used for payback.
Hours handed back to you or your team have real value, and the calculator shows it at the rates you enter. They become cash only if they replace spend or go into work that earns, so they sit next to the payback figure and never inside it.
The build cost is paid once. Savings and running costs recur every year. Year one carries the build cost; the years after it do not.
There are no industry averages, benchmark hours or default hourly rates. The one pre-filled figure is 52 weeks, because that is how many weeks a year has.
Time one real run of the task, including the checking, chasing and copying between systems that usually goes uncounted. Then count how often it runs in a normal week.
Use payroll, not a guess: the hourly wage plus super plus the other costs of employing that person. The wage on its own undercounts the hour.
Pull the last 12 months of bills for the contractors, services or software the build would replace. If nothing would stop being paid, leave it at zero.
Take them from the quote, including every subscription and usage charge the build depends on after it goes live.
Only 39% of respondents to McKinsey’s November 2025 State of AI survey attribute any EBIT impact to AI, and most of those put it below 5%. Testing payback on your own figures before a build is the first check against that. Source: McKinsey & Company, The State of AI, November 2025 (global survey)
The hours one run of the task takes and how often it runs each week, your team’s loaded hourly cost from payroll, any contractor or software spend the build would stop, and the build and running costs from a quote. Echelon’s calculator assumes none of these for you, apart from the 52 weeks in a year.
Because returned hours only become cash if they replace spend or go into work that earns. The calculator counts cash that stops leaving the business in full and shows time returned beside it, so you see both without the return being overstated. Echelon uses the same rule when it builds a roadmap.
Quotes that go out the same day and win the work, jobs that stop falling through and a month-end that closes on time. Those can be worth more than the hours, but they’re hard to price before a build runs, so Echelon measures them against a baseline once the system is live.
From a real quote, including every subscription and usage charge the build depends on after it goes live. Echelon quotes once the scope is clear, because the cost depends on the systems involved and the state of the data. Use the quote’s own figures in the calculator, not a guess.
Yes. Enter the hours your team spends answering, booking and returning calls each week, their loaded hourly cost, and any answering service or after-hours cover the receptionist would replace as cash that stops. Jobs won from calls you miss today aren’t counted, because they’re hard to price before it runs. Echelon measures those against a baseline once the receptionist is live.
Quotes that go out the same day and win the work, jobs that stop falling through, a month-end that closes on time. Those can be worth more than the hours, and they are hard to price before a build runs. We measure them against a baseline once it does.
Our diagnostic produces these figures from your own operation: recurring work timed across every role in scope, and a ranked list of what is worth building.
AI will change how the best businesses operate. The question is whether that change is designed or inherited. We partner with leadership teams choosing to design it.