Automation Business Case Calculator

The question this calculator answers, and the one it cannot

The calculator answers one question. If this repeated task is automated, on the figures you believe today, does the saving cover the cost, and in which month? It does not tell you whether to automate. It tells you whether the arithmetic supports the decision you are leaning toward, and which of your own assumptions that support rests on.

Automation cases tend to fail in the spreadsheet before they fail in production. The saving is booked from the first month, saved minutes are treated as cash, and the build ends on the planned day. The calculator removes each habit.

Who gathers the figures, and how long a sitting takes

Entering the figures takes about as long as a coffee once you have them. Getting them is the real work, and it is shared. The operations lead who owns the task knows how often it runs and how long a run takes, checking included. The finance partner holds the loaded cost of an hour and decides what counts as released. Whoever scoped the build owns the one-off cost, the running cost and the go-live date.

A first pass alone is fine, to see whether the case deserves anyone else's afternoon. A result you will put in front of a finance committee should be run with those people, each owning the lines they know.

What is counted, and what is left out

Each month, the benefit is the number of runs, times the minutes in a run, times the share the automation does alone, times the share you count as money, times the cost of an hour, times that month's level of use. The one-off cost lands at the start. The running cost is paid from the go-live month, whether or not the benefit has caught up.

Left out are errors avoided, faster turnaround, better service, lower risk, inflation, discounting, tax and financing. Left out too are exceptions that return to people beyond the share you entered, and time lost while the system is repaired. Volume, minutes and costs stay level, so a growing task pays back sooner than shown and a shrinking one later.

Why does the ramp move the payback month?

People rarely switch on the day a system goes live. Some run both routes for a while, and some wait for the first fault before they trust it. The calculator reads this as a ramp: use rises in equal steps from the go-live month, while the running cost is paid from the first live month. A ramp costs twice, because the benefit arrives later and the spending starts early. Enter the ramp you would plan for if the sponsor were not watching.

Why does the worst case move volume and cost together?

Projects that miss on benefit rarely miss alone. A smaller volume often means the task was less uniform than assumed, and a less uniform task takes more build and more upkeep, so the worst case lowers volume and raises both costs in the same pass. The best case raises volume only, since a project that runs ahead seldom gets cheaper to build. Both spreads are your own judgment. A wide one says the inputs are soft, and the response is to firm them up.

Reading net benefit, the payback month and the range

Read net benefit with the payback month, which counts from the start of the project and so includes the build. If the cost is not earned back inside the months counted, the result says so in words and does not stretch the period. Then read the range. If the worst case is still positive, the case does not hang on your optimism. If only the expected case is positive, talk about what would make the worst case less likely. The sensitivity table shows which input shifts net benefit most. That input is where the next conversation belongs.

What to do the day after you calculate

1. 01

   Take the shakiest input to its owner

   The row of the sensitivity table you trust least goes to the person who owns that figure, with a request for evidence from real work.

2. 02

   Ask finance which spending it releases

   Name the hire, the overtime or the outside invoice that falls away. If none can be named, lower the share counted as money and run the case again.

3. 03

   Open the meeting with the worst case

   The sponsor then decides against the downside and not against the hope.

4. 04

   Write down what was decided

   Going ahead, narrowing the scope and leaving the task alone are all outcomes. Record the figures and the date, so the next run can be compared with this one.