"This would save us so much time" is the most common justification for building an automation, and it's rarely backed by an actual number. It's said with genuine conviction, which is exactly why it goes unchallenged so often. A rough but honest calculation, done before any building starts, prevents both the mistake of skipping a genuinely valuable automation and the equally common mistake of investing heavily in one that never pays for itself, and it costs nothing more than a spreadsheet and a few honest assumptions.
The basic formula
Time saved per occurrence, multiplied by how often the process runs per month, multiplied by a reasonable hourly cost of the time being freed up, gives a monthly value. Compared against the estimated cost to build and, importantly, maintain the automation, that comparison produces a genuine payback period rather than a vague feeling that automating something must be worthwhile.
The maintenance cost that gets forgotten
Build cost is usually estimated reasonably accurately, since it's the visible, upfront number. Ongoing maintenance, monitoring, occasional fixes when a connected tool changes its interface, periodic review as the underlying process evolves, is the cost that gets consistently underestimated or ignored entirely, and it's often what actually determines whether an automation pays for itself over its real lifetime.
Time saved isn't always straightforwardly convertible to money
Time freed up only translates directly to value if it gets reallocated to something that actually matters, more sales calls, more strategic work, faster response times that measurably improve conversion. Time saved that just becomes slack in someone's day, with no clear reallocation, still has real value, reduced stress, lower error rates from less rushed work, but it's harder to put a precise number on and worth being honest about that distinction rather than overstating the ROI.
A worked example
A task taking fifteen minutes, done twenty times a month, at an effective hourly cost of thirty euros, represents roughly one hundred and fifty euros of monthly time value. An automation costing eight hundred euros to build, with minimal ongoing maintenance, pays for itself in a little over five months, a reasonable timeline for most business automation investments. The same automation costing four thousand euros to build would need a genuinely higher-value or higher-frequency process to justify the same investment within a comparable timeframe.
When the ROI calculation says no, and that's the right answer
Not every process clears this bar, and that's a legitimate, useful outcome of doing the calculation rather than a failure of the exercise. A process that's infrequent, low time-cost per occurrence, or would require expensive custom development to automate reliably sometimes genuinely isn't worth automating yet, and knowing that clearly is more valuable than automating it anyway on vague enthusiasm and discovering the payback never actually arrives. This kind of honest calculation is exactly what should precede any serious automation investment, not follow it as an afterthought. It takes minutes to run and often prevents months of quiet disappointment with an automation that never quite pays for itself, or the opposite mistake of never building one that clearly would have paid off many times over within its first year.
Revisiting the calculation as circumstances change
A process that doesn't clear the bar today might clear it later, if its frequency increases, if labour costs rise, or if a cheaper automation platform becomes available. Keeping a short list of previously rejected automation ideas, revisited every six months or so, catches these shifts without requiring anyone to remember to reconsider them unprompted.