What one automated process is actually worth

Automation ROI gets dressed up in a lot of nonsense — "10x productivity", "digital transformation", percentages with no source. You don't need any of it. For a repetitive process, the value comes from exactly three places, and you can put a number on each one yourself.

1. The hours, priced honestly

Take the hours the process eats per week, multiply by the fully-loaded hourly cost of whoever does it (salary plus taxes and overhead — not just the wage), multiply by 52. A process that takes 8 hours a week at a true cost of €15/hour is €6,240 a year. That's the number a build has to beat — and notice it's an ongoing saving. Compare it with what the automation costs you per month or per year, not just what it costs upfront.

2. The errors that stop happening

Manual data entry produces errors even under careful conditions — around 1% of entries in controlled studies, and more in messy real-world workflows. Errors have real prices: a mispaid invoice, a lost order, an hour of hunting down why two systems disagree. Estimate how many errors the process produces a year and what the average one costs to fix. For document-heavy processes this number is often bigger than the hours number — it's also the one everyone forgets to count.

Source for the error-rate claim: Barchard & Pace (2011), Computers in Human Behavior — single-entry data entry error rates of roughly 1% in a controlled study.

3. The things that stop falling through

Some process failures don't cost time — they cost the deal. The quote request that sat unread for four days. The follow-up that never went out. This value is real but hard to number honestly, so treat it as the tiebreaker rather than the headline: if two processes score similarly on hours and errors, automate the one where delay loses you customers.

What to subtract

Honest ROI subtracts the full cost: the build itself, any tool subscriptions it introduces, and maintenance — automations break when the software underneath them changes, and someone has to fix them. If a vendor's ROI story doesn't mention maintenance, it isn't an ROI story, it's a brochure.

When the answer is "don't bother"

If the process is rare, cheap, and its errors are harmless, automation is a hobby, not an investment. We say this on discovery calls regularly. The point of doing the maths first is that it makes the decision boring — which is exactly what a good investment decision should be.

This calculation is what the number in our 30-day check is built on: before we build, we agree the hours-saved target in writing, and if the automation misses it in its first 30 days live, you walk away and owe nothing. ROI you don't have to take on faith.