"How much time will this save?" is the wrong first question. The right one is "what does that time actually cost, and what else changes when it’s gone?" Here’s the framework we walk clients through.

Start with fully-loaded cost, not just hours

Multiplying hours saved by an hourly wage undercounts the real cost. A fully-loaded employee cost includes benefits, overhead, and management time — often 1.25–1.4x base salary. Use that number, not the number on a pay stub, or you’ll consistently underestimate your own ROI.

Time saved isn’t the only return

The obvious metric — hours reclaimed — is usually the smallest part of the real return. The bigger numbers tend to be:

  • Error reduction — a manual data-entry error that causes a shipping mistake or a compliance issue often costs far more than the minutes saved on entry
  • Speed-to-revenue — faster lead response or faster invoice-to-cash directly affects when money actually arrives, not just how efficiently someone’s day is spent
  • Capacity without headcount — the ability to handle 2x the volume without hiring is often worth more than the time savings on current volume
If your ROI case rests entirely on "time saved," you’re almost certainly leaving the bigger numbers out of the calculation.

The formula we actually use

Annual hours lost = employees × hours wasted per week × 52. Annual cost = hours lost × fully-loaded hourly rate. Recovered value = annual cost × expected automation coverage (rarely 100% — we typically model 60–75% for a first workflow, since some manual judgment usually still belongs in the loop).

Then add error-reduction and revenue-speed impact as a separate line, estimated conservatively — don’t combine it with time savings into one fuzzy number, or you lose the ability to sanity-check either half.

What to be skeptical of

Be wary of any ROI projection (including ours) that assumes 100% of manual time disappears, or that doesn’t account for the ongoing cost of monitoring and maintaining the workflow once it’s live. A workflow that saves 20 hours a week but needs 3 hours of babysitting isn’t a 20-hour win — it’s a 17-hour one, and that’s the number worth building a business case around.