The bars are rarely close
For most mid-market finance functions the manual-time bar is four to ten times the licence bar. Software budgets are argued over; the larger number is not in the conversation at all.
Free tool
Every ERP business case compares a new system's price against a licence renewal. That is the wrong comparison, because the largest cost of the current arrangement is not the software — it is the people spending half their week doing what software could do.
A health check measures the manual-time figure from your actual systems rather than from a slider.
Move the sliders. The manual-processing figure is the one worth arguing about, and it is derived from a loaded cost of $96,000 per finance person.
How to read it
The point of the calculator is not the total. It is the ratio between the two bars, which is usually the opposite of what a software budget conversation assumes.
For most mid-market finance functions the manual-time bar is four to ten times the licence bar. Software budgets are argued over; the larger number is not in the conversation at all.
Each system you add increases licences and — because every system boundary is a reconciliation — often increases manual time too. The two do not trade off cleanly.
Most teams estimate their manual share at 25 to 35% and measure it at 45 to 60%. It is the input with the widest gap between belief and reality.
A vendor cannot invoice for your team’s time, so including it makes their number look worse against a competitor who omits it. Every proposal therefore compares licence to licence, and the largest cost on both sides stays invisible.
That omission has a consequence beyond the arithmetic: it frames the decision as a purchase rather than as a reallocation. The question is not whether the new system costs more than the old one. It is whether the combined figure — software plus the people doing what software could do — goes down.
Manual processing means: keying or coding transactions, matching, chasing approvals, reconciling between systems, rebuilding the same report, and assembling the consolidation. It does not mean analysis, review, judgement calls, or the conversations that follow them.
The distinction matters because the second category is what you are paying a finance team for and the first is what automation removes. Teams that measure this properly — timing a week rather than estimating one — consistently find the manual share higher than they assumed, and the gap is largest in AP and reconciliation.
It will not tell you how much of the manual share is actually automatable. That varies by workflow from 94% down to 66% in our own published benchmarks, and it depends on your vendor concentration, document quality, and policy clarity rather than on the software.
A realistic business case takes the manual figure, applies a per-workflow automation rate, and assumes the recovered hours go into analysis rather than out of the door. Treating them as headcount reduction is a different decision and one most teams do not actually make.
Loaded cost per finance person is fixed at $96,000 — a US mid-market blend of salary, employer taxes, and benefits. If your loaded cost differs materially, scale the manual bar proportionally.
Software costs are per-system monthly figures scaled by revenue, drawn from list pricing and from the 63 quotes in our pricing study. They are typical rather than specific to any vendor, and your negotiated rates will differ.
The revenue slider scales software cost sublinearly, because most finance tools price on volume bands rather than continuously. Below $18M and above $95M the scaling is clamped.
Nothing here is captured, stored, or sent anywhere. The calculator runs entirely in your browser and there is no email gate on the result.
Questions
The question is whether software plus people goes down, not whether the software line goes up.