Quick answer: An OEE software investment is justified when three things are simultaneously true: your bottleneck line has at least 5 percentage points of recoverable OEE, your throughput-hour is worth at least €800 in average selling price, and you have at least one full-time operator or maintenance lead willing to own the rollout. If any of the three is missing, the software does not save the day, it just adds a line to the OPEX bill while the team waits for one of the three to change.
Key takeaways:
A real CFO-grade ROI model for OEE software has seven lines. Anyone showing you a sales spreadsheet with three lines is missing something material.
Inputs (the things you have to know honestly):
The maths the CFO actually wants to see:
That is the model. It survives CFO scrutiny because it uses margin contribution (not full selling price), real annual hours (not theoretical 8,760), and Year-1 true cost (not headline license). Vendors that show you a spreadsheet using ASP × full year × headline cost are not modeling, they are selling.
Related deep-dives: OEE software selection process · OEE software hidden-cost checklist · OEE + CMMS as profit center · Breaking the OEE plateau.
Three honest scenarios where OEE software is the wrong next purchase, plus the work to do first.
Scenario 1, You do not have a named bottleneck line. If the plant director cannot name "Line 3 packer" or "CNC cell 7" in one sentence, OEE software will produce four lines of mediocre data instead of one line of actionable data. The prep work: spend a week with the production schedulers and identify the line that constrains throughput the most days per month. That line gets the pilot. Once it is named, OEE software becomes the right purchase.
Scenario 2, Your throughput-hour is worth less than €500. Below that, the maths in Slot 1 stops working. Either the units are too low-value or the line runs too slow. The prep work in that case is not OEE software, it is a manufacturing engineering review of the line itself. Sometimes a simple line-balancing exercise produces more value than any digital tool.
Scenario 3, No internal owner. The Plant Director is busy, the maintenance lead is part-time, and the operators are stretched. OEE software shipped into that environment becomes a dashboard nobody opens. Prep work: appoint a single full-time owner for 90 days. If you cannot find one, the OEE programme is not real and the software will not change that.
The CFO question is rarely "should we buy OEE software". It is "should we buy it now or in 90 days". The three scenarios above almost always answer that question. Spend the 90 days doing the prep work, then revisit. Most plants that wait 90 days end up with a sharper pilot and a faster payback than the ones that rushed in.
What payback period should a CFO expect? 3-9 months on a real pilot if the three thresholds are met. Anything past 18 months means the inputs are wrong or the bottleneck is not real.
Does the calculation change for multi-plant rollouts? Yes, but pilot one plant first. Get the Year-1 numbers from the pilot, then use those numbers (not vendor projections) to model the rollout. The payback period typically extends from 4-6 months on the pilot to 8-12 months on the rollout because of integration overhead at scale.
What about industries with thin margins? Use margin contribution, not gross margin. Even a 8% margin business protects real money when OEE moves 5 points, the per-unit number is small but the volume is large.
Should I include the avoided-capex argument? Yes, but separately. The model above produces protected revenue. Avoided capex (deferred replacement of a worn-out asset thanks to better PM cadence) is real but lives in a separate line for board purposes. See our board-presentation guide for the framing.
What ROI multiple should we see? Year-1 net contribution is typically 5-15× the software cost on the pilot line. Year-3 cumulative is typically 15-40×. Anything outside those ranges is suspect.
The investment is justified when three thresholds, recoverable OEE, throughput-hour value, internal owner, are simultaneously true. If they are, the ROI model in Slot 1 produces a CFO-grade case with a 3-9-month payback. If any threshold is weak, do the prep work first; the software does not save the day. The single biggest mistake we see is buying the tool before the bottleneck line is named. Don't.
Want the ROI spreadsheet pre-filled with your plant's numbers? Fabrico runs it as a 25-minute working session, bring two inputs (bottleneck throughput per hour, ASP), and we send you back a CFO-ready one-pager. Book the session.
Three numbers gate the decision. Get them on a one-page memo before the next vendor call.
Threshold 1, Recoverable OEE on the bottleneck line ≥ 5 points. If your current OEE is 78% and physical max is 82%, software does not produce a meaningful number. Recoverable points have to exist before software helps surface them. Quick way to check: are the operators saying "the line can go faster but X keeps getting in the way"? If yes, there are usually 6-10 recoverable points.
Threshold 2, Throughput-hour value ≥ €800 at margin contribution. Below that, the absolute euro figure from a 5-point OEE recovery does not justify a software programme. Above that, it does. The maths is brutal but clean. Run it before any vendor conversation. The number sets your budget tier (under €800/hour = no software for now; €800-3,000/hour = mid-market pilot; over €3,000/hour = enterprise rollout).
Threshold 3, Named full-time internal owner for the first 90 days. Plant Director can sponsor; cannot operate. The owner is usually a senior maintenance lead, a continuous-improvement engineer, or a shift supervisor pulled out for the rollout. Without one, the dashboard goes unloved and the data goes unused. If you cannot identify the person in a sentence, you do not have an owner.
If all three thresholds are simultaneously met, the investment is justified and the model in Slot 1 produces the case. If one is weak, the prep work in Slot 2 fixes it in 90 days. If two or more are weak, the conversation is not yet about OEE software, it is about the plant's readiness to digitize at all.
Login
Blog
