Key takeaways
Short answer: OEE ROI is the contribution margin from extra units you produce because OEE went up, minus the cost of the platform that helped you do it. The math is simple. The credibility is in the inputs, baseline OEE measured (not assumed), conservative improvement, contribution margin per unit (not selling price), and three cases (base, upside, downside). CFOs reject single-number ROI promises and accept three-case models. See also OEE vs Utilization.
Annual ROI = (Annual extra units x Contribution margin per unit) − Platform cost
Where:
That is the entire model. Everything else is sensitivity around these inputs.
If you are building this business case to buy your first OEE platform, you do not have a measured baseline. Estimate it conservatively, most plants that have never measured run 45-65% OEE. Use 55% as a default, and note explicitly that the post-implementation actual baseline will refine the number.
If you already have OEE measurement and are upgrading, use your measured baseline from the last 90 days. Do not use a single best shift.
Realistic year-one OEE improvement from a competent rollout:
Higher numbers exist in case studies but are not credible for a CFO model in year one. Lead with the conservative number.
Annual extra units = (Target OEE − Baseline OEE) x Annual planned production time x Ideal cycle rate.
Example: a plant running 4,000 planned production hours per year, ideal rate 120 units/hour, baseline OEE 55%, target 60% (5-point improvement):
Extra units = 0.05 x 4,000 x 120 = 24,000 units/year
The number CFOs want is contribution margin per unit, not selling price. If a unit sells for €100 and costs €60 in raw material, direct labor variance, and energy, contribution margin is €40. The extra units recovered cost only the variable inputs; fixed costs (depreciation, salaried labor, facilities) are already covered by the baseline production.
24,000 extra units x €40 contribution = €960,000 annual recovery (base case).
Platform cost = license + implementation + first-year internal labor. For a mid-market OEE platform serving 5-10 lines, total year-one cost is typically €40,000-€120,000. Use the high end of the range for the model.
€960,000 − €100,000 = €860,000 net year-one return. Payback period: roughly 1.3 months.
CFOs accept three-case models because they reveal assumption sensitivity. Build:
If the downside case still has a positive year-one ROI, the deal is approvable. If it does not, the model is fragile and needs more conservative inputs or a smaller platform investment.
Three patterns get business cases killed:
A one-page model with: measured (or conservatively estimated) baseline, three improvement cases, contribution-margin math, explicit platform cost, payback in months. Bring the spreadsheet. Show the inputs. Defend the assumptions.
Fabrico's OEE module ships with a baseline-measurement workflow so the actual baseline can be locked in within 30 days of deployment, giving the CFO a real number to back the rest of the model.
See how Fabrico captures this automatically, explore OEE for manufacturing or book a demo.
Contribution margin. Selling price overstates the benefit because variable costs scale with extra units.
Use a conservative industry default (55% for discrete manufacturing) and lock in the real baseline within 30-60 days of deployment. Revise the business case once you have the real number.
5-8 points in year one is realistic. Higher case-study numbers exist but are not credible for a CFO model.
Most accept under 12 months for an operations tool. Under 6 months is common for OEE because the math compounds quickly.
Only if the platform measurably reduces maintenance hours or part spend. Many OEE-only platforms do not, keep the case clean and focus on production recovery.
Login
Blog