How do you calculate the ROI of an OEE system?

The question arises sooner or later in every OEE project. Often from the CFO. Sometimes from the CEO. Sometimes it is the production manager themselves who wants to be sure before the decision is made.

What do we get back on the investment and how long does it take? The question is important, as there are examples of systems that have cost more than they have returned. Perhaps because they are not used because they are too complicated to operate, or they have remained merely reporting tools without any connection to daily work.

This guide shows you how to build a credible and reasonable business case or ROI calculation. Which savings can actually be realized, which assumptions hold true and which do not, how to avoid building hot air into the calculation, and how to present the figures in a way that withstands scrutiny.

The basic principle: an OEE system pays for itself through improvement work, not through the measurement itself. The platform is the tool you pay for. The improvements are what you get back.

Where does the money come from?

The money comes from five levers: higher output from the same machinery, lower costs for downtime, less scrap and less rework, shorter changeover times and lower energy consumption per produced unit. The magnitude of these varies by industry and starting point, but the sum is often surprisingly large.

1. Higher output from the same machinery

The biggest lever in most factories. A typical OEE level of 50 to 60% means that 40 to 50% of machine time is not value-adding. Shifting OEE from 55 to 65% yields 18% more production from the same machines, the same hours, and the same staff.

For a factory that would otherwise have needed to run an extra shift, hire more operators, or invest in new equipment, this often becomes the largest single item. It becomes larger the closer to the capacity limit the factory already is.

A line produces 8,000 approved units per shift at 44% OEE. If OEE increases to 57%, with unchanged planned production time, product mix, and ideal cycle time, this corresponds to approximately 10,364 approved units per shift. The calculation is 8,000 × 57/44. That is an increase of approximately 29.5%.

If demand is unchanged, the improvement can instead reduce the required production time. Whether an entire shift can be eliminated depends on production volume, staffing, bottlenecks, and scheduling.

2. Lower costs for downtime

Downtime costs more than what is visible in the OEE report. Every hour of standstill means underutilized labor, ongoing fixed costs, energy consumed without producing anything, and delivery risks that can escalate into express freights or lost orders.

Calculating the cost of an idle machine hour is one of the most valuable exercises in a business case. It varies greatly between industries, but a rough rule of thumb is that the hourly cost of a production line is often 2 to 5 times higher than operator costs. Depreciation, facilities, energy, and management all burden the downtime.

At Sibbhultsverken, technical stops were reduced by 73% and unplanned stops by 63%. In one cell where they shortened losses, OEE improved by 19.4% in just 12 months. In another cell, OEE increased by 40%.

3. Less scrap, rework and quality losses

This item is systematically underestimated. When a product is scrapped or reworked, you have already paid for raw materials, energy, labor time, and in many cases, packaging. Scrap is not just the raw material cost; it is the entire production chain up to the point where the defect was detected.

At Barilla Wasa in Filipstad, product waste was reduced by 15% when the platform was implemented and the improvement work was launched. At Kopparbergs Bryggeri, the number of quality deviations fell by 68%. Both figures translate directly into money and provide better capacity utilization within the same machinery.

4. Shorter changeover times

Changeovers are one of the most hidden areas of loss. They are planned, they appear in the schedule, and they are often not counted as "downtime" in the same way as breakdowns. But a changeover where the machine has been idle for three hours is three hours of capacity that is not being used.

At Bostik in Helsingborg, which manufactures adhesives and sealants with over 80 products in the same facility, changeover time fell by 70% on the filling machines following systematic work according to the Lean Six Sigma methodology DMAIC (Define, Measure, Analyze, Improve, Control). OEE improved by 40%. One of the key insights was that two operators working together completed the changeover faster than two working on separate machines. The platform's measurements showed the difference.

5. Lower energy consumption per produced unit

Energy has gone from being a background cost to appearing in management reports. Higher OEE can reduce energy use per approved unit, as fewer stops and less scrap mean more approved units for the same energy. The effect needs to be verified through measurement. Add to this direct work with idle consumption and comparisons between similar machines. Cutting peak loads can also lower power costs, even if total energy use does not decrease.

A factory with 100 million euros in turnover can realistically save around 125,000 euros per year on energy by combining productivity improvements with targeted energy measures. Climate impact is reduced by 60 to 80 tons of CO2e. For many companies, this is also an item that enters sustainability reporting and strengthens the brand.

How do you build the business case?

Build the business case in six parts: define an honest current state, set quantified goals in realistic intervals, calculate annual savings distributed across the five levers, sum up the investment cost for the first year, state the ongoing cost for year two and onwards, and calculate payback period and net present value. This structure works in most presentations to management.

Current state (baseline). Define a number of key performance indicators before the project starts. OEE per line, number of downtime hours per month, scrap and rework costs, changeover time per product switch, energy cost per produced unit. This is the reference point against which the entire calculation will be measured. Without an honest current state, it is impossible to prove an improvement.

Quantified goals. State realistic intervals, not single points. Example: OEE increase of 5 to 15 percentage points over 12 months. Scrap reduction of 10 to 25%. Changeover time 30 to 50% shorter. Use results from similar factories in the same industry as a benchmark, not the most spectacular examples.

Calculated savings per year, distributed across five levers. Add them up to a total annual saving. Be consistent: if the OEE increase already includes fewer stops, do not double-count the downtime cost in the same calculation.

Investment cost for the first year. Software license, IoT hardware if needed, implementation, training, internal work by own staff. Include the internal time which should be factored into the total cost.

Ongoing cost from year two and onwards. Software subscription, support, any hardware additions. This is the cost to be weighed against the continued annual savings.

Payback period and net present value. Payback period is often the figure management looks for first. Net present value over a three-year or five-year horizon provides a more accurate picture. Use the company's normal hurdle rate.

Which assumptions hold up in the calculation?

Assumptions that hold up are gradual improvements over time: OEE increase in lower ranges in year one, scrap reduction as loss analysis matures, and changeover improvements over 6 to 12 months. Assumptions that do not hold up are quick miracle results, such as a 50% OEE improvement in six months. The most common pitfall is counting on the best-case scenario as if it were the expected case.

Assumptions that hold up:

• OEE improvement in the lower range for year one (5 to 10 percentage points), higher in years two and three once the way of working is established.

• Scrap reduction in step with the maturation of loss analysis.

• Changeover improvements that take 6 to 12 months to take effect broadly.

• Energy reduction in step with data becoming available and measures being implemented.

Assumptions that do not hold up:

• 50% OEE improvement during the first six months. This rarely happens, except from a very low starting point.

• Scrap reduction without prioritizing quality analysis. Scrap does not decrease on its own.

• Energy savings without measurement at the machine level. You cannot improve what you cannot see.

• That the entire organization suddenly starts working differently when the system is deployed. Change takes time.

Build the calculation with three scenarios: pessimistic, realistic, and optimistic. Defend the realistic one. That is where you will land in discussions with management.

What does it cost to do nothing?

Doing nothing costs in terms of lost competitiveness. While you stand still, competitors can increase their capacity. The status quo has a cost that is rarely visible on the income statement, but it is still there.

If a competitor improves OEE from 55 to 65% while you stand still, they have gained 18% more capacity from the same machinery. They can take orders you cannot meet. They can push prices you cannot compete with. They can invest in next-generation equipment while you are still arguing for this.

If your customers report under CSRD, they may request energy and climate data from you. If structured energy data is lacking, you will need to build a separate reporting track, which is both more expensive and less effective than data in the same system as production tracking.

Information security is carrying more and more weight in procurement. Therefore, choose a supplier that is ISO 27001 certified, so the system does not become an obstacle when larger customers audit you.

Factoring in the cost of doing nothing, even roughly, makes the business case more realistic. It shows that the alternative is not "saving the investment," but rather "losing competitiveness over time."

How do you avoid your case becoming hot air?

Avoid hot air through three things: anchor the figures in your own reality, assign responsibility for every figure, and measure continuously and adjust. This is what distinguishes a business case that holds up from one that falls apart.

Anchor the figures in your own reality. Use your own downtimes, scrap costs, and changeover times. Real values are more important than spectacular increases from other people's factories.

Assign responsibility for every figure. Who owns the delivery of the OEE increase? Who owns the scrap reduction? Who owns the changeover improvement? If no one is singled out, the calculation becomes a document, not a plan.

Measure continuously and adjust. The business case is not a one-time exercise. Check monthly in the first year, quarterly thereafter. Adjust targets when reality shows they were set too low or too high. This is how the improvement work, and thereby the payback, is kept alive.

How does Good Solutions work with the business case?

The platform from Good Solutions is built to drive improvement work in daily operations. That is where the money is returned. Machine connectivity provides reliable baseline data. Operator tools, dashboards, and reports make the data useful throughout the organization, from the factory floor to the management meeting. Timeline and loss analysis ensure that the right efforts are prioritized. The energy module allows cost savings and sustainability goals to be tracked in the same platform.

Operational implementation is just as important as the software. The platform is delivered with expert support from consultants with production experience, a dedicated Customer Success Manager who follows the customer over time, and a Swedish support organization. Experience from implementations in over 300 factories ensures that business cases are built realistically from the start and followed up together with the customer.

Among results realized: Kavli produced 5,000 tons more in 2024 than the previous year, without more shifts or more machines. Derome Timber increased OEE by 10% in six months, which according to the company corresponds to more than 250 extra truckloads of finished goods per year from the factory in Kinnared. This type of result is built step-by-step through improvement work in daily operations.

Read more about how others have increased their factory productivity

FAQ

What does an OEE system cost?
It depends on several factors. It can be the number of machines and users, which modules you need, and how the implementation is set up. Good Solutions works with subscriptions that include everything you need: software, IoT hardware, cloud operations, support, and updates.

How long is a typical payback period?
It depends on the starting point and how quickly the working methods get started. The lower the OEE at start, the greater the potential. The more mature the improvement work already is in the organization, the faster the platform translates into results. Build the calculation with a pessimistic, a realistic, and an optimistic scenario.

What should we measure before we start?
Establish the baseline for at least four things: OEE per line, total number of downtime hours per month, scrap and rework costs, and changeover time per product switch. For many factories, energy consumption per produced unit is also added. Without a documented current state, it is impossible to prove an improvement.

How do we factor internal time into the investment cost?
Account for the time of operators, production management, continuous improvements, maintenance, and IT during implementation. For a medium-sized factory, this often involves 100 to 300 hours internally during the implementation period. Add a realistic hourly cost. This figure should be included in the total cost.

What happens if we do not reach the goals?
The most likely reason is that the improvement work has not gotten started in daily operations. The platform then becomes an advanced report instead of an improvement tool. Continuous follow-up, preferably together with the supplier, ensures that problems are detected early and the right actions can be taken.