Why is OEE not enough as your only KPI?

OEE is not sufficient as a sole KPI because the figure shows that production is losing efficiency, but not why. It says nothing about the reason behind the losses, it can also be improved in ways that harm the business, and it cannot be compared fairly between factories.

To create value, OEE needs to be supplemented with loss analysis, quality metrics, changeover times, energy data, and flow metrics. Above all, it needs to be used as a basis for improvement work, not as a grade on a screen.

This guide explains why OEE is not sufficient as a sole KPI.

We look at the most common misconception about what actually stops production, three concrete reasons why OEE becomes misleading if it stands alone, what the metric needs to be supplemented with, and what is required for OEE to create real value.

If you want the basics first, please read the article “What is OEE and how is it calculated?“.

Is production stoppage primarily a machine problem?

No. Data from factories that have worked structured with OEE for many years shows that approximately 80% of the stoppage causes are not directly machine-related. They are about material shortages, changeovers, waiting for orders, quality issues, planning errors, communication between shifts, and lack of operators. A common assumption is that stoppages are mainly due to machine breakdowns, and that the solution is better maintenance and smarter machines. The reality is therefore different.

This has a direct bearing on how OEE should be interpreted. A low OEE figure says nothing in itself about where the problem lies. If 80% of the losses are not in the machines, a factory that only invests in better machines and more maintenance will be disappointed with the result. The great potential lies elsewhere, in working methods, planning, material flows, and in the operators' knowledge of what is actually happening on the floor.

It also has a bearing on how data should be collected. Automatic collection from machines can show that a stoppage has occurred. It can rarely show why. Operators need to be able to code the cause, quickly and easily, at the very moment the stoppage occurs. Otherwise, the largest part of the losses ends up in the "Other" category and becomes impossible to work with. At Orkla Nidar, a candy manufacturer in Norway since 1912, operators were able to eliminate the "Other" category from reporting as soon as they were given tools that made it easy to code the correct cause directly.

What are the reasons why OEE is not sufficient as a sole KPI?

OEE is not enough on its own for three reasons. It says nothing about why the losses occur, it encourages wrong behaviors when it becomes the only metric, and it cannot be compared between different factories. We go through them in turn.

1. It says nothing about why

An OEE of 62% does not say whether the problem is due to breakdowns, changeovers, quality, or staffing. Without loss analysis, OEE becomes just a thermometer showing fever without a diagnosis. You know something is wrong, but not what, and therefore not what to do about it.

The value only arises when the figure can be broken down. By stoppage cause, by cause group, by line, by article, by shift, by machine part, and as a trend over time. Then it is possible to answer the questions that actually lead to action. Why does article A perform worse on line 1 than on line 2? Which stoppage causes have increased over the past month? Which losses are frequent but short? Which are rare but expensive? It is in the breakdown, not in the total figure, that the improvement work gets its direction.

2. It encourages wrong behaviors

If OEE becomes the only metric, factories start optimizing the figure instead of the business. This rarely happens on purpose, but it happens.

A common example is deducting time that "should not be counted" until the figure looks good without anything actually being improved. Another is running machines that really should not be run, just to keep availability up. A third is reducing the number of changeovers by running longer series. This raises OEE in the short term but lowers flexibility, increases inventory, and binds capital. The factory looks more efficient on paper while becoming less competitive in practice.

When a metric becomes a target, it often ceases to be a good metric. OEE is no exception. Therefore, it needs to be balanced against metrics for flow, delivery precision, inventory, and quality.

3. It cannot be compared between factories

OEE values depend on how "planned production time", "ideal cycle time", and stoppage causes have been defined. Two factories that calculate differently can have the same actual efficiency but completely different OEE figures, or the same figure despite large differences in actual performance.

Comparing factories' OEE values against each other is therefore almost always misleading. It can also damage the improvement culture. When factories or shifts are compared on a single figure, incentives arise to make the figure look good, rather than to improve what lies behind it. Competition over the number replaces collaboration on improvements. For a group with multiple facilities, this is an important insight. Follow-up at the group level should focus on development over time within each unit, and on sharing learnings, not on ranking units against each other.

What should OEE be supplemented with?

OEE should be supplemented with loss analysis per cause, quality metrics, changeover times, energy consumption per unit, and flow metrics. OEE should not be abandoned, but supplemented, so that the picture becomes complete and difficult to manipulate.

Loss analysis per cause. The very foundation for OEE to become actionable. Without it, OEE is a figure without an address.

Quality metrics. Scrap, rework, and quality deviations, preferably linked to article and cause. Quality is not just a part of OEE, but a separate dimension worth tracking.

Changeover times. One of the most hidden areas of loss, and often one of the greatest levers for increased capacity.

Energy consumption per unit. An increasingly important dimension, both for costs and for sustainability reporting. Energy per produced unit decreases as OEE rises, which means the metrics are closely linked.

Flow metrics. Lead time, work in progress, and delivery precision. They capture what OEE misses, especially the risk of a high OEE being achieved at the expense of flow.

Together, they provide a picture that is hard to cheat with. That is the point. When multiple metrics point in the same direction, the development can be trusted.

What is required for OEE to create value?

Five things determine if OEE creates value: data quality, operator involvement, daily management, in-depth loss analysis, and a clear implementation process. Experience from 300+ factories shows a clear pattern. Technology is rarely the bottleneck.

Data quality. If operators do not trust the figure, they will not use it in their improvement work. This requires consistent definition of planned production time, correct handling of changeovers and planned stops, and clear stoppage cause codes that operators understand.

Operator involvement. Since 80% of stoppage causes are not machine-related, operators must be able to code them. This requires simple interfaces on the shop floor, not reports to be filled out the next day.

Daily management. OEE only becomes valuable when we use it in daily pulse meetings, shift handovers, and improvement meetings. A figure in a monthly report changes nothing.

In-depth loss analysis. Being able to break down OEE per line, article, shift, stoppage cause, and time is the difference between reporting and acting. Sibbhultsverken increased OEE by 19.4% and reduced technical stoppages by 73% by working systematically over a year, not through a single effort.

Clear implementation process. Implementing an OEE system is not just about installing software. It is about defining what is to be measured, ensuring that signals from machines and operators are reliable, and building working methods that actually use the data. Without that, the system becomes another source of information that no one looks at.

How does OEE go from metric to improvement work?

OEE goes from metric to improvement work when the figure is used in a recurring cycle of measurement, analysis, prioritization, action, and follow-up. The common thread in all of this is that OEE is a tool, not a goal. The goal is to reduce production losses and costs, and to increase factory productivity. OEE is one of the best tools we have to see where the losses are, but the figure itself changes nothing.

That is why OEE belongs in a context of continuous improvements, what in lean is called continuous improvement or Kaizen. Measure, analyze, prioritize, act, follow up, and start over. OEE and loss analysis provide the factual basis. The improvement work provides the results. A factory that measures OEE but does not work structured with improvements gets a nice dashboard without the car moving forward.

How does Good Solutions work with this?

The platform from Good Solutions is built so that OEE goes from reporting to action. It is not a measurement tool that produces a figure, but a tool to drive improvement work in daily operations. The purpose is concrete: to reduce production losses and costs by increasing factory productivity.

The platform combines machine connectivity, operator tools, dashboards, timeline, reports, quality management, maintenance, Andon, energy, and business implementation in one coherent tool. Operators, production managers, improvement leads, and management work from the same facts, each in the view that suits their role. The loss analysis ensures that the right efforts are prioritized. The energy module allows cost and sustainability to be tracked in the same platform as production.

Just as important as the software is the business implementation. A proven process, led by experts with production experience, ensures that data quality holds and that working methods get in place. A Swedish support organization and a dedicated Customer Success Manager follow the customer over time.

The results speak for themselves.

Bostik improved OEE by 40% and shortened changeover times by 70%. Barilla Wasa increased net production by 15% while CO₂ consumption decreased by 28%. Kavli produced 5,000 tons more in one year, without more shifts or more machines. In all cases, the results came from the improvement work, with OEE as one of several metrics pointing the way.

Read more about how others have increased their factory productivity

FAQ

Is OEE sufficient as the sole KPI for production?
No. OEE is powerful but says nothing about why losses occur. It should be supplemented with loss analysis per cause, quality metrics, changeover times, energy consumption per unit, and flow metrics. Following only OEE often leads to the factory optimizing the figure instead of the business.

Why can't we compare OEE between our factories?
OEE values depend on how each factory has defined planned production time, ideal cycle time, and stoppage causes. Two factories that calculate differently can have the same actual efficiency but different figures. Instead, compare each factory's development over time, and use the differences to share learnings, not to rank.

What does it mean that 80% of stoppages are not machine-related?
That the largest part of production losses is due to things other than machine breakdowns: material shortages, changeovers, waiting for orders, quality issues, planning, and staffing. This means that a factory only investing in better machines misses the greatest potential, and that operators' coding of stoppage causes is crucial for the losses to be addressed.

How do we prevent staff from optimizing the numbers instead of the business?
By never letting OEE stand alone. When the figure is balanced against quality, changeover time, flow, and delivery precision, it becomes difficult to improve OEE in a way that harms the whole. Just as important is to use OEE as a basis for improvement, not as a grade for shifts or individuals.

How do we get started with using OEE for improvement and not just reporting?
Start by securing data quality and giving operators simple tools to code stoppages. Then raise OEE and loss analysis in daily management, such as morning meetings, shift handovers, and weekly improvement meetings. Prioritize a few losses at a time, address them, and follow up. It is in this recurring rhythm that OEE goes from report to result.