Every plant manager I have ever worked with knows their OEE number. Most can tell you the OEE of their best and worst lines. Many have weekly or even daily OEE targets. Some have OEE on the screens in the production hall, updated in real time.
Very few have a reliable answer to the more important question: what is going to happen to our OEE next week?
This is not a technology problem. It is a measurement philosophy problem. OEE is an outcome metric. It tells you what happened. By the time you see it on the dashboard, the losses have already occurred. The equipment has already stopped. The quality defects have already been produced. The scheduled production has already been missed.
The difference between leading and lagging indicators
A lagging indicator measures an outcome. It is backward-looking by definition. OEE, EBITDA, customer satisfaction scores, accident rates — these are all lagging indicators. They are important. But they cannot be managed directly. You cannot improve OEE by focusing on OEE. You improve OEE by improving the things that cause it.
A leading indicator measures something that predicts a future outcome. It is forward-looking. It gives you the opportunity to intervene before the outcome occurs. This is the entire point.
"The goal of a management system is not to measure what happened. It is to prevent what you do not want to happen — and to accelerate what you do."
Most operational management systems are built primarily around lagging indicators. They are excellent at telling you how last month went. They are much less useful at telling you what to do today to make next month better.
What actually predicts OEE
OEE has three components: Availability, Performance, and Quality. Each component is predicted by a different set of leading indicators.
The relationship between leading and lagging indicators is not always linear, and it varies by equipment type, product mix, and operational context. But the principle holds: if your autonomous maintenance compliance is high and your abnormality detection rate is healthy, your availability losses will tend to be low. If they are not, you have an early warning signal — before the breakdown occurs.
Why most plants do not measure leading indicators well
The honest answer is that leading indicators are harder to measure than lagging indicators. An unplanned breakdown is an event — it has a start time, an end time, and a cause. It is easy to record. Autonomous maintenance compliance requires someone to track whether each task was completed, by whom, and when. Abnormality detection requires a culture where operators are actively looking for problems rather than just running the line.
Leading indicators also require a different kind of management attention. A lagging indicator triggers a reaction. A leading indicator requires anticipation. Managing a plant primarily through OEE is reactive by design — you see the loss and you investigate. Managing a plant through leading indicators is proactive — you see the signal and you intervene before the loss occurs.
This requires a different leadership posture. It requires daily management routines that review leading indicators, not just outcomes. It requires a culture where finding a problem is valued, not just fixing it.
Building leading indicators into daily management
In the Tiered Daily Management system that I have implemented across multiple organisations, leading indicators appear at every tier:
- Tier 1 (shift/line level): Autonomous maintenance task completion, abnormalities raised, near-misses reported
- Tier 2 (plant/department level): AM compliance trends, open abnormality age, SPC alerts, preventive maintenance adherence
- Tier 3 (management level): Leading indicator trends across the plant, root cause analysis completion rates, CAPA effectiveness
The key is that at each tier, the leading indicators connect visibly to the lagging outcomes. A supervisor can see the relationship between last week's AM compliance rate and this week's availability losses. A plant manager can see which departments have healthy leading indicators and which are building up risk.
The five leading indicators that matter most
If you are building a leading indicator system from scratch, these five are the highest-value starting points for most industrial environments:
- Autonomous Maintenance compliance rate — Are operators completing their daily/weekly checks? This single indicator predicts more availability losses than almost any other leading measure.
- Abnormality detection and closure rate — Are people finding problems before they become failures? How quickly are tagged abnormalities being resolved?
- Near-miss reporting rate — A low near-miss rate is rarely a sign of a safe environment. It is usually a sign that people do not feel safe reporting. This leading indicator predicts accident rates with high reliability.
- Planned vs reactive maintenance ratio — What percentage of maintenance work is planned? A plant where more than 40% of maintenance is reactive is in firefighting mode. The trajectory of this ratio tells you whether the system is improving or degrading.
- CAPA closure rate and effectiveness — Are corrective actions being completed? More importantly, are they working? Recurring failures are the clearest evidence that CAPA is closing records without solving problems.
These five indicators, tracked consistently and reviewed at the right level of the organisation, give you a forward-looking picture of operational health that OEE alone cannot provide.
What the Control Tower adds
The Capabilium Control Tower is built around the distinction between leading and lagging indicators. The CEO Agent does not just report OEE — it monitors the leading indicators that predict OEE, tracks their trends, and alerts when the signals suggest a problem is building before it becomes visible in the outcome metrics.
More importantly, it connects the leading indicators across domains. A declining AM compliance rate combined with increasing operator near-miss reports and falling engagement scores is a different signal than any of those three indicators in isolation. The system sees the combination and acts on it.
This is not magic. It is systematic application of a principle that good plant managers have always understood: the best time to address a problem is before it becomes one.
Vítor Vila Verde is the founder of Capabilium Partners. He has implemented leading indicator systems across dozens of industrial plants, most extensively during his tenure as VP Operational Excellence at Logoplaste.