Which KPIs to track during an OEE pilot

Without clear indicators, a pilot decides nothing

An OEE pilot only has value if it leads to a decision. And a decision is taken on indicators, not on impressions. Too many improvement efforts fail not for lack of data, but for lack of indicators chosen in advance and tracked rigorously. The first job of a well-run pilot is therefore to define, from the outset, the KPIs that will settle the matter.

These indicators must be few in number, legible, and tied to a possible action. A dashboard crowded with curves is useless if nobody knows what to do with it. Conversely, three or four well-chosen KPIs, tracked in real time and discussed as a team, are enough to turn measurement into a concrete lever for improvement.

Real OEE, the central indicator

The master KPI of a pilot is real OEE, measured continuously, machine by machine and shift by shift. Its strength lies in being calculated from objective data recorded to the second, rather than reconstructed after the fact from manual readings. It tells the truth of the line, without the flattering smoothing of self-reporting.

The first use of this KPI is comparison with declared OEE. The gap, often several dozen points, makes the starting point objective and creates the wake-up call. It is from this gap that the rest of the pilot is built: you no longer argue about whether the figure can be trusted, you work on real performance and on the potential it reveals.

The split of losses by cause

Measuring OEE is not enough: you have to know where the minutes go. The second essential KPI is therefore the split of losses between long stops, micro-stops, speed losses and quality defects. This breakdown turns a global figure into an action map: it shows which family of loss to concentrate the effort on to get the biggest gain.

It is the direct application of the Pareto principle. Very often, two or three dominant causes explain the majority of a line’s losses. Tracking their evolution over time lets you check, week after week, whether the actions taken really reduce the targeted loss. Without this breakdown, you would act blind, on assumptions rather than on facts.

The real duration of changeovers

On lines with high product variety, a third KPI becomes essential: the real duration of changeovers, and the ramp-up phase that follows them. This lever is often underestimated because manual recording does not measure it finely. Yet every changeover combines a stop time and a period of under-pace, two losses that must be tracked together.

Measuring these durations in real time lets you compare changeovers with one another, between teams and between products. You then discover gaps in practice that are sometimes considerable, each one a piece of good practice to identify and spread. It is a particularly profitable improvement field, because it acts on a recurring loss with no material investment at all.

The tracking rhythm: a KPI lives in real time

A pilot KPI only has value if it is tracked at the right rhythm. Real-time data allows daily tracking, even at every shift, where manual recording forced you to wait until the next day. Shortening this loop between measurement and action is one of the major contributions of a well-instrumented pilot.

In practice, a short ritual of reading the KPIs, on the floor, in front of the screen, is enough to anchor the approach. The team sees the previous day’s performance, spots the anomaly, decides on an action, and checks its effect the next day. This fast cadence is what distinguishes a living pilot from mere passive monitoring that improves nothing.

The final gain, expressed in OEE points and capacity

At the end of the pilot, the summary KPI is the gain obtained, expressed in OEE points recovered and in additional capacity freed up. It is this figure that justifies the decision to deploy. Because it rests on a real before/after measurement, it is neither estimated nor promised: it is observed on the pilot line itself.

This gain then translates into the language of leadership: production hours gained, capacity freed with no capex, additional margin. Hutchinson improved its OEE from 42% to 75% with the same headcount and machines, sensor installed in under an hour. Converting OEE points into capacity and value is the step that turns a technical result into an economic argument, understandable by every decision-maker.

Avoiding false indicators

Not all indicators are equal. Some reassure without saying anything: a flattering declared OEE, an underestimated number of stops because micro-stops are not counted, a quality rate that ignores start-up losses. A rigorous pilot sets these false indicators aside in favour of objective, complete measurements.

The rule is simple: a KPI must be measured, not declared; complete, not partial; and tied to an action, not merely to an observation. Only on that condition does it become a decision tool rather than a piece of reassurance. The quality of the KPIs determines, at bottom, the quality of the final decision.

Breaking OEE down to act better

Global OEE is useful for measuring the result, but it is its breakdown into three components (Availability, Performance, Quality) that guides action. The same OEE of 60% can hide very different realities: a line weighed down by stops does not call for the same decisions as a line penalised by under-pace or by scrap. Tracking the three components separately avoids fighting the wrong battle.

This breakdown is one of the great contributions of reliable automatic measurement. Where a manual reading gives an aggregated, approximate figure, continuous measurement isolates each component and its evolution. You immediately see which one is improving and which is stalling, which lets you redirect the effort mid-pilot rather than at the end, when it is too late to correct.

Sharing the KPIs to build buy-in

A KPI kept in an office improves nothing. The value of a pilot indicator also lies in its sharing: displayed on the floor, discussed with the teams, it becomes a common object around which improvement is built. Operators, who know their line better than anyone, bring the context that data alone does not provide.

This sharing transforms the perception of measurement. It stops being experienced as a control and becomes a tool in the service of the teams, who finally see objectified the difficulties they were suffering without being able to document them. This buy-in is decisive for what follows: a successful rollout rests as much on ownership of the indicators as on the quality of the measurement. More than 450 plants across 30+ countries already monitor their OEE to the second with TeepTrak.

The discipline of indicators, not the technology

An OEE pilot is steered with a few well-chosen KPIs: real OEE by line and by shift, the split of losses by cause, the real duration of changeovers, and the final gain in OEE points. They all rest on objective real-time measurement, tracked daily in a short ritual. It is this discipline of indicators, more than the technology itself, that turns measurement into a decision.

A team that meets briefly every day in front of the same screen ends up reasoning differently. The figure stops being a subject of debate and becomes a shared basis for action. That cultural shift, born of a handful of reliable indicators read together, is often the most durable outcome a pilot leaves behind, well beyond the OEE points it recovers.

FAQ

What is the main KPI of an OEE pilot?
Real OEE, measured continuously and compared with declared OEE. The gap between the two makes the starting point objective and serves as the basis for the whole improvement approach.

Why track the split of losses?
To know where the minutes go and act on the two or three dominant causes following the Pareto principle, rather than on assumptions. That is what makes the action effective.

Which KPIs should I track during the 60 days?
Real OEE by line and by shift, the loss split (stops, micro-stops, speed, quality), the real duration of changeovers, and the final gain in OEE points recovered.

At what rhythm should these indicators be tracked?
In real time, ideally every day or at every shift, in a short reading ritual on the floor. Real-time data shortens the loop between measurement and action.

How is the final gain measured?
In OEE points recovered and in additional capacity freed up, measured before/after on the pilot line. This gain then translates into production hours and margin, with no capex.