OEE Benchmarks and Factory KPIs: World-Class Targets and How to Hit Them
World-class and typical ranges for OEE and its three factors, plus throughput, downtime and payback KPIs, with the levers for each and a cited source where one exists.
Use these benchmarks as reference points for spotting a gap, not as targets. OEE figures are widely misquoted, so anchor to the accepted ladder, then let your own trend over many shifts tell you where your line really stands.
Every figure shows its basis. A superscript number links to a public source in the Sources list, and the sentence gives that source's scope, because several of the measured averages come from a single national sample. A figure marked no cited source is kept exactly as originally published because no public source stating it was found, so treat it as illustrative.
At a glance
OEE and factory performance KPI benchmarks at a glance
Words in parentheses give each figure's original framing or its source's scope. Not stated means the original guide gave no figure for that column. The Swedish averages come from one study of 94 OEE values from Swedish manufacturers4.
Method: How the figures in this guide are sourced
A figure with a superscript number is stated by the cited source, listed under Sources with its publisher, location in the document and access date, and the sentence gives its scope. A figure marked no cited source is kept exactly as originally published because no public source stating it was found. It is illustrative, not measured industry data. Worked examples show arithmetic only and are not benchmarks.
OEE
OEE (overall equipment effectiveness) multiplies availability, performance and quality into one score. World-class OEE for a discrete manufacturing line sits at 85 percent2, 3, a figure that traces to Seiichi Nakajima's work on total productive maintenance.
Most plants run well below that. Vorne, a maker of OEE monitoring equipment, reports that most manufacturers are closer to 60 percent1, and that it sees more companies below 45 percent than above 85 percent6.
Measured samples tell the same story. OEE averaged 51.5 percent across 94 values from Swedish manufacturers7, and the averages in published OEE studies range from 40 to 65 percent8. A second Chalmers study, covering 884 machines at 23 companies, found a median of 70 percent and a mean of 65 percent9.
The 85 percent world-class figure decomposes into roughly 90 percent availability, 95 percent performance and 99 percent quality2, and their product is what makes 85 hard: 0.90 x 0.95 x 0.99 is 84.6 percent10. The world-class table in Raju et al., which they attribute to Nakajima, sets the quality rate higher still, above 99.9 percent3.
Establish your baseline over 20 to 30 shifts no cited source before you name a target. A single good shift tells you nothing about the sustainable number you should hold your line to.
How to measure
Multiply availability, performance and quality, each measured over the same planned production time.
Use a validated ideal cycle time for the performance factor.
Record the baseline per line, across many shifts, before setting a target.
What moves it
Availability losses from breakdowns and changeovers.
Performance losses from small stops and slow running.
Availability benchmarks separate the disciplined plants from the rest. World-class availability is 90 percent or better3, 2, and anything below 75 percent signals reactive maintenance and long changeovers no cited source.
Typical plants sit well short of world class. In the Swedish sample of 94 OEE values, with data from 2006 to 2012, availability averaged 78.9 percent, with unplanned stops costing 9.6 percent and setup losses 11.5 percent of planned production time4.
Measure it against planned production time, and track two contributing KPIs: MTBF (mean time between failures) and MTTR (mean time to repair). Raising MTBF and cutting MTTR are the direct levers.
SMED (single-minute exchange of die) changeover reduction is usually the fastest availability gain on a high-mix line. Its target is a changeover under 10 minutes11.
For example, pushing MTBF from 40 hours toward 100, cutting MTTR from 45 minutes to 20, and moving a 40 minute changeover under 10 minutes all hand planned time back to the run.
Maintenance discipline sits underneath those levers. The Department of Energy's Federal Energy Management Program (FEMP) O&M guide, using NASA benchmarks from 2000, sets schedule compliance and preventive maintenance completion above 90 percent and the emergency maintenance percentage under 10 percent12.
How to measure
Divide run time by planned production time.
Log every stop with a reason code, changeovers included.
Track MTBF and MTTR per asset alongside it.
What moves it
Changeover reduction with SMED.
Longer MTBF through planned and preventive maintenance.
Measured performance is far lower. In the Swedish sample, the performance factor, which that study calls operational efficiency and which covers minor stops and speed losses, averaged 67.1 percent, the lowest of the three factors4.
Reported figures can mislead. In that second study, 80 percent of the 884 machines recorded a performance rate of 100 percent, the system default, so their OEE mostly measured availability13.
The gap is almost always small stops and slow running that never make it into the downtime log. Minor stops last from a few seconds to a minute or two and are cleared by the operator14, 15. Every point of performance recovered is capacity you already own and pay for.
For example, a line whose operators quietly run 10 percent under standard rate loses a tenth of its performance factor, and so a tenth of its OEE.
How to measure
Compare actual output with ideal cycle time multiplied by run time.
Validate the ideal cycle time on the line before you rely on it.
Capture short stops automatically, because manual logs miss them.
What moves it
Root cause work on jams and misfeeds to eliminate minor stops.
Rate discipline, so the line runs at its standard speed.
Quality benchmarks are the tightest because scrap compounds through every other factor. World-class first-pass quality is 99 percent or higher2, and the world-class table in Raju et al. puts it above 99.9 percent3. Anything under 95 percent bleeds margin fast no cited source.
In the Swedish sample of 94 OEE values, the quality rate averaged 96.9 percent4. That is the OEE quality factor, good units over units run, not a yield rolled across several process steps.
The improvement levers are SPC (statistical process control) to catch drift before it produces defects, mistake proofing at the station, and tightening the process capability index Cpk toward 1.33 or above, the minimum many industries use as a benchmark5.
For a centered process with two-sided limits, a capability of 1.0 produces roughly 2,700 defects per million; at 1.33 that falls to 64 per million16.
How to measure
Count only units that pass the first time, without rework.
Divide first-pass good units by total units run.
Track Cpk on the characteristics that drive most of your scrap.
What moves it
SPC to catch drift before it produces defects.
Mistake proofing at the station.
Higher process capability on critical characteristics.
Beyond OEE, throughput and balance KPIs tell you where capacity actually leaks. Line efficiency is the ratio of work content to the number of stations times the bottleneck time.
Line efficiency should target 85 percent or higher no cited source; typical assembly lines run 70 to 80 percent no cited source, and the gap is balance loss. Track throughput gap as the percentage below capable rate and hold it under 10 percent no cited source.
Both point straight at the constraint, and the lever is always the same: add capacity or reduce time only at the bottleneck station. Improving any other station moves nothing.
How to measure
Sum the work content of every station on the line.
Divide it by the number of stations times the bottleneck cycle time.
Measure throughput gap as the shortfall of actual output against capable rate.
What moves it
Rebalancing work away from the bottleneck station.
Added capacity at the bottleneck, and only there.
Shorter cycle time at the bottleneck, and only there.
Downtime KPIs deserve their own targets because they drive both availability and cost. Track unplanned downtime as a percentage of planned production time and aim for under 10 percent no cited source.
That aim is close to measured averages rather than beyond them: unplanned stops took 9.6 percent of planned production time in the Swedish sample4. World-class lines hold it near 5 percent no cited source.
For comparison, the DOE FEMP O&M guide benchmarks equipment availability above 95 percent12. That benchmark comes from NASA in 2000 and counts all downtime on facilities equipment against total hours, not only unplanned stops against planned production time12.
Pair it with a Pareto of downtime reasons, since the top three causes typically account for 60 to 80 percent of lost minutes no cited source.
Then rank stoppages by dollars rather than minutes, which reorders priorities fast. A rare 90 minute stop can cost more than dozens of two minute jams, and dollars are the honest tiebreaker.
How to measure
Divide unplanned stop time by planned production time.
Give every stop a reason code and a duration.
Convert each stop to dollars as well as minutes.
What moves it
Fixing the top causes on the Pareto first.
Ranking fixes by cost, not by the count of stops.
The MTBF, MTTR and changeover levers from the availability section.
Smart manufacturing KPIs get benchmarked on payback, not on technology adoption. A sound IoT or monitoring rollout targets a payback under 12 months no cited source and a first-year availability lift of 3 to 8 points on a monitored line no cited source, measured against the pre-install baseline.
The improvement lever here is measurement itself. Plants that move from paper logs to automated capture routinely find their true OEE is 10 to 15 points below what they believed no cited source, and closing that visibility gap is often the single largest one-time gain available.
For comparison, DOE's O&M guide cites unnamed independent surveys in which functional predictive maintenance programs cut maintenance cost 25 to 30 percent, eliminated 70 to 75 percent of breakdowns and reduced downtime 35 to 45 percent17.
EPA estimates that most companies can realize a 15 to 25 percent increase in equipment efficiency rates within three years of adopting TPM (total productive maintenance)18. EPA states it as a percentage increase in efficiency rates, not as a gain in OEE points.
How to measure
Record baseline OEE and availability on the line before installation.
Compare the same line over the same planned time after go-live.
Count payback from measured gains, not projected ones.
What moves it
Automated capture of stops, counts and rejects.
Acting on the newly visible losses, not just displaying them.
Set targets as a staircase, not a leap, and name which factor moves at each step and by how much.
For example, if your baseline OEE is 55 percent, the next milestone is 65, then 75, then a run at 85: 8 points of availability from changeover reduction, then 5 points of performance from minor stop elimination, then 2 points of quality from mistake proofing.
Measure OEE and its three factors over many shifts before you name a target.
Name the factor and the lever behind each step of the staircase.
Tie each target to one owner and one lever.
Review the KPI set weekly against these ranges.
Re-baseline every quarter.
Benchmarks only matter when every number on the board maps to a specific action someone is accountable for this week.
World-Class OEE: Set Targets To Drive Improvement. Vorne Industries (OEE.com). Section 'What Is World-Class OEE?', figure (image alt text) and following paragraph; attributes the numbers to Nakajima, Introduction to TPM (1984; English 1988). Accessed September 30, 2026.
Data-Driven Decision Support for Maintenance Prioritisation: Connecting Maintenance to Productivity (PhD thesis). Chalmers University of Technology, Department of Industrial and Materials Science (M. Gopalakrishnan). published 2018. Thesis p. 22 (PDF p. 34), Table 3 'Average OEE results, with components' (Study A, 94 OEE values, Swedish manufacturers); summary text p. 34 (PDF p. 46): 'availability component (avg. 78.9 percent)', 'operational efficiency component (avg. 67.1 percent)'. Accessed September 30, 2026.
OEE Six Big Losses. Vorne Industries (OEE.com). Sections 'Idling and Minor Stops' and 'Equipment Failure' (threshold rule of thumb). Accessed September 30, 2026.