Is world-class OEE really 85 percent?

Many vendor pages say world-class OEE is 85 percent and credit Seiichi Nakajima, who named total productive maintenance (TPM). The number is a round target, not a measurement. Whether a machine reaches it depends on how you count.

Where does the 85 percent come from?

The usual story says Nakajima set it in a 1980s TPM book, as 90 percent availability, 95 percent performance, and 99 percent quality. We opened nine English and German vendor and training pages that repeat the number. None gives a page of the book or a quotation from it.

The pages also disagree with each other. Two put quality at 99 percent and two at 99.9 percent. Either way the product rounds to 85:

Quality target	Product
99 percent	0.90 × 0.95 × 0.99 = 84.6 percent
99.9 percent	0.90 × 0.95 × 0.999 = 85.4 percent

The origin stories differ too: Japanese plants that won a TPM prize (two pages), high-volume car making (one page), "the TPM literature" (one page). A 2015 case study in a business journal prints the same four numbers and credits them to a software vendor, Vorne Industries (2008), not to Nakajima.

The book exists. Introduction to TPM (Productivity Press, 1988) translates a Japanese book, probably Nakajima's Seisan kakushin no tame no TPM nyumon, which Japan's National Diet Library dates to 1983. Vendor pages date the book to 1984 or 1988.

What do the sources actually say?

We searched book catalogs, two maintenance handbooks, and open research for the figure in Nakajima's own words. We could not read his book, so we can report what others say it holds, not what it holds. What we could read:

• Gulati (2021) says no process is likely to reach 100 percent OEE and that many manufacturers set a challenging target. He calls 85 percent "not uncommon" and names no source. Nakajima's book is not in that chapter's reading list.
• Wireman (2015) lists OEE values seen in benchmarking from under 20 percent to over 85 percent, with a best practice that "varies". Reaching 85 percent costs too much for some processes, he writes, and is too little for others. He names no source.
• A 2015 master's thesis at KTH says Nakajima (1988) held 85 percent to be the minimum for a world-class company. It cites two older studies with lower values: about 50 percent as an ideal (Kotze, 1993) and 30 percent upward in plants called world class (Ericsson, 1997). We could not open those two.

JIPM, the Japanese institute that runs the TPM Excellence Award, describes it on its English pages with no OEE threshold. So the vendor story that prize winners all scored above 85 percent has no support in what we opened.

Is 85 percent OEE world class?

Verdict: folklore. As a rule for what "world class" means, 85 percent has no source we could open. As a target some plants set, it exists: two handbooks mention it, and neither backs it with data.

A real source would still not settle the question, because OEE has more than one definition. Gulati reports that the SMRP guideline calls the usual calculation, based on scheduled hours, OEE 2. It calls the version based on all 24 hours of all 365 days OEE 1. His example machine scores 77.7 percent on the first and 53.2 percent on the second.

What should you use instead?

We recommend that you compare each machine with its own past, under one written counting rule. Write down what counts as planned time, which ideal cycle time you use, and how you code stops. Then follow the trend and the three factors, and work on the largest loss. This is principle 10 of the Maintenance Manifesto, "Manage the work, not the metric".

Wireman names the opposite habit as the first trap of benchmarking: treating a benchmark as a performance indicator, so that people chase the number and lose the goal. If someone asks you to reach 85 percent, ask three questions. For which machine? Counted how? What does the last stretch cost?

Availability explains why OEE availability differs from the standard's. Uptime and downtime shows which stops EN 13306 counts as downtime. OEE counts every stop in scheduled time. More in Maintenance.

Sources

• Seiichi Nakajima, Introduction to TPM: Total Productive Maintenance, Productivity Press, 1988. Not read; book records only: Open Library, from Library of Congress data (LCCN 88061394, "Translation of: TPM nyumon") and the National Diet Library records for the English edition and the 1983 Japanese book.
• Ramesh Gulati, Maintenance and Reliability Best Practices, 3rd ed., Industrial Press, 2021, chapter 7 (OEE, TEEP, SMRP naming; reading list).
• Terry Wireman, Benchmarking Best Practices in Maintenance, Reliability and Asset Management, 3rd ed., Industrial Press, 2015, chapter 12 (equipment performance, Figure 12-7) and chapter 2 (traps to benchmarking).
• Quentin Fernandez, Performance indicator design and implementation on semi-automated production lines, master's thesis, KTH Royal Institute of Technology, 2015, section 2.3.4.
• Aysha Herdiwan and Sri Widiyanesti, An analysis of machine effectiveness on the production line by using overall equipment effectiveness (OEE) method based on total productive maintenance (TPM) principle, Journal of Corporate Governance, Insurance, and Risk Management 2(1), 2015, pages 106 to 118 (Table 1, credited to Vorne Industries, 2008).
• JIPM, TPM Excellence Awards, Japan Institute of Plant Maintenance, accessed 2026-10-07.
• Maintenance Manifesto, 2025, principle 10.