Eight lean metrics cover most of what matters on a line: takt time, cycle time, lead time, work in process, value-added ratio, first time through, changeover time and OEE. Pick the ones tied to your current problem, usually three or four, and show them where the crew can act on them.
Start from flow, not from a dashboard
Lean asks a simple question: how smoothly does work move from start to finish, and how much of the time is waste? A good lean metric answers part of that and points to something a team can change.
A long list does neither. So below are eight metrics worked through on one line, then advice on which to choose.
The example line
Example: a line assembles a small pump, with 450 available minutes a day and customer demand of 900 units a day. All numbers on this page are hypothetical.
| Metric | Formula | Example |
|---|---|---|
| Takt time | Available time / demand | 27,000 s / 900 = 30 s |
| Cycle time | Run time / units, at the slowest station | 25 s |
| Work in process | Units started, not finished | 1,800 units |
| Lead time | WIP / throughput | 1,800 / 900 = 2 days |
| Value-added ratio | Value-added time / lead time | 5 / 900 min = 0.56% |
| First time through | Units with no second touch / units started | 900 / 957 = 94.0% |
| Changeover time | Last good unit to first good unit | 38 min |
| OEE | Good units x ideal cycle time / planned time | 900 x 21 s / 27,000 s = 70.0% |
Takt time and cycle time: the pace
Takt is the pace demand asks for, and cycle time is the pace the line can hold. If the slowest station is above takt, the line cannot meet demand without overtime.
Here, 25 seconds against a 30-second takt looks comfortable, with 5 seconds of margin. Our guide on takt, cycle and lead time gives the full differences, and the takt time calculator does the math.
Lead time and work in process: the wait
These two are tied together by Little's law: average lead time = average WIP / throughput. You can read lead time from the floor by counting units, with no stopwatch needed.
Example: 1,800 units waiting and moving, with 900 units a day leaving, gives a 2-day lead time. Cut WIP to 900 with the same output, and lead time halves to 1 day.
The math holds for averages over a stable period. For a first look that is enough, and manufacturing lead time covers the details.
Value-added ratio: the eye-opener
Also called process cycle efficiency, it compares time spent changing the product with the total time in the plant. The result can be a shock the first time people see it on a value stream map.
Do not chase this figure as a weekly target, because it moves slowly and depends on how you define value. Use it once a year on a value stream map to show where the big wait is.
First time through: the quality of the flow
Each second touch breaks the flow and adds work. First time through counts units that needed no scrap, rework or retest.
Here 957 units start to give 900 clean ones, so 57 units a day take a detour. Across a whole route, multiply the steps, as in rolled throughput yield.
Changeover time: the price of small batches
Lean wants small batches, and small batches need quick changeovers. Time it from the last good unit of one product to the first good unit of the next, with the method in how to measure changeover time.
Example: one 38-minute changeover a day takes about 8% of the 450 minutes. Move to three smaller batches and setup takes 114 minutes, or about 25%, which is why SMED comes first.
OEE: how well the line uses its time
OEE comes from Total Productive Maintenance, launched by Nakajima in the 1980s (see Muchiri and Pintelon, 2008), and many lean teams use it too.
It multiplies availability, performance and quality, the method set out in Overall Equipment Effectiveness (R. C. Hansen, 2001).
Here is a point the table hides: this line meets takt with an OEE of 70%. With an ideal cycle of 21 seconds, its required OEE is ideal cycle / takt = 21 / 30 = 70%, so any stop above today's level misses demand.
That is what the 5-second margin was hiding. At 25 seconds, 957 units need 399 of the 450 minutes, and the changeover and other stops take the remaining 51, so the line has no spare time.
Which metric shows which waste
Lean's 8 wastes give you a better way to choose than a template. Find your biggest waste, and take the metric that makes it visible.
Track lead time and WIP. The waste is waiting and inventory.
Track OEE and cycle time against takt. The waste is waiting and motion at the line.
Track first time through. The waste is defects.
Track changeover time. The waste is overproduction and inventory.
Common mistakes
- Tracking all eight on every board. Three or four tied to a current problem get more action than eight that nobody reads.
- Treating takt as a target for speed. Takt comes from demand, and running faster than it just builds stock.
- Lowering WIP with no plan. Buffers protect against stops, so cut them as stability improves, or the line starves. See blocked and starved.
- Pushing OEE on every machine. A machine that runs ahead of the constraint only makes WIP. Focus on the bottleneck.
- Reporting and not walking. The number starts the question, and the answer is on the floor, which is why the gemba walk remains part of the routine.
Getting started
Choose one line and calculate all eight once, by hand, in a single morning. Then keep the three or four that point to your biggest problem, and post them at the line.
For the wider management set, including delivery and cost, see production KPIs. For the improvement routine that uses these numbers, see kaizen.
Most of these metrics can start on paper, but stops and speed are hard to catch by hand. That is the part a light MES for OEE like Fabrico handles: stops, speed and counts come live from the line, and operators tag the cause at the line. See how Fabrico tracks OEE on a line.
What people ask next
What are the most important lean metrics?
On a line, the usual core is takt and cycle time, lead time and work in process, first time through, changeover time and OEE. Choose three or four that match your current problem, not the full list.
Is OEE a lean metric?
OEE comes from Total Productive Maintenance, which is a close relative of lean, and lean plants use it widely. It shows how well a line uses its scheduled time. See what is OEE.
How do you calculate process cycle efficiency?
Divide value-added time by total lead time. With 5 minutes of value-added work inside 900 working minutes of lead time, the result is 0.56%. A value stream map gives you both numbers.
How do you calculate lead time from WIP?
Use Little's law: average lead time equals average work in process divided by throughput. 1,800 units of WIP at 900 units a day gives 2 days, as covered in manufacturing lead time.
What is the difference between lean metrics and production KPIs?
Lean metrics focus on flow and waste. Production KPIs are a wider group that also covers output, cost and delivery, and the two lists overlap.
Fabrico is a light MES for OEE. It takes stops, speed and counts from PLCs or IoT sensors, with add-on AI cameras counting output where a machine has no usable signal, and operators tag the cause at the line.
Related guides: Takt Time vs Throughput: How to Compare Them · Assembly Line Efficiency: Formula and Example