RTY = FPY1 x FPY2 x ... x FPYn. Rolled throughput yield is the share of units that pass every step right the first time. A four-step process with first-pass yields of 98%, 95%, 99% and 92% has an RTY of 84.8%, much lower than any single step suggests.
Rolled throughput yield calculator
Enter each step in order. The defaults are the four-step example worked through below. Updates as you type.
Only one step? The first pass yield calculator does a single step in more detail.
Good first time, step by step
The working
FPY uses the units entering each step, not the units started at step 1. Reworked units do not count as good first time.
What is rolled throughput yield?
Rolled throughput yield (RTY) is the share of units that get through every step of a process right the first time, with no rework and no scrap.
You get it by multiplying the first-pass yield of each step. It is never higher than the lowest single step, and often a lot lower.
If you need the yield of a single step, start with how to calculate first pass yield. This page is about the whole chain.
The RTY formula
FPY of a step = units good first time / units entering the step
It works like a chain of coin flips. Each step must pass for the unit to count. Many small losses multiply into one big one.
Worked example
Example: a product goes through four steps. The first-pass yields are 98%, 95%, 99% and 92%.
RTY = 0.98 x 0.95 x 0.99 x 0.92 = 84.8%.
Here is the same thing for 1,000 units started:
| Step | First-pass yield | Good first time (of 1,000 started) |
|---|---|---|
| 1 Cut | 98% | 980 |
| 2 Weld | 95% | 931 |
| 3 Paint | 99% | 922 |
| 4 Assemble | 92% | 848 |
| RTY | 84.8% |
Now compare. If you only look at the last step (92%), you think the process is healthy. The true first-pass result is 84.8%. About 15 of every 100 units cost extra work somewhere. That is a hidden factory.
Many of those 152 units were not thrown away. They were fixed. The fix cost labor, machine time and floor space, and none of that shows in a final-yield number.
RTY, final yield and normalized yield
People mix these up. They are different numbers.
| Measure | What it counts | Example |
|---|---|---|
| Final yield | Units that pass the last check, even after rework | Can be near 100% |
| First-pass yield | Good first time at one step | 92% at step 4 |
| RTY | Good first time at all steps | 84.8% |
| Normalized yield | The average step yield that would give the same RTY (RTY to the power 1/n) | 96.0% |
Normalized yield is the geometric mean of the steps: 0.848 raised to 1/4 is about 0.960. It helps when you compare processes with different numbers of steps. A 10-step process at 96% per step has an RTY of only 66.5%, far worse than a 4-step process at the same level.
When defects are random, RTY is close to e raised to minus DPU (total defects per unit across all steps). Read it the other way too: an RTY of 84.8% matches a total DPU of about 0.165, because e to the power of -0.165 is 0.848.
RTY and OEE
RTY and the quality factor of OEE are close cousins, but not the same. OEE quality counts good units at the end of the scheduled time on one machine or line. RTY looks across all steps.
In the method of Overall Equipment Effectiveness (R. C. Hansen, 2001), product that fails first pass counts as a loss for the original run time. That is first-pass thinking, so it fits RTY. If your OEE counts reworked units as good, your OEE quality will look better than your RTY. See availability, performance and quality.
For the quality loss in OEE, also read how to calculate scrap rate.
What the gap costs
The 152 units that failed first time in the example are not free. Say each needs 0.2 labor hours to fix, at 30 euros an hour (both numbers are examples).
152 x 0.2 x 30 = 912 euros per 1,000 units started. And that counts only the labor. The fix also takes machine time, floor space and often a second inspection.
For the money view of defects, see cost of poor quality.
How to measure it on the floor
- List the steps of the process, in order. Include inspection and test steps that can fail a unit.
- At each step, count units entering, units good first time, units reworked and units scrapped.
- Calculate FPY per step: good first time / entering.
- Multiply the FPYs. Do it per product family, not for the whole plant.
- Rank the steps by loss. The step with the lowest FPY is not always the one to fix first. See below.
Which step to fix first
Look at cost and position. A defect found at the end has already used all the earlier steps. A defect found at step 1 only wasted step 1.
In the example, step 4 loses 74 units (922 - 848), each already carrying four steps of work. Step 2 loses 49 units carrying two steps. Here step 4 comes first on both counts. Watch for defects made early but found late: a weld flaw caught only at the final test has used all four steps, so charge it to step 2 and weigh it at four steps of cost.
A value stream map helps to see where rework loops sit. Then use root cause analysis on the biggest one.
How to use RTY in a weekly review
Put the RTY of each product family on one page, next to its OEE. Add a column for the worst step and its FPY.
Then ask three questions each week. Which step lost the most units? Did it get better or worse than last week? What was changed there?
If RTY falls while OEE quality holds steady, your rework is probably being counted as good output somewhere. Find that gap before it grows.
Common mistakes
- Counting reworked units as good first time. That is final yield, not first pass.
- Multiplying across steps that do not apply to every unit. Use one product family at a time.
- Taking FPY from the wrong denominator. The denominator is units entering the step, not units started at step 1.
- Skipping the test and inspection steps. They fail units too.
- Averaging step yields. The average of 98, 95, 99 and 92 is 96, which hides the 84.8%.
Rework hides when counts are typed in at the end of the shift. Fabrico is a light MES for OEE that shows counts and stops per line and shift, so the quality loss of each line sits in the same view as its downtime. See how Fabrico tracks OEE per line.
What people ask next
Is RTY the same as first pass yield?
No. First pass yield is for one step. RTY multiplies the first pass yield of every step, so it shows the whole process. RTY is never higher than the lowest step, and usually well below it.
What is a good rolled throughput yield?
There is no single target. It depends on the number of steps and the cost of a defect. Track your own trend and set a target step by step. Use the first pass yield calculator for each step.
Does RTY include scrap and rework?
Yes. Any unit that is scrapped or reworked at any step does not count as good first time. That is why RTY is lower than final yield.
How is RTY different from the OEE quality factor?
OEE quality is good units divided by total units for one machine or line in the scheduled time. RTY chains the first-pass yields of all steps. If rework units are counted as good in OEE, RTY will be lower. See how to calculate OEE.
How can I raise RTY?
Fix the steps with the largest first-pass loss, weighed by the cost in the unit so far. Mistake-proofing at the source is often the cheapest fix. See poka-yoke and how to reduce scrap and rework.
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: How to Calculate Throughput: Formula and Worked Example · How to Calculate Production Yield by Process · First Time Through: Formula and FPY Difference · Throughput Time vs Lead Time vs Cycle Time · Manufacturing KPI Formulas: A Full Cheat Sheet