How do you calculate pilot yield?

Divide the number of good (first-pass) units by the total units started in the pilot, then multiply by 100. With 8 good units out of 250 started, pilot yield is 8 ÷ 250 × 100 = 3.2%.

What is a good pilot yield?

For a mature design transferring to volume, pilots often target 90-95% first-pass yield. The example here yields 3.2% against a 95% target — a 91.8-point gap that means the build is nowhere near launch-ready.

Should pilot yield count rework as good units?

No. Pilot yield is a first-pass metric. Units that needed touch-up, retest or rework should be excluded from the good count, because at volume that rework time becomes a real cost and bottleneck.

Pilot yield vs production yield — what's the difference?

Pilot yield measures a small pre-production build on near-final tooling and instructions; production yield measures steady-state volume. Pilot yield is usually lower and is used to predict and fix issues before they scale.

Why is my pilot yield so low?

Common drivers are immature work instructions, prototype-grade fixtures, incoming material variation, and unresolved DFM/DFA issues. A 3.2% yield like the example almost always points to a systemic problem, not random defects.

How many units do I need in a pilot to trust the yield?

It depends on expected defect rates, but a few hundred units is typical for a first read. Treat the number as directional and pair it with Pareto analysis of failure modes rather than reading it as a precise production forecast.

NPI, DFM/DFA & Engineering Change calculator

Pilot Yield Calculator

Pilot yield is the percentage of units from a pre-production pilot build that pass all inspection and test criteria on the first pass, with no rework. NPI engineers, launch managers and quality teams use it as the primary go/no-go gate before authorizing volume ramp. A pilot that yields well below target signals unresolved DFM issues, immature work instructions, or fixturing problems that will multiply at scale. Catching that gap at the pilot stage is far cheaper than discovering it after the line is running at rate.

What this calculator does

Estimate pilot yield for npi, dfm/dfa and engineering change using production-ready inputs so teams can track KPI performance and decide whether corrective action is needed.
Use it when pilot yield in npi, dfm/dfa and engineering change needs a clean rate and gap-to-target you can put on a tier board.
It computes first-pass pilot yield as good units divided by units started, then the gap in percentage points to your launch target.

Formula used

Pilot yield rate = pilot yield count ÷ total pilot yield population × 100
Pilot yield gap to target = pilot yield rate - target pilot yield rate

Inputs explained

Good units off the pilot build:
Total units started in the pilot run:
Target first-pass pilot yield:

How to use the result

Use it after each pilot or pre-production build to decide whether a product is ready to move to volume ramp or needs another build loop.
A single pilot is a small sample, so a yield number from 250 units carries wide statistical uncertainty and can swing on one or two defect modes.

Common questions

How do you calculate pilot yield? Divide the number of good (first-pass) units by the total units started in the pilot, then multiply by 100. With 8 good units out of 250 started, pilot yield is 8 ÷ 250 × 100 = 3.2%.
What is a good pilot yield? For a mature design transferring to volume, pilots often target 90-95% first-pass yield. The example here yields 3.2% against a 95% target — a 91.8-point gap that means the build is nowhere near launch-ready.
Should pilot yield count rework as good units? No. Pilot yield is a first-pass metric. Units that needed touch-up, retest or rework should be excluded from the good count, because at volume that rework time becomes a real cost and bottleneck.
Pilot yield vs production yield — what's the difference? Pilot yield measures a small pre-production build on near-final tooling and instructions; production yield measures steady-state volume. Pilot yield is usually lower and is used to predict and fix issues before they scale.
Why is my pilot yield so low? Common drivers are immature work instructions, prototype-grade fixtures, incoming material variation, and unresolved DFM/DFA issues. A 3.2% yield like the example almost always points to a systemic problem, not random defects.

Last reviewed 2026-05-12.