Understanding Yield in WATS: Calculation and time filter

Yield is the most important metric in WATS, because it tells you how well your production and test process is performing. It can help you identify quality issues, unstable test stations, or process steps that create rework.

This article explains:

• What yield means in WATS
• The two main ways of looking at yield
• How yield is calculated
• How the time filter affects yield calculations
• The key concept: “First Seen in Process”

What is yield?

In simple terms, yield is the percentage of items that pass testing.

However, in WATS, yield can be calculated in different ways depending on whether you are measuring:

• Test performance (how many test reports pass)
• Product performance (how many physical units pass)

This is why WATS provides two main yield types.

Two ways of looking at yield in WATS

1. Test Report Yield (TRY)

Test Report Yield (TRY) is calculated using test reports only:

TRY = Passed test reports / All test reports

This calculation does not consider the unit serial number.
That means every test run counts as a separate data point.

When is TRY useful?

TRY is often the best way to evaluate:

• Test station performance
• Operator performance
• Test program stability
• Whether a specific tester produces unusually many failures

2. Unit-Level Yield (FPY / SPY / TPY / LPY)

Unit-level yield takes the unit serial number into account.

Instead of counting test reports, WATS tracks each individual unit across multiple test runs.

This is the yield type used to answer questions like:

• How many units pass immediately?
• How many units require retests?
• How many units never pass at all?

In WATS, unit yield is calculated using different yield levels based on how many attempts were required.

How Unit-Level Yield is calculated

• FPY (First Pass Yield): Units that pass on the first run / Total units
• SPY (Second Pass Yield): Units that pass within 2 runs / Total units
• TPY (Third Pass Yield): Units that pass within 3 runs / Total units
• LPY (Last Pass Yield): Units that eventually pass / Total units

FPY is typically the best indicator of true process quality, while LPY is often used to measure final output quality even when retesting or rework occurs.

Why the time filter matters

The time filter in WATS is critical for yield calculations, but it behaves differently depending on which yield type you are looking at.

TRY time filtering

For Test Report Yield, the time filter works the way most people expect:

• WATS includes all test reports within the selected time range
• Passed reports and failed reports are counted directly

So if you filter for “Last 7 days”, TRY uses only test reports that occurred in those 7 days.

Unit-Level Yield time filtering

For unit-level yield, the intuitive interpretation of the time filter is incorrect.

This is because unit yield depends on the full history of a unit, including its first run and all later runs.

To solve this, WATS applies the concept:

First Seen in Process

When calculating unit-level yield, WATS does two things:

1. Only includes units whose FIRST test run occurred within the time filter
2. Includes ALL available test data for those units, even if some later runs happen outside the filter

This ensures that WATS can correctly determine whether a unit passed on:

• Run 1 (FPY)
• Run 2 (SPY)
• Run 3 (TPY)
• Or eventually (LPY)

Why WATS uses this method

If WATS only included test reports inside the time range, yield would be misleading.

Example:

• A unit fails on Monday
• It is retested and passes on Friday
• You filter the data to show only Friday

If WATS only looked at Friday, the unit would appear as a first-pass pass, which is incorrect.

By using First Seen in Process, WATS ensures:

• FPY is truly “first attempt”
• SPY and TPY are calculated correctly
• LPY reflects eventual outcomes

Summary

WATS provides two main yield perspectives:

Test Report Yield (TRY)

• Based on test reports
• Ignores serial numbers
• Best for evaluating station or test performance
• Time filter includes reports strictly inside the range

Unit-Level Yield (FPY / SPY / TPY / LPY)

• Based on serial numbers and test run history
• Best for evaluating product/process quality
• Uses the First Seen in Process principle
• Time filter selects units based on first run, then includes all their data