Design Win Attrition
Design-win attrition is the fraction of announced design wins that never actually ramp to volume production, and it exists because a design win locks in the customer's intent, not their eventual outcome. The customer's own product can still be delayed, redesigned around a different part, or cancelled outright for reasons that have nothing to do with the supplier at all, and every one of those outcomes converts a real, previously-counted design win into a number that will never show up as revenue, which is exactly why treating a design-win pipeline as a direct revenue forecast systematically overstates what will actually ship.
The causes of attrition sort into two categories that a forecaster has to treat very differently. Attrition caused by the customer's own program, their end product getting delayed, redesigned, or cancelled for reasons unrelated to the supplier's part, is largely outside the supplier's control and has to be estimated statistically from historical attrition rates across similar customer segments. Attrition caused by the supplier's own part, failing qualification, missing a committed sample date, or losing a late re-selection to a competitor, is at least partly within the supplier's control and is the kind of attrition a design-win program can actually work to reduce.
This is why a mature forecasting process discounts the raw design-win pipeline by a historical attrition rate rather than counting every win at face value, and why that attrition rate itself has to be segmented by customer type, end market, and product maturity, since a design win inside an established customer's high-volume consumer product survives at a very different rate than one inside a startup's unproven first product.
| Design-win stage | Typical cause of loss | Within supplier's control? |
|---|---|---|
| Design-in / layout | Customer re-selects a competing part | Partially — depends on support responsiveness |
| Qualification | Part fails the customer's own qual testing | Yes — directly tied to part quality |
| Pre-ramp | Customer's end product delayed or cancelled | No — driven by the customer's own program |
| Post-ramp | Demand forecast was simply wrong | No — ordinary demand-forecast risk |
Design-win attrition is the discount factor that turns the design-win pipeline covered in the previous row into a realistic revenue forecast — tracking it separately from the raw pipeline count is what keeps that forecast honest.