Overview of Warranty Prediction

Use Warranty Prediction to forecast future warranty claims or returns based on historical warranty data. A warranty analysis uses information about past warranty claims to predict the number and cost of warranty claims in the future. By fitting a distribution to your warranty data, you can estimate the number of expected failures in the next month, the next year, or other period of time. Using the results of the analysis, you can better allocate resources to address future product failures adequately.

Suppose the quality manager of a lawn equipment manufacturer needs to budget for future warranty claims. Lawn mowers are covered by a one-year warranty and the cost to the company for each warranty claim is $1000. Each month, the manager records the number of lawn mowers from each shipment that fails within the warranty period in order to project future warranty claims.

The manager performs the following sequence of analyses to evaluate the warranty data:
  • Pre-Process Warranty Data to convert the data from triangular matrix format to arbitrary-censored format
  • Distribution ID Plot (Arbitrary Censoring) to determine which of 11 distributions best fits the data
  • Warranty Prediction to predict the number and cost of future claims

Where to find this analysis

To perform warranty prediction, choose Stat > Reliability/Survival > Warranty Analysis > Warranty Prediction.

When to use an alternate analysis

  • If your data are in triangular matrix form in the worksheet, use Pre-Process Warranty Data to reformat your data into arbitrarily-censored time-to-failure data before you perform this analysis.
  • If you do not know which distribution fits your warranty data, use Distribution ID Plot (Arbitrary Censoring) before you perform this analysis.
By using this site you agree to the use of cookies for analytics and personalized content.  Read our policy