Chrysler planned $45B of incentives with a dealer-data pricing model

Automakers keep manufacturer's suggested retail prices and wholesale prices fixed through the model year, customizing price with incentives - roughly $45 billion per year in the US market.

Chrysler, a pioneer in science-based pricing, engaged J.D. Power and Associates to implement an incentive planning model.

The approach uses a random-effects multinomial nested logit model of vehicle, acquisition (cash, finance, lease), and program-type choice, trained on sales transaction data collected daily from approximately 10,000 dealerships.

Predict response before spending incentives

The logit model predicted how incentives change share

Daily dealer data kept the model current

Nested choice captured vehicle, finance, and program trade-offs

Planning against response, not budgets, priced the $45B market

When a market spends tens of billions on one lever, build the demand model first: predicting share response per incentive dollar turns promotion planning into an allocation problem.

Chrysler Group and J.D. Power and Associates were named 2007 Franz Edelman Award finalists for PIN. The Interfaces paper documented the model and its role in planning automobile manufacturers' pricing and promotions.