Forecast Wizard

Peter Fader of the Wharton School and Bruce Hardie of the London School of Business have developed an ingenious approach to produce marketing forecasts based on commonly available customer data.  As one example, the Forecast Wizard can turn scanner data into projections for new product adoption.  In another example, the Forecast Wizard can calculate each customer's lifetime value to the firm based on either annual counts of active customers or data on recency, frequency and monetary value of purchases.

In all cases, Fader-Hardie models simultaneously produce forecasts at the individual and group levels based on assumptions about each individual's buying process.  Those assumptions in turn depend on the nature of the purchase transactions (contractual vs. non-contractual) and the type of data (counts, continuous time duration, discrete time duration or choices).  The Wizard helps guide you through data setups and technique execution.   Forecasts produced by the resulting models tend to be remarkably accurate.

  • From This
  • To This

Forecasting: Customer Relationships (video)
Forecasting : Cruises & Concerts Example (video)

Forecast Wizard Benefits

  • Greatly reduced time in building and estimating models
  • Accurate forecasts using a limited amount of data
  • Automatic technique selection, reducing analysis time
  • Facilitiates learning about Fader-Hardie probability models
  • Typically pays for itself in one project

Forecast Wizard Features

Ease of use

  • Templates serve as a guide for organizing data
  • Context-sensitive access to academic papers and lectures
  • Wizard format reduces the need for training
  • Running the Wizard with a blank workbook leads to a database that helps decide on which analysis to run
  • Automatic technique assessment

Automatic recommendations

  • Technique
  • Data type
  • Relationship with customer

Data flexibility

  • Wizard senses the need for data reorganization and performs it automatically
  • Variation from the templates is tolerated
  • Tables and charts can be cut and pasted into other applications

Breadth of applications

  • Identifying underlying customer segments
  • Calculating customer lifetime value (CLV)
  • Forecasting contract renewal rates
  • Forecasting trial and repeat sales of new products
  • Understanding the drivers behind the frequency of visiting a web site