Situation:
Two divisions of a domestic manufacturer wanted to develop relationship marketing programs. In order to tailor communications to their prospects, a method was required to determine when the prospect was likely to be in the market.
Solution:
Polk created in-market timing models for both divisions. Multiple time-to-market periods were created, based upon three-month increments:
|
0 - 3 months |
10 - 12 months |
19 - 21 months |
|
4 - 6 months |
13 - 15 months |
22 - 24 months |
|
7 - 9 months |
16 - 18 months |
25 months or more |
The statistical technique used to create the models was a multinomial predictive segmentation model, an extension of the traditional Polk Predictive Segmentation Model to situations where the outcome has more than two possible values.
A telephone follow-up asked prospects when they were likely to be in the market for the particular brand.
Results:
- Telephone survey responses matched up with what was predicted 44% of the time. An additional 10% were off by only one category (i.g., the model predicted 4 - 6 months, but the prospects stated that they would be in the market in 7 - 9 months).
- At least one of the division’s vehicles had its best sales month ever.