Parfitt–Collins model was first presented at the
1961 ESOMAR Congress by Baum and Dennis, and subsequently by John Parfitt and Barry Collins
at the 1967 MRS Conference.
Using real data from a household panel, the model estimates market share
with these three metrics:
- Cumulative brand penetration (cumulative brand trial).
- Repeat purchasing rates.
- Buying Index or the heaviness of buying.
Market Share = Trial×Repeat×BI
where:
- Trial: Projected proportion of homes that would try the new product.
- Repeat: Projected proportion of those who continue to repurchase the brand
after having tried it.
- Buying Index: Index of repeat purchases, in terms of weight of purchase.
The model has some limitations, especially when dealing with
products in categories with low penetration or long inter-purchase
intervals. The repeat measure also has its own set of constraints.