TITLE:
The Role of Seeding in Multi-Stage vs. Two-Stage Diffusion Models
AUTHORS:
Yair Orbach, Gila E. Fruchter
KEYWORDS:
Diffusion of Innovation, Seeding, Sampling, Purchase Process
JOURNAL NAME:
Modern Economy,
Vol.8 No.3,
March
31,
2017
ABSTRACT: Seeding the market by
delivering free samples is a common strategy marketers use to promote new
products, shorten the introduction stage and accelerate diffusion. There are several models for assessing seeding effectiveness and
determining the optimal seeding that can maximize profits, given costs,
diffusion parameters’ values and interest rates. In practice, however, managers
use seeding mainly as a competitive tool, as a way to improve market position rather
than to promote and accelerate the diffusion of a new product category. Actual
seeding is quite small in volume compared to the optimal levels recommended by
existing models. Seeding is also accompanied by other steps which do not
involve giving away products. It is argued here that the problem is not the
fact that marketers do not implement an optimal strategy, but rather the
modeling itself. We present a method for analyzing seeding effectiveness and
optimization, based on a diffusion model with a multi-stage purchase process,
as suggested by Kalish. We compare the diffusion acceleration and optimal seeding
assessments of our model against existing seeding assessment and optimization
diffusion models that are based on a two-stage purchase process as described in
the Bass model. To achieve optimization our model recommends seeding, as a
means to accelerate the diffusion of new product categories, only in rare
cases. This coincides much better with firms’ actual implementation of seeding
strategies than seeding levels recommended by models in the literature. In
particular, it is aligned with managers’ intuition that diffusion models with a
two-stage purchase can lead to oversampling and an overestimated sales
acceleration forecast.