The goal of this paper is to evaluate the potential of marketshare models as decision-support tools. We apply systematically different versions of market-share models to a reasonably large set of state-of-the-art quality data (excluding, however, scanner data). In this process, we investigate a series of problems inherent in the specification and estimation of these models.
During the last few years market researchers endeavoured to create a new approach of the consumer behaviour, for instance in the field of life style analysis or more generally of the attitudes of the consumer towards the product or its distribution. Panel operators whose main wish is to offeranas wide as possible knowledge of the consumer- have tried to integrate these new means of analysis in their permanent samples, thus creating new more efficient ways of using consumer panels. Our paper will describe some recent applications on the Secodip consumer panel . First we will present some examples of a better knowledge of the consumer : 1. Geo types 2. Life styles 3- Specific typologies 4. Consumption circumstances In a second part, we will show consumer panel analysis can better integrate the actual supply multiplicity 1. In terms of stores 2. In terms of assortment.
Our objective then is to determine whether consumer panels are truly representative of the market place for soluble coffee. To explain this low coverage, we looked closely to: 1) The kind of consumers sales that the panel was providing us. The kind of consumers sales : only household consumption at home is provided. 2) The kind of product we wanted consumers sales for : Soluble Coffee. The convenient character of soluble coffee is irrefutable, so we can imagine a higher usage when convenience is required : Vacation; Working place; Week; and when living alone.
In this paper we describe how we tried to solve a marketing information system need. First, we explain why we were gradually convinced it was necessary. The second step in our thinking was to list all the data we had and wanted to integrate in it. After that, obviously, we asked ourselves "what do we want to do with such a tool?" The following step was to try to find a package to meet our needs and, given the fact we found no one on the market, to try to develop one. This package has been developped by Sligos and Secodip and is now currently used in L'Oreal and Secodip since one year.
The information necessary for the marketing decision making is generally issued from many chronological and spatial independent systems. It is the case for SECODIP where we find several information systems consumer's panels, retailer panel, advertising expenditure research, ad hoc surveys (schema 1). The main object of the SECODIP data bank is to integrate all these data at the basis. The general organisation of the new system is resumed in the schema 2. We can see that this new system does not take the place of the present one, but completes it. The two main ideas of this new system are the following: 1) The integration of the data is realised at the basis, on the consumption unit (the household); 2) The system restores a specific information adjusted to a given demand itself function of the marketing decision.