This paper investigates the potentialities of the new technology based on neural networks to realize a DSS that might help the marketing and product planning management of a factory in conceiving strategic decisions regarding prices and product performance levels in an oligopolistic market. A possible oligopolistic market scenario assumes a multiplicity of customers whose choices on a given product set have to be explained; the individual choices of these subjects contribute to defining some market shares to be estimated—with their possible evolution—by a multicriteria model (management's problem formulation). Very few computer based systems to support such a complex problem are found in literature; developed essentially with “classic” methodologies and techniques, they suffer from considering either one single decisional subject (the market/consumer) or a multiplicity of monocriterion choices under uncertainty conditions, without taking into account any integrated representation wich might permit a multiactorial interpretation of “the market” to be given. An approach based on neural networks seems to allow such limits to be overcome, as it has been applied successfully to design and develop a prototype of DSS (MUTANT) for the problem formulated. Some first computational experiments allow some positive judgements to be expressed about the system's validity and the promising potentialities of a new “modelling and problem solving” approach that might be usefully applied so as to enlarge some potentialities of other multicriteria decision aid (MCDA) methods.