Alfred Taudes

Stochastic models of consumer behaviour

Abstract During the past thirty years much research has been carried out in order to produce decision aids for managers. Considerable efforts have been dedicated to models that describe how consumers make purchase decisions. Within this vast area of research we review stochastic approaches to consumer behaviour, i.e. probabilistic laws relating the observed strings of consecutive purchases to explanator variables. We start out by discussing the basic concepts underlying this field. Then we deal with models concentrating on brand choice, whereby we trace the development from the early ideas in the Fifties to todays sophisticated models. Subsequently we turn to the Negative Binomial Distribution, the ‘classical’ purchase incidence model, and its extensions. Then we describe combined purchase timing-brand selection models and introduce models of store choice and purchase quantity selection. After briefing on estimation and validation methods for stochastic models we finally critically review the state-of-the-art in this field. Special emphasis is placed on empirical applications throughout the review.

Integrating Pricing and Inventory Control: Is it Worth the Effort?

In this paper we first show that the gains achievable by integrating pricing and inventory control are usually small for classical demand functions. We then introduce reference price models and demonstrate that for this class of demand functions the benefits of integration with inventory control are substantially increased due to the price dynamics. We also provide some analytical results for this more complex model. We thus conclude that integrated pricing/inventory models could repeat the success of revenue management in practice if reference price effects are included in the demand model and the properties of this new model are better understood.

Integrating inventory control and a price change in the presence of reference price effects: a two-period model

Demand and procurement planning for consumer electronics products must cope with short life cycles, limited replenishment opportunities and a willingness to pay that is influenced by past prices and decreases over time. We therefore propose the use of an integrated pricing and inventory control model with a two-period linear demand model, in which demand also depends on the difference between a price-history-based reference price and the current price. For this model we prove that the optimal joint pricing/inventory policy for the replenishment opportunity after the first period is a base-stock list-price policy. That is, stock is either replenished up to a base-stock level and a list-price is charged, or it is not replenished and a discount is given that increases with the stock-level. Furthermore, we use real-world cell phone data to study the differences between an integrated policy and traditional sequential optimization, where prices are initially optimized based on the expected demand and ordering cost, and the resulting demand distribution is used to determine an optimal inventory policy. Finally, we discuss possible extensions of the model.

A hedonic wage model‐based methodology for evaluating the benefits of IT investments in public‐sector organisations

The economic justification of investments in information technology (IT) is a basic issue for IT management in private and in public‐sector organisations. Usually, the expenses made for any investment are compared to the cost saved. While the costs for implementing a new system are uncertain, only a small percentage of the benefits accrues as cost savings given the type of IT systems used today and the particularities of the public‐sector. In this paper, we present a methodology for the monetary quantification of the benefits resulting from the introduction of a modern IT application and demonstrate its use on the basis of a case of prison administration.

Real option valuation with neural networks

We propose to use neural networks to value options when analytical solutions do not exist. The basic idea of this approach is to approximate the value function of a dynamic program by a neural net, where the selection of the network weights is done via simulated annealing. The main benefits of this method as compared to traditional approximation techniques are that there are no restrictions on the type of the underlying stochastic process and no limitations on the set of possible actions. This makes our approach especially attractive for valuing Real Options in flexible investments. We, therefore, demonstrate the method proposed by valuing flexibility for costly switch production between several products under various conditions.

Bundling decisions in supply chains

Firms often sell products in bundles to extract consumer surplus. While most bundling decisions studied in the literature are geared to integrated firms, we examine a decentralized supply chain where the suppliers retain decision rights. Using a generic distribution of customers’ reservation price we establish equilibrium solutions for three different bundling scenarios in a supply chain, and generate interesting insights for distributions with specific forms. We find that (i) in supply chain bundling the retailer’s margin equals the margin of each independent supplier, and it equals the combined margin when the suppliers are in a coalition, (ii) when the suppliers form a coalition to bundle their products the bundling gain in the supply chain is higher and retail price is lower than when the retailer bundles the products, (iii) the supply chain has more to gain from bundling relative to an integrated firm, (iv) the first-best supply chain bundling remains viable over a larger set of parameter values than those in the case of the integrated firm, (v) supplier led bundling is preferable to separate sales over a wider range of parameter values than if the retailer led the bundling, and (vi) if the reservation prices are uniformly distributed bundling can be profitable when the variable costs are low and valuations of the products are not significantly different from one another. For normally distributed reservation prices, we show that the bundling set is larger and the bundling gain is higher than that for a uniform distribution.

Technological Efficiency and Organizational Inertia: A Model of the Emergence of Disruption

We study the influence of technological efficiency and organizational inertia on the emergence of competition when firms decide myopically. Using a multi-agent computer simulation model, we observe the competitive reaction of a former monopolist to the advent of a new competitor. While the entrant uses a new technology, the monopolist is free either to stick to his former technology or to switch to the new one. We find that—irrespective of details regarding the demand side—a change of industry leadership occurs only if the new (“disruptive”) technology is not too efficient and organizations are inert.

Implementing Branch-and-Bound Algorithms on a Cluster of Workstations — A Survey, Some New Results and Open Problems

Networks of workstations running under a multiuser, multitasking operating system like UNIX are an increasingly commonplace personal computing environment. Due to their use as personal computers these workstations are typically underutilized most of the time. Thus it is attractive to develop software to use the ample free computing resources to configure a loosely coupled multicomputer to solve computation intensive problems in a distributed fashion. In this paper we discuss the feasibility of implementing Branch-and-Bound algorithms for combinatorial optimization on a cluster of workstations. There by we use experiences made by us when solving the Vertex-Cover-Problem on a cluster of 8 HP 9000/330 workstations under HP-UX connected via Ethernet and reports from literature about combinatorial optimization on multicomputers. Besides presenting performance results we discuss programming techniques for balancing the workload, for interprocess communication and for distributed termination. Based on this evidence we conclude that given proper tuning a distributed Branch-and-Bound algorithm can yield satisfactory speed-up on a cluster of workstations. However, tools are needed that make the development and run-time control of such applications easier while preserving the favourable efficiency.

Microdynamics of new product purchase: A model incorporating both marketing and consumer-specific variables☆

Abstract The Negative Binomial-Dirichlet Model of brand choice and purchase incidence has been found to be capable of describing consumer behaviour in a large number of empirical applications. However, a major drawback of this model is the postulate of stationary market conditions. The multivariate Polya Model generalizes this approach in order to overcome the before-mentioned shortcoming. The present paper proposes a further extension: the incorporation of consumer-specific variables. This Extended Polya Model is thus suited to describe dynamic changes at the micro level of individual buyers. Amongst others a useful feature of the Extended Polya Model is its capability to account for asymmetries between brands. It is shown that the multivariate Negative Binomial Distribution Model and the Polya Model with homogeneous explanatory variables are nested within this model. This way the market analyst is provided with a general and flexible approach that allows the modelling of consumer behaviour with various degrees of complexity. The models are applied to empirical data concerning the introduction of a new brand in the German detergent market, and the results obtained are compared with regard to descriptive power and computational effort. It is found that the buyers of the new brand are more price sensitive (assessed in terms of direct and cross elasticities) than consumers of the well-established brands.

Organizational learning in production networks

Abstract If one accepts that a firm’s behavior is determined by history-dependent capabilities that adapt in a goal-directed way, one would like to know how a firm’s organizational structure influences the manner in which this distributed and partially tacit organizational memory evolves over time. In this paper, we study the impact that alternative information systems, incentive systems and modes of learning coordination have on the efficiency and generality of priority rules for job-shop scheduling which are learnt by a network of production agents modeled by neural networks. When modeling the alternative organizational structures by different input layers, feedback and training methods, we find that efficient rules evolve when global incentives and synchronized learning are employed even if the system state is only partially known to an agent. However, organizational learning fails when it is performed asynchronously with local goals.