U bent hier


Opgelet! Dit event heeft al plaatsgehad.

Competitive Location Models and Consumer Spatial Behaviour

woensdag, 29 september, 2010 - 17:00
Campus: Brussels Humanities, Sciences & Engineering campus
Faculteit: Social Sciences and Solvay Business School
Lieselot Vanhaverbeke

Location, location, location. It is an often-heard saying when retailers
are asked about the key success factors for their business.

The ‘where?’-decision for a retailer has a long-term impact and requires a
substantial investment. Changing place is unlike changing product,
price or promotion a very drastic matter for a retailer in a competitive

We propose an optimisation and a simulation approach to advice
retailers on the most interesting location opportunities without losing
sight of the “raison d’être” of competition in a retailer market, the

First we address the uncertain competitive environment in which
location decisions have to be taken. We analytically formulate and
solve three optimisation models that take anticipation on the future
actions of new competitors into account. In line with wellknown game
theory strategies, we propose a worst case maximin location model, a
minimum regret location model and a von Stackelberg location model
for two players. Computational tests and an efficient procedure to deal
with large consumer demand populations in location models demonstrate
that these formulations provide interesting results for applications
in practice.

Next we focus on the increasingly segmented and complex consumer
markets that form the rapid evolving context in which retailers nowadays
operate. The “fishing where the fish are” rule of thumb is put into
perspective in a literature review. We discuss two streams of traditional
modelling of consumer spatial behaviour and then apply a general
framework for spatial models to the particular context of consumer
behaviour. We move to more recent modelling approaches and elaborate
on the recent evolutions in the modelling of consumer spatial

Particularly the aspect of increased mobility draws our attention and
in our survey in the city of Ghent we zoom in on the shopping behaviour
of consumers during commuting trips. Clearly, going where your
consumers are does not fully grasp the complex challenge that consumer
spatial behaviour modelling presents.

So we infer that two important concepts are revealed in our literature
study of consumer spatial behaviour: mobility and bounded rationality.
Next we link these recent evolutions in consumer spatial behaviour
to the changing retail location patterns with an agent-based approach
for location modelling.

Therefore we first introduce the methodology of agent-based
modelling and simulation and review related work in relevant domains.
Next we develop a conceptual model that encompasses mobility and
bounded rationality on the demand side and heterogeneity in the
assortment of products on the supply side.

We implement a prototype model in NetLogo. Our simulation results
show the interplay between the consumer spatial behaviour and retail
location decisions.

Finally, we extend the analytical and simulation approach with the
simultaneous handling of location and price decisions to more accurately
estimate future revenues. We successfully incorporate the mill
price decision in the maximal covering location problem for revenue
maximization in a competitive environment. The optimisation model
seems computationally intractable, but we propose an intelligent
enumeration technique that allows to solve the model to optimality in
a very efficient way.

Finally we return to the agent-based approach. Similarly as in the
optimisation model, we first introduce the price in the utility function
of the consumer agents. Next we also incorporate the price as a
decision element in the retailer agents.

We conclude that it is imperative to adopt a quantitative point of view
for taking profitable business decisions in a competitive retail context,
but we also recognise that another good location for a retailer is in the
minds and hearts of its best consumers.

PDF icon 20100929i.pdf