Abstract
In the design of a new urban retail network, the size of each office can be determined — once the number and the location of outlets have been fixed — by means of a location-allocation model. In order to carry out this task, two different solutions have been considered: the best solution in the opinion of the firm's managers and the solution obtained by maximizing the outlets' accessibility, based on a spatial interaction model. Our biobjective program bridges the gap between both solutions by enabling the generation of a finite set of non-inferior points, and constitutes, therefore, a valuable decision-support tool. The paper closes with a case study in the banking sector.