An indicator must be defined according to at least three points of view. It must be a parameter of a model according to which the objective of the decision can be represented; it must be an argument of a decision function; and it must be estimable through a function of available data. The whole set of estimates of the parameters of a model is a multidimensional statistic, i.e. a set of functions of available data. This statistic is often seen as a list of indicators; it can also be considered as a set for indication that may replace the original data set for any further purpose of indication. Using as a case study the Senegalese artisanal fishery, an illustration is given with several control variables in the context of a multicriteria objective dealing with biomass levels, incomes and nominal activity. These results may be useful for decision making if they are considered as a step in the context of dialogue(s) between actors involved in the fisheries system.