Area-based needs-assessment methods tend to reinforce the status quo in service or program delivery rather than challenge it. Data collection easily becomes routine and self-justifying. To rid ourselves of ritualistic methods of data collection, we must turn attention to the process of change that should result from it. That is, we must think in advance of the types of decisions that must made about service delivery and the nature of the change process itself. There are two routes to this end, depending on how change is conceptualised. If one adopts a conflict model of change, then felt need or, more crudely, what consumers or groups and communities want becomes the primary focus. This would place consumers in the chief role within an action-oriented advocacy process. Alternatively, if a more functionalist model of change is adopted, then the focus of activity is on the information needs of managers and the type of data they require to justify a resource shift. Typically these data are not about "needs' but marginal health gains and costs. In either event, the data collection procedures should be far more specific and directed than the existing broad-brush procedures that presently serve only a limited purpose and effectively dilute or detract from both agendas for change. There are several possible pathways to resolve some of these issues, in particular, using hypothetical results to foresee if or how change could result from any proposed data collection activity.