Will the future of operational DSM be to endlessly refine highly centralized soil database (e.g. Global, continental or national) managed by a small group of happy fews? To my opinion, this scenario is neither philosophically acceptable nor efficient. Digital Soil mapping must go down progressively at more local levels which will ensure an easier access to the input data (including the local knowledges on soil variations), a better proximity to the user needs and, finally, a better appropriation of the produced maps by the local people. The challenge is to make emerge a viable DSM activity populated by DSM-trained soil surveyors that could answer with agility to the needs in soil information expressed by the users and formalized through local specifications compatible with the higher levels ones. The methodological limitations to this scenario are not the greatest ones. We must increase our effort in capacity building and also act for overcoming the legal and economical obstacles that could hamper the growing of such DSM activity. Developing local DSM is far to be antagonist to the development of national and Global soil databases. A fruitful collaboration including data exchanges can be set between the different levels, each of them having its own utility for solving soil-related problems.