As with European experiments, in various regions in France, territorial intelligence projects have been initiated since 2003. (see the regions of Lower Normandy, Lorraine, Réunion Island, the Aquitaine region, etc.).The objective of these is to gather and exploit information which is not confined to particular sectors and the collective processing of which can contribute to durable development. Apart from institutions, civil society and the inhabitants of the territory, it is observed that companies and in particular small and medium sized enterprises are natural partners who show interest in such initiatives. Both the different economic chains and the participating organizations thus derive considerable benefit in terms of the anticipation of threats and in the reaffirmation of the territory as a common resource worth defending. Above and beyond the information processing systems operating within these organizations or economic chains, the articulation of internal actions to generate informational capital in terms of local territorial intelligence, produces a leverage effect with visibility of European or even worldwide visibility (Herbaux, 2007)52. Nonetheless these experiments lead to widely differing results, of which the progressive abandonment of the project by the companies involved is one of the most commonly observed. To support a theoretical contribution as a thread for this communication, we report on the results of a Delphi type survey completed in 2006 and covering 53 companies in the Nord- Pas de Calais region involved in a process of territorial intelligence since 2003. This revealed that 43 companies out of the 53 concerned had not followed through on their internal information sharing project and contented themselves, by default, with the results by economic sector derived from public regional surveillance Beyond this apparent disengagement from the process initiated, we may be curious about this apparent discretion of a group of actors concerning local government. This work nonetheless did generate a consensus around certain observations among the actors questioned, particularly as regards an initiative for which they did not deny the final utility but for which the requirements necessitated a significant modification to their internal culture. After the initial conventional responses: "security of patrimonial data, new choices in investment of time, lack of means, different priorities, etc.", repeated and differentiated questioning of those concerned revealed that the progressive abandonment of these practices and commitments bore a relationship with a number of human factors of relational and cognitive nature, thus depriving the project of its founding principles. This observation echoed that of the implication suggested by Girardot en 200553 on the theme of multi-level governance. Although the financial aspect is a factor in the long term survival of regular investments of man-hours, this criterion appeared progressively more marginal to the general project among the actors surveyed, as against several positions cited as prerequisites. Based on a synthesis of the results of the study, we propose five key success factors to promote within organizations to promote the logic of information sharing. To this effect, our proposal for a model named in French "CADIE" (Communication, Appui, Durée, Implication, Ecoute - or, in English, Communication, Support, Duration, Attentiveness) suggests several attitudes to which organizations must adhere to develop long-term integration in a territorial intelligence network. The limitations of our proposal arise from the small size of the sample at our disposal and the regional limits of our data gathering. This experimentation, duplicated in various European regions would benefit from a multi-cultural gloss and thus would provide the template for a preliminary European approach to the logic of territorial intelligence.