All four analysed innovations were born as initiatives of TSOs. In all cases there has been one (or several) TSOs that wished to do something in a specific new way that have either started to do it on their own and then found the way to obtain public support (cases 1, 2 and 4), or have convinced public partners to start the project (case 3). In all three cases, public administration (local or regional) has had a significant role in supporting and sustaining the initiative, in some cases expanding it beyond its initial scope.

All four cases show an attempt to change the ways of working and addressing users. In all cases, the idea of customising or personalising services to specific needs or preferences of service users is present. In two cases (1 and 4) this adaptation goes further to redefine the boundaries between users and practitioners, be it by blurring the distinction (case 4) or by a more profound redefinition of the nature of service users (case 1).

Most cases show some degree of changes in governance and the relationship between TSOs and local/regional government. These relationships have a long tradition of particularistic, case-by-case selective discretionary support to the activities of each TSO. Cases 1 and 2 show a move towards a system of (a) clearer definition of what activities may be subsidised and with which criteria and (b) a negotiated procedure for setting goals and overall criteria for the programmes. This has not eliminated completely a certain degree of discretion, but it constrains it into a more objective system.

There is a degree of political ambivalence in all cases. On the one hand, there is a clear “political” aim by the TSOs to “work differently” in ways they consider to be better and more adapted to service users’ needs. In some cases these alternative ways of working are very different to the dominant views in public services (cases 1 and 4), in others they are just different from the mainstream programmes. On the other hand, governments may accept such different or challenging approaches because they help to offer a limited exception for some users (with whom they are not being very successful) or they enhance the view of a more complex welfare mix in which TSOs have stronger roles. This may explain the (apparent) lack of political controversy in some cases (cases 1 and 2) or a very limited one (case 3). Case 4 is somewhat atypical, since it has seen stronger controversy, although external political reasons (connections with radical groups and with Basque nationalism) may be crucial in this case, and some political positions have been contradictory (the project got its public support thanks to the only political party that has had a critical stand afterwards).