South African researchers are used to being told to 'bridge the gap between science and policy-making' and to do research that will lead to 'evidence-based policies'. A lot of research grants require projects to be useful in some way. There is also a burgeoning literature on the theory and practice of doing, communicating and ensuring that research is useful. ACHWRP is an example of how HEARD has worked with these sort of demands and, we like to think, learned valuable lessons along the way.

Facilitating a child welfare management plan in Amajuba district has involved much more than the standard activities of 'stakeholder consultation', 'focus group discussions' and 'dissemination workshops' ; indeed, what these activities mean is much more than what is usually inferred. All are necessary things to do and have been done by the ACHWRP team. However, there is no straightforward link between producing research results and using them to influence government and NGO child welfare plans and programmes.

HEARD has learned that the work is less about 'community outreach' and more about 'reaching in' to the human disposition to be constructive; to do good, if you will. There is, of course, a discernable line beginning with research, leading to communicating the 'evidence' and, finally, assisting a community, be it a government for example or a group of a group of NGOs, to use the results for practical purposes. That path can even be summarised as research, advocacy, marketing and technical support. What these terms hide is that the process is also about learning (research); about building trust, credibility and awareness (advocacy); about building knowledge, conviction, ability and, in turn, motivation and opportunity (marketing); as well as being about securing commitment, checking that a practical plan is working and training people to implement a plan (technical support).

HEARD's ACHWRP team has also learned that research projects can and should morph from research into advocacy and support and back again; in a phrase, be flexible. For instance, research and consultation in the past has left the team with a database of over 400 child welfare agencies. That has drawn the attention, most recently, of a steel company and the Department of Labour for advice on where and how, respectively, to start corporate responsibility projects and to train 15-20 community organisations. These are just two developments amongst many others that took the team from advocacy about improving child welfare in the district, to creating infrastructure for things to happen. Inevitably, there is growing interest locally in learning whether and how child welfare interventions are working and so there is more research work.