When foundations talk about intermediaries they are referring to organizations that do re-granting for them or that provide technical support to large numbers of direct service nonprofits. Both of these are services - run by middlemen organizations that can serve a lot of nonprofits either by region or topic.
Its time to start thinking about intermediaries that provide products, not services. The MIT Poverty Action Lab is - in my opinion - the gold standard in this regard. The lab conducts third-party research on poverty alleviation initiatives, culls through the data for generalizable findings, and makes them available to others working to end poverty in other places.
In addition to the MIT Poverty Action Lab another Expert Intermediary is in the formative stage. This is a new effort in international development, called 3IE - The International Initiative for Impact Evaluation. 3IE will coordinate work from foundations (Gates, Hewlett, Google, UN), national governments (Mexico, Uganda, Britain, Netherlands, Canada) and multinational NGOs including CARE and the International Rescue Committee.
The tricks to doing this kind of work successfully, are many. Here are three to think about.
First, we need to counter the "not invented here" mentality in philanthropy. Using strategies or data or interventions or tools that others have created is key to ever having a noticeable impact. One way to do this might be to try to celebrate what might be called (at least by Apple fans) the 'Microsoft' approach - don't be first, be bigger. Take proven ideas, repackage them, and move them broadly.
Second, somehow we have to bring credible data to people in a way that they can use them when thinking about philanthropic strategy. Our first step in doing this should be to recognize the complementary roles that passion and pragmatism play in philanthropy - donors do want data, do use data, and do make rational decisions. And they apply these to issues and causes that they care about. Using data to inform passion and direct strategy is what we should be aiming for - not redirecting passions and interest, but driving them forward with good information.
Third, large funders are important investors in these efforts, and they need to think about the downstream funders of the work they intend to improve. Individuals are a significant part of the revenue streams for civic sector work in poverty and on other issues. Bringing this kind of information to individual donors in the form, depth, and frequency that they can use, is a strategy that could strengthen, align, and improve the effectiveness of billions of dollars in philanthropic support and remittances.