Measuring Success | Linking the (Big) Data Silos for Collective Impact
single,single-post,postid-16251,single-format-standard,ajax_fade,page_not_loaded,,wpb-js-composer js-comp-ver-4.3.2,vc_responsive

Linking the (Big) Data Silos for Collective Impact

18 Nov Linking the (Big) Data Silos for Collective Impact

By Sacha Litman

Funders are investing billions of dollars trying to move the needle on a variety of social issues. How do they know, though, if they are funding the right people, programs, and organizations to achieve their desired goals?

Take at-risk youth. CEOs of after-school programs and education non-profits ask me all the time how they can demonstrate to funders their impact on kids’ graduation rates from high school when they have little control over (or knowledge of) what hap- pens to the kids after, say, 7th grade. Demonstrating impact is very difficult, especially if an after-school provider cannot obtain graduation information from a student’s alma mater. Right now, that information enters a black hole. Even if the program could get its hands on the data, there would be insufficient evidence to prove that its own program, which occurred 5 years prior to graduation, made the difference among the dozens of interventions.

The solution to this problem lies in linking the silos of data main- tained by each program. In the words of the Gates Foundation, “It is the combining, linking, and connecting of different ‘data islands’ that turns data into knowledge – knowledge that can ultimately help create positive change in our world.”

In six communities across North America, we are helping pioneer an approach that takes collective impact to a new level. In Dallas, for instance, we are working with a diverse set of education program providers—public schools, after school programs, preschool programs, social service agencies, and non-profits—to improve outcomes for at-risk youth. After gaining permission from families and ensuring strict compliance with FERPA, HIPAA, and other privacy measures, we are building a meta-data sys- tem that links the information that each organization stores about a child and his/her family. This enables us to compile a longi- tudinal data-store of information—that shows inputs and patterns over time—to figure out which interventions are working and which are not at the macro level.

The real-time analytics and reporting engines allow funders and program providers to measure impact in the near and long term. Interestingly, we are finding out more than just “this program works, this program does not.” Rather, we have identified which interventions, or which sequence of interventions, work for which type of child. We are getting to the point now with our efforts that we can use predictive analytics to identify that a youth with slipping grades in school, whose mother is being treated for drug abuse, would be best supported by a specific intervention (for example, a mentoring program). As the old adage goes: different strokes for different folks. After all, a doctor does not prescribe the same medication to each patient! A good doctor reviews your medical history and performs diagnostic tests, and on that basis, recommends a certain course of action. (more about e-health records in another article!).

Critics say that this approach of linking the silos of data interferes with individual privacy. Yet we’ve found that this is a red her- ring because the whole system relies on permission from parents. Families have overwhelmingly appreciated this holistic ap- proach to their children’s wellbeing and have signed over waivers to that effect; those who are uncomfortable can have their data removed from the system (in practice, fewer than 5% fall into this bucket). Also, the data is held confidentially. Foundation reports are presented from an aggregate viewpoint (so that no funder or provider can see the results for an individual family, unless permission is granted).

The real challenge, we’ve found, comes when program providers refuse to share their data due to territoriality or the fear that their programs will not show impact. These are understandable concerns in a competitive world where performance matters.

The solution there, however, is to remind program providers that transparency breeds confidence; openness and collaboration allows programs to perform better in the long run, even if it exposes some dirty laundry along the way.

As funders, we know that we have to shift the focus from program-centric to person-centric. If linking silos of data can help Netflix provide subscribers with personalized movie choices, Amazon to encourage customers to buy more books, and the Obama campaign to micro target voters, do we not believe that our efforts to help those in need deserve the same benefit? Using these tools of data aggregation and predictive analytics, we can prove that yes, funders can move the needle.

Sacha Litman is Managing Director and CEO of Measuring Success in Washington, DC.