Data dividends

Have you heard of the design company Threadless? You may know it as a t-shirt company. Or the former home of the Obama campaign's current CTO, Harper Reed. Or you may never have heard of it. But if you work in foundations as a program officer, grants manager, communications director, evaluator or other executive, I think you should take a look.

Threadless is a company that prints designs (primarily on t-shirts and iPhone cases). The community suggests the designs. The community votes on the designs. The community buys the products. The products sell very well. The process repeats. Buyers, designers, voters - some people play all these roles, some play only one or two. But they're all part of the community.

Why should foundations pay attention to this? How many of you (raise your hand) have found a cool idea, funded a successful project, evaluated something, or photographed a great event and then said "we should share this information?" So you publish something - maybe a pdf or a blog post. (Maybe you tweet about it or you share it on scribd, tumblr, scoop.it, or RebelMouse if you are really cutting edge)

And then you wonder, where did it go? Did anyone read it? What happened?

I had a conversation yesterday with two foundation executives about this problem. Their organization had funded three different projects over several years to the tune of many millions of dollars. They had information about what worked, a good handle on the communities that might be interested in this information, even funding partners interested in helping get the info to those communities. What they didn't know was "What did the communities want?" Threadless (and Kickstarter and IndieGoGo and others) never has this problem. It knows what the community wants. It only builds (prints) what the community wants. And it then provides it in a variety of forms that the community has already said it will use.

Why is this so important? Foundations have limited resources. They should use them really well. They typically spend a lot of time and talent making informed funding decisions. They take care with their money. I think they should think about the information sharing possibilities just as carefully and with the same expectation of some outcome. In fact, I think the careful building of networks and sharing of information can be seen as a "data dividend," from which a funder should be looking for some kind of return just as they expect some kind of result from their dollars. For $X million the foundation funded several demonstration projects. For a data dividend of $Y (where Y is going to be significantly less than X), they can spark, inform, catalyze, accelerate more of the kind of things they funded with X, if they do it right. If they get the right information to the right people at the right time in the right form.

And if they do it right, the math is really exciting (because of the economics of information). It may cost $X to get a few demo programs running or to build a few prototypes or to conduct an evaluation. If the process of gathering the information and the plan for distributing it after the funding is built in from the start and the information is designed to 1) help someone get something done and 2) take advantage of in-person and online sharing systems then the return on the $Y data dividend could be even higher. Suppose you spend $3 million on a school program and $250,000 documenting, learning from, and building a community of two dozen other school professionals who have said they want to learn from that original program. Where $3m gets one school program the $250K might spark change in two dozen others. And the nature of (well structured) information is that it can keep being used, reused, shared, re-applied. Think of it this way - well structured data and networks for sharing it extend the value of the grant dollars - you get more bang for the buck. That's the data dividend.

How can foundations do this better? They need to ask their potential customers what they want before they start paying to evaluate, document, convene or create toolkits. The network of potential "info users" knows far better than the foundation does what it can use to learn. Is it a case study? A seminar? A site visit? Online access to an open data set? Professional development and worksheets? All of the above? None of the above?  A foundation's network of arts organizations and artists might want videos and site visits and it's partners in school reform might want curriculum for existing professional development networks. Different networks will use information differently - which is one of the reasons designing digital information upfront is so important, it gives you the most flexibility down the road.

The Threadless community has made it clear over time - they buy t-shirts and iPhone cases. Threadless knows the format in which to provide the information, and it lets the community then suggest and promote the actual designs. Threadless knows it has buyers for its designs from before they print anything, because the buyers tell them (Again, this is the same principle at work with Kickstarter). If foundations planned for their data dividend from the beginning of a grant they would get better and better at knowing which communities use what form of information. They'd get better at knowing what information those communities need. They'd involve that "end user" community early in the conversation so they can get the most bang for the "publication buck" and not spend a lot of time and money documenting something that no one can find, use, or care about.

Foundations with communications officers, evaluation experts, program officers and grants managers have all the talent they need to work toward "data dividends" but they need to re-allocate the way those folks spend their time and blur the boundaries between who does what when. Grants managers can be vital partners in helping program officers map the landscape of organizations working on an issue. They can help identify the potential community of users for foundation data. Communications professionals can (and many are) already at work trying to connect the foundation to the right networks to accelerate the institutions mission - here's a great "tweet" example of one step in that network building:

The key to this is building the network first, then using it to use information. Foundations have a great source of information to build these networks - the proposals and inquiries they get from people in communities trying to make the world a better place.

Evaluators will have (some of) the data that will be useful to the end users, and need to be part of the early conversations so they can help both the funded (and evaluated) programs and format the information so it can be useful to that broader network. And program officers setting strategy and selecting grantees can also be thinking about, from the beginning, what happens with what gets learned from this grant? What happens with the data it produces and how can I get the greatest dividend on it?

In this way each grant is not just part of a program portfolio but it's also a step toward a network of informed peer organizations, sharing information and working toward shared goals. What is learned from and produced from each grant should be available for others to put to use - this is what's exciting about data philanthropy. Given the economics of "one-off" grants compared to multi-use data and information, the potential benefits of working this way are worth the costs of change. A foundation doesn't need to have all of those staff people I just described (it might be easier if you don't) but it does need to have a mindset of working in networks and using both its money and the data that flow through it to achieve its mission.