A few months ago, we wrote about (and signed onto) the "Integration Proclamation" calling for data standards to promote interoperability between tools in the nonprofit space. There's been a lot of conversation since then, on lists and off, about what that standard should look like: XML? vCard? Something new? What about the fields and the data structure? Unfortunately, in the real development of standards, these questions are so secondary that they're almost beside the point. These are discussions about data formats -- particular schemes of organizing information. But data formats are easy. The hard part about standards is the collective adoption of one format in particular. Standards don't result from clever programming or even adroit diplomacy. They result from profit-sustaining incentives that drive users and developers to the same format. XML and other formats are a dime a dozen -- we at DIA have worked with and created a bunch of them before. We've got our own XML format, for example, and we came up with another one with Green Media Toolshed and others at a conference three years ago, which was a combination of many existing XML formats to handle internationalization. It's pretty easy to write and modify XML formats -- many of which are good, adaptable, and/or comprehensive. For constituent data, extending the vCard format is always useful, because then it plugs directly in with existing mail servers. The hard bit is calling any such format a 'standard'. For a given format to be a standard, others have to use it or develop software against it. That, in turn, requires an incentive to do so -- which means that for a format to be a standard, there has to be a clear path to profitibility (either monetary or movement profit) for the software builders. Without the incentives, there's nothing. Ever wonder why all our phones have different power adapters, when clearly it would be better for the consumer if they were all the same? There's monetary incentive for companies to build and deploy their own. Same with printer ink. With standard-supporting incentives, on the other hand, you could have the worst format in the world (e.g., vCard) become a standard just because there is immediate benefit to using it (it works with existing mail servers). So the biggest question for establishing data 'standards' is -- regardless of the beneficiary format, how can we create that incentive? One path is have an organization no one can ignore deploy that standard -- say, the Democratic Party, or Library of Congress, or W3C. Then it is the standard. Other paths involve someone ponying up the development time and effort to write importers and exporters for many major and minor providers, so that it is essentially free to implement the standard, and there is no business advantage to building your own. That's what we need to figure out. What are those tangible incentive paths -- with real dollars, and real costs? Then, we've got a shot at a format that we can call a standard.