FlockData at Collaborate

CollaborateFriday’s sessions at the Washington D.C. Collaborate conference involving government employees from many levels all revolved around one notion: the current speed of government operations is too slow to accommodate the quick and forever changing tech landscape.

How do we solve that? Many presentations on Friday either danced around or directly asked this question. Xavier Hughes, the Chief Innovation Officer at the U.S. Department of Labor, wondered if the current model of hiring out large companies to fulfill large contracts that generate large sums of money didn’t quite fit this 21st century world.

As the country’s largest employer, the federal government would struggle to mimic the nimble flexibility with which many tech startups operate. However, some of that gap can be bridged by eliciting feedback, support, and even tools from the startup IT community. By holding several National Data Contests, the Department of Labor, according to Hughes, has been able to better build their IT infrastructure to suit the times.

But even these are smaller steps than are perhaps necessary to adapt to the private sector’s standards. David Bray, CIO of the FCC, lamented how much aversion to risk exists in the public sector. To him, it makes a degree of sense, any mistake or failure from a government official is scrutinized by media, both social and otherwise, to such an extent that moving as though one were treading through treacle is preferable to any risky operation.

Hughes echoed this notion, “If I release something from the government that isn’t ready just to generate feedback and people hate it, I might as well just hand in my letter of resignation.”

(Incidentally, Bray discussed the development of applications in this context, where the data all exists but it’s difficult to corral it. FlockData can help with that.)

This narrative of being afraid to fail permeated the Friday sessions. Which is what made Keynote Speaker Tom Chi’s talk all the more refreshing.

Chi specializes in helping large organizations realize their full potential by undergoing an intensive two-day exercise where chief executives, builders, and strategists rethink their entire process.

A common problem, according to Chi, is that executives will gather a meeting and try to build an entire product on paper before putting it into development. The result will be something that “100 percent optimizes 25 percent of the problem.” However, if instead you come up with a testable hypothesis, a small portion that you can build right away and test, you end up definitively answering questions a lot faster.

To drive home this concept, Chi discussed an impoverished town in the Mexican state of Aguascalientes. Three teams traveled there in an effort to build jobs in the town, one coordinating and two in the field. Giving themselves approximately 30 minutes for each stage of the process, they: Found an expert carpenter, developed a business around that carpenter, recruited people potentially willing to fill out that business as apprentices, discovered those people were more willing to be mechanics, found a mechanic, and found cars for the mechanic to teach the willing. In a matter of three hours, eight business ideas were created and improved on, and the town’s prospects were as bright as the state’s water was hot.

Failure was central to Chi’s speech. It’s a cliché that one learns much more from their failures than their successes, but when put into practical use, as Chi’s group had done, the results are inarguable. A good way of thinking about it is if you can fail a bunch of times in the preliminary stages of development, you can avoid failing spectacularly at product release.

From the impressions given by Hughes and Bray, the government thinks too big. The key might be to answer a bunch of small questions to discover what the big question should even be.

By the way, this is very similar to FlockData’s approach to data profiling and integration. You have a bunch of data and an idea on what to do with it. But by profiling it and putting it all in one place, FlockData essentially answers a bunch of small questions that lead you to a bigger question.

Look for us in the Exposition Hall at Collaborate on Saturday to learn more about how FlockData can help you answer all those little questions.