If you've got two million employees, asking all of them for suggestions about how to do things better can be a little bit like trying to drink from a firehose. And while that might be intimidating for even the most high-tech of companies to consider, the challenge gets tougher if the organization you're trying to improve is the Federal Government. So we're excited to take a behind-the-scenes look at the President's SAVE Award initiative. This initiative is about soliciting ideas from federal employees for how to save taxpayer dollars, but there are some larger lessons from SAVE that can show how broad input from federal employees and the public can make government run better.
SAVE's mandate is simple: Find ways to save taxpayer dollars, and make the government run more efficiently and effectively. But having a free-for-all where any one of the millions of Federal employees could throw their ideas up on a web page wouldn't just be a mess, it'd reduce the chances of success. So, what the Office of Management and Budget (OMB) did was create a system that works a little bit like American Idol, if that popular TV show were a little more focused on the "American" part, and a little less focused on the "Idol".
First, the idea of the President's SAVE Award was announced to nearly all Federal employees, using technologies that were already in place, like email announcement newsletters within individual departments and agencies. This might seem obvious, but it's a pretty important bit of learning when we're thinking about how to make crowdsourcing technology: The community you're trying to reach might already exist. Just because we tend to be talking about Web 2.0 and social networking, it can be easy to overlook traditional connecting technologies like email lists. This broad announcement acted like an open call, casting a wider net of potential contributors.
Then, once the idea was announced, idea submissions were captured from employees at the agency or department level. This serves a few useful purposes:
In all, this process yielded nearly forty thousand idea submissions, from across the country and from all levels of the government. We talked to an OMB official about what it was like to process all of these submissions, and one of the most interesting insights was discovering that some of the ideas that were submitted could actually turn into government-wide improvements, even simple ones like making sure every government building allows its employees to turn the lights off at night when they go home. So OMB held on to some of those ideas that were best implemented across the entire government, and then processed the majority of submissions through a full review, to help whittle that list down to the Final Four for everyone to vote on.
The key thing to mention here is that technology did not make the decisions about which ideas to promote: Humans did. What technology can do is provide platforms to quickly filter ideas, group similar or redundant ones into categories where they can be easily processed, and get lists of responses down to manageable size so that humans can efficiently review them. A lot of times, crowdsourcing is misunderstood as letting technology make decisions about ideas, but what the SAVE Award has shown is that having technology serve to amplify the good judgment of motivated people is actually quite effective. And while we'd like to imagine the human vetting process for these ideas sounded a bit like the judges on American Idol, it's likely it was probably a little more dignified an affair.
Once the broad pool of ideas was narrowed down, the White House stepped in to lend its considerable voice to helping promote these finalists, and to encourage the public at large to vote on which ideas resonated best. Macon Phillips, the Director of New Media at the White House, gave us some insights into how his team helped to encourage the public to build on OMB's efforts.
First, and most obvious, was using the President's voice to promote the idea of increasing government efficiency. This isn't just a pragmatic move to draw attention to the award, but a good sign to federal employees that this initiative is being taken seriously. It can be easy to forget, if you're outside the government, that the President is in some ways the CEO of a giant corporation with 2 million employees, and if your boss's boss's boss says "This Is Important", that's a pretty good sign that it is. So clearly, the SAVE Award was being taken seriously, and the idea of highlighting and rewarding federal employees for their good ideas gets a boost from that message.
But just as importantly, the public at large wants a voice in this kind of effort. And the White House web team was able to create a simple voting system for picking which idea to reward. From a technological implementation standpoint, the voting system is minimal: The four ideas are presented in random order in slideshow format, and a range of scores from 1 to 5 is presented underneath the idea for voting. The descriptions of the ideas are brief but clear and voting on them all takes less than a minute or two.
Voting's still open, so we haven't yet seen who's going to win, but the winner will get to meet the President, and their idea will be incorporated into the Federal Budget for fiscal year 2011. And that brings us to perhaps the most important motivator for a crowdsourcing initiative: Knowing that your submission can have an impact.
Again, this has nothing to do with internet technology or cool social networking sites, but the biggest motivator for people to contribute to a collaborative effort can be the combination of recognition and the knowledge that input will be taken seriously and have a real impact. In this case, once we see a crowdsourced idea bubble its way up from a pool of two million employees to the final four SAVE Award nominees to eventually being part of the federal budget, we'll have a remarkable example to learn from.
We'll be watching closely at Expert Labs to see how the SAVE Award succeeds, and we're excited to have examples like this to learn from and report on as we begin our own work on making it easier to tap into collective expertise.
First, the idea of the President's SAVE Award was announced to nearly all Federal employees, using technologies that were already in place, like email announcement newsletters within individual departments and agencies. This might seem obvious, but it's a pretty important bit of learning when we're thinking about how to make crowdsourcing technology: The community you're trying to reach might already exist. Just because we tend to be talking about Web 2.0 and social networking, it can be easy to overlook traditional connecting technologies like email lists. This broad announcement acted like an open call, casting a wider net of potential contributors.
Then, once the idea was announced, idea submissions were captured from employees at the agency or department level. This serves a few useful purposes:
- Ideas could be vetted by OMB experts on those agencies, where there's the knowledge and context to judge if the idea submission was feasible, practical, and legal.
- A lot of the best ideas might be submitted multiple times. By having a review process at the agency level, a lot of redundant ideas could be filtered out.
- We're always focused at Expert Labs on how to connect with people who have the right expertise — in the case of many government agencies, they already have the expertise in-house, but just need help discovering it and amplifying the ideas of those experts.
In all, this process yielded nearly forty thousand idea submissions, from across the country and from all levels of the government. We talked to an OMB official about what it was like to process all of these submissions, and one of the most interesting insights was discovering that some of the ideas that were submitted could actually turn into government-wide improvements, even simple ones like making sure every government building allows its employees to turn the lights off at night when they go home. So OMB held on to some of those ideas that were best implemented across the entire government, and then processed the majority of submissions through a full review, to help whittle that list down to the Final Four for everyone to vote on.
The key thing to mention here is that technology did not make the decisions about which ideas to promote: Humans did. What technology can do is provide platforms to quickly filter ideas, group similar or redundant ones into categories where they can be easily processed, and get lists of responses down to manageable size so that humans can efficiently review them. A lot of times, crowdsourcing is misunderstood as letting technology make decisions about ideas, but what the SAVE Award has shown is that having technology serve to amplify the good judgment of motivated people is actually quite effective. And while we'd like to imagine the human vetting process for these ideas sounded a bit like the judges on American Idol, it's likely it was probably a little more dignified an affair.
Once the broad pool of ideas was narrowed down, the White House stepped in to lend its considerable voice to helping promote these finalists, and to encourage the public at large to vote on which ideas resonated best. Macon Phillips, the Director of New Media at the White House, gave us some insights into how his team helped to encourage the public to build on OMB's efforts.
First, and most obvious, was using the President's voice to promote the idea of increasing government efficiency. This isn't just a pragmatic move to draw attention to the award, but a good sign to federal employees that this initiative is being taken seriously. It can be easy to forget, if you're outside the government, that the President is in some ways the CEO of a giant corporation with 2 million employees, and if your boss's boss's boss says "This Is Important", that's a pretty good sign that it is. So clearly, the SAVE Award was being taken seriously, and the idea of highlighting and rewarding federal employees for their good ideas gets a boost from that message.
Voting's still open, so we haven't yet seen who's going to win, but the winner will get to meet the President, and their idea will be incorporated into the Federal Budget for fiscal year 2011. And that brings us to perhaps the most important motivator for a crowdsourcing initiative: Knowing that your submission can have an impact.
Again, this has nothing to do with internet technology or cool social networking sites, but the biggest motivator for people to contribute to a collaborative effort can be the combination of recognition and the knowledge that input will be taken seriously and have a real impact. In this case, once we see a crowdsourced idea bubble its way up from a pool of two million employees to the final four SAVE Award nominees to eventually being part of the federal budget, we'll have a remarkable example to learn from.
We'll be watching closely at Expert Labs to see how the SAVE Award succeeds, and we're excited to have examples like this to learn from and report on as we begin our own work on making it easier to tap into collective expertise.
Fantastic article and very inspiring. Please message me any other experiments of mass-crowdsourcing going on out there. I wish to learn more.
Posted by: JWilfong | 01/02/2010 at 12:56 AM
Speaking as a former federal employee, here are my thoughts/questions about the SAVE Initiative:
If 38,000 suggestions were submitted by federal employees, but only four (4) were selected for voting by the public, then what happens to all the rest?
That is, if the SAVE Initiative is being "taken seriously", then what happens now to the remaining 37,996 ideas?
Granted, OMB may use "some" of them later on, but if federal employees see that over 99% of their submitted ideas go into a black hole, then what message does that send?
And, just for the sake of discussion, let's say that "only" 1,000 ideas (out of 38,000) had real potential for substantial savings.
So, then, what is the purpose of figuring out which one is the "BEST idea" when ALL of those "top-1,000" ideas can save money?
Does there have to be a "winner" (a la American Idol) as if this were some type of competition? No, there does not unless, of course, this is trying to be marketed to the public in an simplistic, entertaining way.
Many federal employees will recognize that aspect of the SAVE Initiative, and when they are asked for more ideas, in the future, the number will be much, much LESS than 38,000.
ALSO: I'm wondering whether the online system allowed federal employees the option of remaining anonymous. It can be very dangerous to one's career to point out inefficiencies in one's office. Many bosses tend to look at that as open criticism of their management.
Posted by: Transpartisan | 01/20/2010 at 12:06 PM
Dear Anil,
I always wince when I hear people talk about how a blog is a great way to have a "conversation".
And it could be that you may respond to 99% of the comments left on your blog-postings.
But, all I can gather from the lack of a response to my comment posted almost a years ago, is that, like most bloggers, what YOU want to say is more important than the response from others.
I sincerely hope that this is not the case here, but that is up to you to clear up.
vr,
Stephen Buckley
Posted by: Transpartisan | 11/29/2010 at 02:41 PM
Hi Stephen,
Actually I'm very interested in two-way conversations -- but sometimes the tools make it inefficient to carry those on, as in this case your response slipped through the cracks. Apologies for that.
To answer your questions as best I can (keep in mind, Expert Labs isn't part of the government so we're just fellow citizens when it comes to insights into how these programs work):
* It is a shame that unused ideas go into a black hole. Ideally, all submissions would be visible, and even more optimally, the criteria used for filtering them would be public. The tools we make at Expert Labs, like ThinkUp, are designed with those goals in mind, but I'd expect that the SAVE program's efforts were so early that some of these ideals we'd like to see were omitted in order to get the project launched at all.
* Picking a few ideas out of many to operate on is exactly the sort of prioritization we *want* our federal agencies to be doing. The issue for any agency isn't "are there lots of worthy projects to tackle?", because the answer to that is always an emphatic yes. Rather, we want to make visible the decision-making process by which these agencies rank and sort those priorities. In this way, there will be accountability, because frankly there's never going to be enough resources to say yes to 1000 great ideas all at once.
* Finally, I think your concerns about privacy/career impacts of sharing ideas points out the complexities in your initial request that all submissions be made public. Focusing exclusively on ideas that can be publicly shared will necessarily skew some of the feedback, and will have a censorious impact on the kinds of ideas suggested. Perhaps a complementary campaign to SAVE could be done with the explicit expectation of all submissions being non-anonymous and public, to see if that elicits different types of suggestions.
Again, my apologies for the lack of response -- sometimes the intention of conversation is overcome by the constraints of time and technology. And, though I share many of your criticisms of SAVE, I think this situation is parallel in some ways; Sometimes we have to be happy with good-faith, but imperfect efforts, instead of demanding the perfect ideal solution in a situation, because making progress towards the ideal is the most important first step.
Posted by: Anil Dash | 11/29/2010 at 03:30 PM