Last week, we announced that we were going to bid on Recovery.gov. Here, 9 days later we have a few pages of information on our bid, but I don’t think it makes sense to turn anything in tomorrow. So I’m declaring this experiment a failure. Most people confuse “failure” with being “wrong” but here in the Labs, we’re into experimentation — and you can’t experiment or push the ball forward if you’re afraid to fail.
This was a completely worthwhile experiment and we’ve learned a lot. I want to share some of the reasons why we failed, and what we’ve learned with all of you.
Why we failed
There is a language problem between us (developers) and Government.
This is the critical reason why we failed. Not only did I sign us up to write a sixty-page term paper, I signed us up to write a sixty-page term paper in a language we don’t speak in a country we’ve never been to. Before taking this on we didn’t know what CLINs were, what an ALLIANT GOVERNMENTWIDE contractor was or anything of the sort. We didn’t know the difference between a Statement of Objectives, a Statement of Work and an RFP. There’s a vast language disconnection between the Government (which speaks heavily in an enterprise dialect) and the open source community which speaks fluently in the dialect of simplicity, pragmatism and “straight forward-ness.” This is a great chasm between us and Government that must be crossed if we want Government to start adopting new technology and letting developers have more of an impact. It is a huge problem that effects every open door into the Government.
I didn’t manage the community well from the start
I didn’t know I wasn’t managing the community poorly, but I’ve learned a few lessons from this that I have to share. But what I did was put an RFP up, put a page on a wiki up, and expected stuff to happen. In hindsight, that was naive. I should have been a better organizer (and will talk more about that in a bit.)
We’re not qualified to do this.
At the heart of it, rebuilding Recovery.gov is complex and the turnaround time is six weeks. The contractor with the winning bid will have to work night and day for six weeks, and while a small team could likely get it done with some heroic effort on the cheap– Government expects two colocation facilities to be built-out in geographically separate areas for redundancy, not to mention a redesign, a fully blown content management system deployment, a reporting interface to create arbitrary reports based on the data and all kinds of other things. The Sunlight Labs team, even if all of us got together and pitched in to build this, would not be able to deliver something great. We’d be doing a disservice to America.
What we got out of it
While we did fail, we got a lot out of it and we’d do it again in a heartbeat. Here’s what we got:
We got contacts at 8 of the biggest government contracting firms
6 of whom want to work with us and teach us more about the contracting business. That’s a great offer. The small ones said to us things like: “this process needs to be opened up so that we can compete fairly” and the big ones said to us: “this process needs to be opened up so that we can compete on merit” and the end result for us is the same. At least some contractors are willing to talk about the process, how it works and how it can be refined. Clearly the process isn’t working well for anyone. An acknowledgment of that from some big players is a big deal.
Three of them said “we’re actively stealing ideas off your wiki” and I said “good.”
We learned how to organize better
In the future, when we do things like this, we need to be specific and direct with requests and probably focus more on organizing than on production. This is always a tough call for us, but in this case we jumped into writing the proposal and should have run around organizing a solution from ISVs that could have each contributed. Or at the very least made specific requests to individuals early on. Next time, we’ll do better at that.
We started learning the language
As I said before, this RFP was written in a different language. That language isn’t something we spoke fluently. We just took our first 101 class and have begun to learn the language and the nuance involved and we can start sharing it with you as we learn more. Hopefully this will make us capable of being more of service to our community here and serving as a gateway for good developers to start creating change from the inside.
We started shining the light on Contracting
I think we all knew light needed to be shined here and there’s much more that needs to be done. I think we’re going to see a more nuanced problem than “government contractors are evil” and “a culture of cronyism” that’s been the typical public perception, and instead start figuring out what’s really going on here. So far it looks like there’s significant issues with regulation, procurement, language and talent. But, you can bet we’ll be paying a lot more attention to the contracting business as we move forward, and likely FOIA all the bids that did make it in, so we can check them out.
The eventual winner may give us a seat at the table
We’ve talked with several of the contractors planning on bidding tomorrow. And while we’ve made no arrangements with any, some have joined us in our chat room on IRC, and many have added in their proposal that they will consult with us if they’re selected. I think that’s a huge victory for us.
We shed light on the Recovery Accountability and Transparency Board
We’re generally uncomfortable with the recent decisions by the RATB. From their decision to fast-track this effort into a six-week process, to having text that has been copied and pasted from one of the Alliant Vendor’s website, to not requiring bulk data access we are starting to question whether the RATB is taking Recovery.gov in the right direction. This is more of the same-old-stuff that really needs to change for Recovery.gov to be meaningful to its citizens.
We are learning how to fail
Like I said earlier, we’re proud to fail. Too often, especially in this community, people wring their hands and try and figure out what’s possible before trying something. While some may use the cliche of “don’t let perfect be the enemy of good, ” I think it may be more appropriate to say: “don’t let what’s possible get in the way of doing something.”
In politics, people argue both sides of just about anything. And while this is a great and wonderful skill, this kind of work trickles out into a community that can also see a side to everything. My take: pick a side, take a first step, see what happens. That’s why we’re called Sunlight Labs, not Sunlight Technology Services Corporation.
So, what’s next? For one, I want to get some of these Alliant contractors who by the way, have been really friendly thus far, together to chat about transparency from their side. To see how they can educate us about this business of contracting, and maybe teach us how to speak the language a bit better.
Secondly, Sunlight may want to think about the most ethical and transparent way to engage in this type of business going forward. While it would be neat to serve as a disruptive market force in government contracting, we’re also a government watchdog group. That kind of conflict could hurt our ability to serve the public. But there’s got to be a way for us to both be of service to you all and continue the great work we’re doing on the outside. We need to figure that out.
So to all of you who helped out– thanks for participating. This was a great experiment. Don’t be disappointed because we’re not going to get the bid: we did a great job here learning more about this field and it is only going to make us more effective down the road. Thank you so much for your participation in this project — your advice and suggestions meant a lot. In time, I think your contributions to this project will pay off in ways we won’t even expect.