#rfpez

In addition to rethinking big data in government in an attempt to save billions of dollars through the Office of Management and Budget, OMB; I, along with two other Presidential Innovation Fellows, inherited “RFP-EZ”. This was a highly successful initiative from the round one fellows sponsored by the Small Business Administration, SBA, in partnership with a larger “White House Small Business Procurement Group” through OMB’s Office of Federal Procurement Policy, OFPP. A majority of our focus on the RFP-EZ project was IT procurement reform.

Many in and outside of government blame the true failure of HealthCare.gov on the IT procurement process; being on the inside, I certainly can’t argue that! There are huge issues with how government procures IT services; and, the more we looked into the problem, the larger that we realized the problem was. So large in fact that we started with renovating “RFP-EZ” and quickly evolved into institutionalizing a “Procurement as a Service” revolution throughout the federal government. This revolution is a big part of the President’s “Management Agenda for Government Innovation.”

Our “tour of duty” was only six-months long, that’s not a whole lot of time to revolutionize anything, but it’s long enough to “get shit done”. My understanding of the inter-workings of government was rather limited, to say the least. However, within the first few weeks I realized that there were two primary ways in which things like this get done, through policy and through loopholes. Policy change takes significantly longer than six-months and requires much more knowledge of “the system” than I could ever wish to obtain, so, the “loophole” became my best friend – find it, understand it, expose it, exploit it … make change. 

There are three types of companies out there: those that win federal contracts, those that don’t and those that don’t even bid. The first lesson we learned was that the best person at winning almost always wins -- that kind of makes sense, right? It’s actually a really big problem. For example, if FEMA needs an iPad app developed for real-time disaster recovery efforts on the ground, who is going to win that contract? The company whose core-competency is making iPad apps or the company whose core-competency is winning federal procurements? You don’t have to work inside government to know the answer to that question; those best at winning always win. Not only does the best service provider who bid not win, the actual best service provider doesn’t even bid at all, because of the way the system functions … that is actually the largest problem of all.

There is an old-adage, “you can have it fast, cheap and good; but, you can only pick two”. That’s actually not true anymore, comparatively speaking. Today’s high-growth, typically small, technology companies operate at such efficiency that they now offer all three (compared to the larger competitive alternatives, dinosaurs). It’s the only way to function in an industry that moves at the speed of innovation. My first start-up was a perfect representation of this model; it’s how we ended up working with Apple, Google, Adobe, Oracle, Mashable, NASDAQ, Twitter and why we never worked with the federal government.

Jason Calacanis tweeted at the start of my fellowship something that became a constant reminder of this reality:

When I was young you wanted a 100-person company doing $20 million, today you want a 20-person company doing $100m. #efficiency

— jason (@Jason)

July 6, 2013

There is such inefficiency within government inherently that highly efficient technology companies actually can’t operate successfully; the model of efficiency breaks-down in an environment of inefficiency. The government is actually tremendously efficient, just at an incomprehensibly enormous scale that moves too slow for technology innovations, which contradicts its request and leads back to the original IT procurement issue at hand. Innovation and efficiency have become synonymous, if the government wants innovation but lacks the efficiency for it … something has to change, and it’s not going to be innovation.

In addition to meeting with just about every federal agency we could, I personally met with over one-hundred technology companies during my tenure. I met with everyone, from the one-man freelancer to the largest government contractors everywhere across the country from Atlanta to San Francisco. I got a deep understanding of how these organizations operate with their clients outside of government and how they operate with the government, if at all. The objective was to learn how the big guy wins, why the little guy thinks he loses and why the best guy never plays the game.

Rewiring the system

We all know the game is wired, but, once we started to understand how it was wired, we started to change everything. It’s no secret that many IT procurements are pre-wired; heck, even many government jobs are pre-wired … the winning candidate pre-determined and the public bidding process just an illusion of being “fair and open”. If many contracts are wired, that means there is probably commonalities among how they are wired … so, we started to create an algorithm to identify those commonalities. We, almost instantly, went through all of the data and found ten key metrics for flagging an opportunity as “wired” (there’s likely many more) – like completely omitting the title, description or scope of work; you’d have no possible way of knowing what to bid on unless you already knew exactly what to bid on. My personal favorite was using a term that didn’t exist to describe that same term. Several procurements actually copy-and-paste the services requested from specific vendor websites, letting them know “in-code” that they are the pre-determined vendor or service requested and thus alienating others. Identifying potential wired opportunities and the contracting officers that wire them was a huge first step forward.

However, the biggest thing wired of all, was the process itself … unintentionally so. Ensuring that all opportunities are fair and open isn’t helpful when the process to bid can feel overwhelmingly unfair and closed. For example, to get paid by the federal government, you must be in a system called SAM; it’s so bloated, broken and complicated that even most people working in the Office of Federal Procurement Policy have themselves struggled to sign-up for the system, many in government fail to even try. And because of this common knowledge of a flawed system, most contracting officers won’t even review a bid from a vendor not already in the SAM system, despite that being a big no-no. So, we had to develop a solution … it was basic “on boarding” strategy; remove all barriers to entry!

We started to develop a system called “SAM Helper”; a guide built on top of SAM to help make the process more intuitive. We couldn’t just fix SAM, because it was so broken, it’s code-base has been locked-down … a “don’t spend money working on what’s broken” mentality. SAM Helper was the “sign here” sticky-tabs of the real estate industry. For anyone that’s ever purchased a home before, the paper-work is overwhelming and daunting, but, a good agent puts little stickies to tell you what’s most important (because everything is technically important) and where the major call-to-actions are – where to sign and how to quickly and effectively get through the process. The goal was to get more companies through the system. And while quantity wasn’t the objective, it was a by-product of helping quality through the door.

Accelerating government

However, it wasn’t just unwiring a system, getting through a complicated process and removing restricted barriers. It was about getting the right companies, providing the right services, to bid and to win; and, most importantly, to want to do so. The solution, one of many, was acceleration. We started to develop the federal government’s first Start-up Accelerator (a.k.a “accelerate.gov” and “AccelerateUSA”).

This accelerator program isn’t your traditional accelerator program; there’s no equity given up, there’s no communal office space offered and there’s no pitch deck required. It’s also not for your traditional “tech start-up”; the program is designed for any company looking to do more business, or do business for the first time, with the federal government … and to accelerate the growth of that business through those opportunities. There is over 74 billion dollars in IT spend every year within the federal government, that’s a plethora of opportunity!

The program was designed to address three core areas of need: education/training, guidance/mentorship and outreach/networking. The strategy was to leverage pre-existing communities that have mutually beneficial incentives for local economic growth/stimulation to host, support, and lead pre-developed programs throughout the country.

Programs such as “Meet SAM” that would guide companies together through the SAM process – through hands-on and digital walk-tru experiences. Programs such as “GovConnect” that allowed larger and more experienced companies working with the federal government to connect with those new to government – you might ask why the big guy would help the little guy, but, consider sub-contracting and recruitment and it becomes synergistically obvious. Programs such as “N.I.TR.O” (just an internal name – New Innovation Technology on-Ramp Opportunity; love me a good acronym) that provide a new purchasing vehicle for new companies to quickly and more efficiently get into the system and win contracts on an accelerated schedule – this was primarily a way to work against the use of the pre-existing GSA Schedule 70 to circumvent the more traditional (open) procurement process. And, programs such as “MarketPlace” that allowed for easier discovery of opportunities to work within the federal government, making every opportunity an accessible one – even bringing forward opportunities from grants, challenges and research funding.  

One of the most fascinating projects I worked on while a Presidential Innovation Fellow was “PricesPaid”, an initiative of the Office of Management and Budget, OMB, within the Executive Office of the President, EOP. It was just one other developer and myself tasked with building a powerful web application that at its core simply exposes the prices paid for goods throughout all agencies within the federal government. A simple objective, with results of potential savings in the several tens-of-billions! Knowledge is power and while you’d think this information was readily available, it took a Presidential mandate for us to get access – it’s kind of a big deal.

Thanks to the tireless efforts of the OMB, we would have access to more data than you could consume in a lifetime, and it was our task to make sense of that data. My primary focus, as a designer and product lead, was on ensuring that we could see past the limitations that were the specific variables in the data-set; “that isn’t data being returned back to me from the database” is an common response to just about anything remotely “innovative”. However, in the end, it wasn’t just about our making sense of the data; it was about making the data sensible to our audience. The development-centric instinct was to take big data and develop a means of distributing that data … the power however was less in the data distribution and more in the data consumption! Delivering a record of a billion past transactions is meaningless without influencing buying behavior to reduce excess spending, increase competitive bidding and promote collaborative purchasing power in an effort to save billions.

We stood up a working prototype, MVP, within the first 28 days. Well, it was a good headline for the press; all we really did was customize open-source software to take one dataset and output it to a web interface … a more data-rich version of “Hello World”. The real power is when we started to evaluate how we outputted that data. There were immediate findings when sifting through the data: such as items being purchased for zero dollars, vast amounts of excess spending on items with low quantities and unnecessary IT overspending and redundancies. Immediately we were able to identify, exploit and then increase zero-dollar spending situations (how that occurs is actually fascinating and for a much longer-post), we were able to connect contracting offers buying like-minded goods to increase purchasing power through volume discount pricing and we were able to put the data in context which allowed a better understanding of Moore’s law and the depreciation of technology to adjust model and pricing recommendations with respect to time, supply and demand.

We like to say that within the first three-thousand lines of code, we were able start saving over three-billion dollars; but, that’s just another fun sound-bite. The truth is, however, that the program has the ability to save tens of billions of dollars over the next several years as it continues to evolve … no big deal!

The hurdles of government innovation

One of the largest issues within government is that there is such a strong focus on wanting to be innovative that it actually becomes a distraction from developing innovative solutions to real problems. It’s how we get so entrenched in buzzwords like “Open Source”, “Open Data”, “API” or “Design Thinking”. I was in a meeting on my last day where someone from the FDA was presenting on a pretty impressive web application that they had developed; the first question asked was, “are you going to open source this” … to which, the response was simply, “it’s built on open-source”. Sometimes you try to open so many doors that you can’t see the ones open to you. Have one presentation on “Story Telling” and the next 72 hours is nothing but how we tell our stories, until the next shiny thing catches our attention, like agile.

The biggest problem with buzzwords is when they become synonymous with solutions. Especially when the buzzword is easier to sell than it is for the salesman to accurately implement, or even truly understand. “Agile” was a constant development-centric solution with no understanding of it’s meaning or execution. You can’t sell agile as a holistic process when you’re not working within the holistic process. When most developers say “agile”, what they really mean to say is “agile development”. They require every aspect of the process to be more “waterfall”, except for the process in which they specifically work … it’s an irony that halters innovation in government most. I had a self-proclaimed “Samurai” (I think it’s like a “ninja”) who preaches agile as the savior of government tell me he was stopping all his work and doing nothing more because he didn’t have “all” of the designs … he wanted every single view designed and delivered before he’d begin; that doesn’t sound very agile to me. This particular Samurai certainly had the best intentions, but, it’s certainly no way to innovate government. Work hard, work fast and constantly be iterating on the process holistically along the way; you’re never done learning and there is no “finished product”. 

Understanding the weaknesses early became essential to avoiding the unnecessary, increasing efficiency and knowing whom to move-forward and when to move-on. Everyone in government wants to do “innovation”, but no one wants to do anything that is new, untested, uncomfortable and requires any level of blind-trust or risk. There is the buzzword and then there is the reality. 

Helping government make sense of itself

The PricesPaid initiative was a fantastic early example of how the simple distribution of big data can prove more harmful than helpful without the understanding of context to create new, more relevant, data. It was also a great example of how the buzzword wasn’t the solution, a solution was the solution … go figure!

The simple goal was to release the spending data of prices paid on goods throughout agencies in the federal government. The basic premise was that if you could see how much others paid for an item, you’d be better educated on how much you should be paying for a like-item. However, that premise requires a lot of effort on the user and can be very misleading, especially in the context of technology. If you knew what you were searching for, you could get a glimpse of the high-end and low-end of the spectrum of what others paid … obviously you’d want to see if you could get the lowest price possible, but, you’d feel comfortable knowing you landed somewhere in the middle (the average – or you could try to roughly calculate some version of the mean, median or mode). However, by simply providing a historical transaction record you’re basing purchasing behavior off of very misleading data. The data can be 15 to 24 months old, and we all know that you shouldn’t be paying for an iPhone 4S today what someone else paid 18 months ago; because of the depreciation of technology. Furthermore, if there is a downward slope of iPhone 4S purchases exactly inversely related to an upward slope of the iPhone 5 purchases, we could more accurately assume that there is a new model of iPhone and that you shouldn’t be buying an iPhone 4S at all!

Simply opening the data is like giving someone a set of encyclopedias. By actually understanding the data, considering context and focusing on the influence and changing of behavior is like giving them Google … you get exactly what you were after worst case scenario and you discover something more relevant that you didn’t even know you were searching for best case scenario.