ICYMI: Government has a technology problem

By Andrew McMahon, Director of Strategic Partnerships, Dcode42

Part 1 of 2.

Over the last 18 months, there has been a growing number of voices pointing out the issues inhibiting the government from effectively building and buying modern technology. The consequence of this failure is costly, causes harm to individuals, and in our opinion is one of the major causes of the distrust and lack of confidence in government. The public cannot (and should not) trust an organization that fails to deliver services and protect personal information in a manner that they expect. While rubbernecking at failed IT can be interesting, what hasn’t been tackled in depth are two important questions:

1. Why is it so difficult for the government to accept change and the risk associated with it?

2. What can we do about it (including the press, private sector, and Congress)?

Bureaucratic inertia

A former boss used to say that inertia was the most powerful force in government. Over six years working across almost every major agency, trying to break this cycle of failure made me aware of this inertia time and time again. And because the status quo is so sticky; some companies, media organizations, and most importantly parts of the government’s bureaucracy begin to benefit from its predictability. There is value derived from a lack of progress. This is one reason why we have inane systems and increasingly irrelevant companies that persist, even in the face of repeated failures. Beyond those set on maintaining the status quo, there are structural issues in the way as well.

The structural problem is threefold, (1) agencies try and avoid risk outright rather than manage it and the punishment for failure far exceeds the reward for success; the government maintains (2) a procurement process that has become more important than the outcome; and (3) a hiring process that is torturous for the applicant and rarely relies on subject matter experts to identify talent.

Moving the needle

Pointing at failure is a bit of a sport in D.C. but rarely do the newspaper articles, audit reports, or hearings promote the smaller steps that can be taken to remedy the larger issues. There are steps that can improve the adoption and management of technology in the short term. There seem to be four major stakeholder groups that influence the adoption and management of emerging technology. In this post I will discuss the Executive Branch agencies and private sector companies. In future blogposts, I will cover the role of the press and Congress. Each of these groups can provide a significant contribution to the progress of our government.

In agencies

Agencies, in the end, are in the driver’s seat here. There is enough money and talented people in agencies to make progress. GSA was successful at modernizing its technology starting in 2011 with no more than a strong Working Capital Fund and dedicated leadership. The agency has saved hundreds of millions and modernized many pieces of core technology. Just to be clear, I am not simply patting my former employer on the back, this should be the minimum bar. Agencies must go beyond modernization and savings and those that are doing less are failing.

Every agency has core mission areas that need to be constantly improved. Without creating a space for emerging technology to be adopted, progress will be tepid. The DHS Science and Technology Directorate contracting work is a great example. By identifying areas of interest (i.e. border protection, information security in the financial sector, even wearables for working dogs) and finding companies/products that meet needs in these areas they are avoiding the requirements trap. Here, DHS accepts that companies have likely already built product to meet most of the need and DHS shouldn’t try and reinvent the wheel for themselves by building out requirements.

In many agencies though, procurement rarely goes beyond the normal course of doing business. To quote Jeff Bezos, “The process becomes the proxy for the result you want. You stop looking at outcomes and just make sure you’re doing the process right.” This is government in a nutshell. And the procurement process may be one of the most powerful. Agencies need to stop putting compliance and timeliness of the process and prioritize the outcome.

“Do we own the process or does the process own us?” — Jeff Bezos

Even with creative procurement methods, without expertise to differentiate between companies with great products and mediocre ones there’s little an agency can do other than do the things that didn’t get it into trouble before. This invariably favors incumbent vendors. That makes hiring critical and unfortunately agencies hire in the same way they buy: the process rules the outcome. Agencies must prioritize the hiring of technology and information security experts and put the decision making in the hands of people who know good talent from average talent.

This, in my opinion, is the magic of USDS and 18F. Recruitment, applicant interactions, assessment of quality, and selection were controlled by people who knew a good designer from a management consultant. HR helped with the paperwork, documenting, guiding between the rules and policy considerations but wasn’t involved in talent selection. This shouldn’t come as a slight towards HR offices in government, if they had to know how to select for quality in every role an agency has needs for, it would be impossible. They should and are experts in the federal hiring process. The two cannot go on being confused in agencies.

From the private sector

Companies rightfully bemoan the procurement process as being far too long, complicated and cumbersome. Its an honest assessment. It is out of their control and that’s why so many companies sit on the sidelines at the detriment of the government’s technology. Unfortunately, tried and true enterprise methods of business development like the free-trial, beach-head, and proof of concept do not have the same benefit and success in government.

By taking a bit more forward leaning approach, it would be great to see companies that could have real impact in the public sector to demonstrate their product’s potential in government. Even without the government involved or asking for it. For example, the many geospatial imaging companies that can count cars in Walmart parking lots to predict quarterly earnings, could likely predict with a level of accuracy where wildfires may be the most catastrophic. A company could do this analysis upfront, share it with the world (and the U.S. Government), and then when the company produces better results than the government, point it out and ask for a conversation.

A level deeper would be for companies to bid on an RFP that is loaded with laundry-list requirements but where your widely available solution could meet most of the need. When you lose the procurement (which should be anticipated), talk publicly about why the RFP got it wrong and how your product could have more quickly and cheaply solved the problem. Talk to the press, Congress, whoever will listen. This approach is a bit riskier as you are likely to annoy the buyer but highlighting the insanity of the requirements process is needed.

The Veterans Affairs scheduling app fiasco is the perfect example of where this could have a positive impact. Notwithstanding a company’s desire to step into the fray here, there are dozens of scheduling, patient experience, and practice management products that could meet the VA’s need. If the VA had a grasp of the digital health market or worried more about outcome than process, any and all of the companies below should have space to prove their worth in the VA hospital system.

The trouble with “innovation”

There has been a great deal of energy focused on “innovation” in government recently. In my experience, you cannot just call something innovative and make it so. Innovation is not a means nor an end, it is an adjective you use to describe the means or the end once the dust has settled. It certainly helps with branding and press but be wary of self-branded “innovation agendas”. Accepting risk and failure as a road to a better government should be the message.

Moving quickly with smaller actions, breaking hardened process, and scaling successful products is the only workable way to escape the pull of inertia while working in the public sector. I believe that true reform won’t come from a grand policy proposals, “SWAT” teams, or saviors with arbitrary titles. Rather, it will take thousands of small actions from within government and from without to disrupt the procurement industrial complex.

In early May, Dcode42 will begin our Spring cohort, which will support companies that are working on solutions in Artificial Intelligence and Machine Learning. These solutions have the power to change every aspect of society — from national security and economics, to health care and citizen engagement. Our hope is that each of these companies makes a small dent in the ecosystem. We know that government won’t change overnight and we’re working to prepare companies so that they can better win contracts and change government through their disruptive products.

Stay tuned as I discuss the role of press and Congress in moving the government forward.