Even before the increased demand caused by COVID-19, much of our digital infrastructure was so poorly designed as to be hostile to users.
Despite years of modernization efforts, the COVID-19 crisis shows that governments at all levels have failed to adequately fund the social digital infrastructure we need to ensure our well-being.
The first bridges to collapse have been unemployment insurance systems across the country. The Small Business Administration is now similarly unprepared to process the loans many companies need to stay above water.
This crisis is likely to again expose many systems as inadequate to serve the needs of our country.
Even before the increased demand caused by COVID-19, much of our digital infrastructure was so poorly designed as to be hostile to users. Needlessly complicated user interfaces, unnecessarily probing questions, limited processing hours, incompatibility with common devices and browsers, and frequently offline systems are all-too-common traits of government support systems.
It doesn’t have to be like this. Many of these problems have well-known solutions. As part of the original rescue team for HealthCare.gov, we saw first hand what it takes to turn a failing government system into one that can scale and adopt the flexibility and high-quality user experiences of consumer technology.
The U.S. Digital Service, 18F, Code for America, and a handful of new digital services vendors have helped forward-thinking career civil servants make progress on these goals, but we need to be honest about how much has changed and how far we still must go to adequately resource critical digital services.
The vast majority of money spent on technology projects goes to traditional procurements with meticulously planned lists of requirements that routinely fail to create the simple, reliable services the public expects. In spite of a desire to have government function more like consumer technology, outside of a few pockets of innovation, government technology today mostly functions like it did 20 years ago.
As agencies find their way through this current crisis and begin preparing for the next one, they must include substantial investments in our national digital service infrastructure.
First, governments need to take advantage of the unique ability for digital infrastructure to scale in the way that call centers and physical infrastructure can’t. If there was a bridge design that could automatically add extra lanes for an evacuation, we’d use it to build every major bridge. Scalable, programmable commercial cloud infrastructure is the expanding bridge for government services. Nearly every government service and application should be on the cloud, and the vast majority of those can use commercial cloud providers rather than ones built specifically for government.
Second, governments need to abandon outmoded models for procuring systems. Right now, most government technology contracts are structured to place the risk of failure on a large system integrator who manages other vendors. Government must take the leading role in day-to-day work and define contracts in terms of outcomes for users rather than outputs from vendors. To do this effectively, they’ll need to hire civil servants with a deep understanding of consumer-level digital services and pay them enough to stay in government.
Third, the government needs to adopt agile software development. It’s not worth arguing over the margins of competing methodologies when the status quo is so bad. Attempting to plan for every risk in a 10-year contract is a massive risk in itself. Circumstances will change. Your contract might assume people can safely leave their homes and the full economy is open for business.
Instead of planning for every contingency, plan for flexibility. It’s the only way we’ll be able to start building the digital infrastructure we need to support our country through the worst of this crisis and prepare for the next one.
Greg Gershman is the CEO and co-founder of Ad Hoc. Before founding Ad Hoc, Greg was a software engineer, member of the inaugural class of Presidential Innovation Fellows and part of the team sent by the White House to help stabilize HealthCare.gov.
Paul Smith is the CTO and co-founder of Ad Hoc. Paul also co-founded EveryBlock, one of the first hyper-local news websites and was part of the White House rescue team for HealthCare.gov.