How the president's American Tech Council should tackle reforming government tech
Total Response time: 0.22962093353271
One of the most common problems that arise at startups is the development of what I call technical debt -- a predicament created when second-rate code or software is deployed in the short run to speed up the product development process.
The intention is to fix or replace the “short-term” code later, but what often happens is that the tech debt grows when the “short-term” code is never upgraded or fixed, and sometimes convoluted skyscrapers of code are built on kludgy foundations. One well-known example is MySpace, which for many years struggled with a snarled "Millennial Tower" of legacy code built on top of a poorly designed architectural foundation.
Like Silicon Valley startups, the Federal Government grapples with technical debt - but on a massive scale. Look at the U.S. Department of Veterans Affairs, which is spending 75% of its technology budget to maintain outdated legacy software systems, or the FAA’s air traffic control, which is still actively using the paper “flight progress strips” to track planes. The Air Force’s nuclear ICBM missile launch process runs on floppy discs.
On Monday, the Trump Administration will host the newly created American Technology Council (ATC) in Washington DC. The aim of the group is to “transform and modernize” the federal government, and it also reflects President Trump’s oft-stated desire on the campaign trail to run government more like a business. The council is composed of high-level Trump administration officials (including Jared Kushner and V.P. Mike Pence), and yet-to-be-named Silicon Valley executives.
They will discuss five major topics: how Washington can better procure technology; adaptation of private technology to government problems; tech workforce improvement; an exchange program to bring more engineers into the government; and finally, H1B visas.
Having served as a senior officer in the federal government (in both the Bush and Obama Administrations), and after working privately as a venture capital investor, try to act as a Silicon Valley-to-DC translator and “bureaucracy hacker.”
As I see it, the two most challenging initiatives are: 1) deploying innovative technology to solve long-standing government problems and 2) solving the procurement problem. These two initiatives are inextricably connected; you can only deploy the new technologies if you can procure them.
To address the Federal tech debt, here are four “bureaucracy hacking” best practices that the Administration and its new advisors might consider for the Tech Council meeting:
Streamline the Procurement Process: Unlike the process of building a business, our government’s procurement process is cumbersome because our democracy requires treating all opportunities with a level of fairness, equal opportunity, and transparency. Businesses can often forego these in favor of efficiency, and in Silicon Valley, there are few constraints (which also contributes to the Valley’s well-documented age and gender discrimination). That said, there is no question that the government has construed these standards in ways that make the process opaque, overly complex, time-consuming, and difficult for anyone but Beltway incumbents to access. To rectify these issues, procurement process reform is needed to swing the balance on the government pendulum back towards greater efficiency.
Know Your Customer: Valley executives are well-aware that if you want to disrupt a traditional mode of operations, you have to “know your customer”. But imposing changes from on high without “knowing your customer’s needs, mode of operation, and culture” won’t mobilize civil servants in the Federal bureaucracy. Palantir is one company that mastered this “Know Your Customer” approach when working with the Federal Government. When Palantir was a little startup with great technology, they spent time in D.C. figuring out the data analytics needs of the intelligence community, educating them on Palantir’s capabilities, and engaging the relevant stakeholders, including operations folks in the bureaucracy, senior officials, and elected representatives. They quickly became experts on how the government does business, and the rest is history: contracting vehicles have become more flexible for easier adoption of innovative technologies, the government now actively looks outside its walls for solutions, and Palantir is a multi-billion dollar company.
Bring the Right Stakeholders to the Table: Since the list of tech executives who will be meeting with the Tech Council hasn’t been released yet, I’d like to add my two cents as to who should be in attendance- specifically, executives who understand the government as a customer and have proven capable of tackling and transforming government culture to implement change. While C-suite executives from the big tech firms will likely be at the meeting (Google, Apple, Facebook, etc.), it would be helpful if more technically-inclined executives were invited. Take for example, Bill Vass, the VP of Engineering at Amazon Web Services, who has overseen the deployment of Amazon’s GovCloud and has past experience as a CIO and CTO across the Department of Defense and Armed Services. Or Susie Adams, CTO of Microsoft’s Federal Government business unit, working across the Defense, Intelligence, and Civilian Agencies. Execs from Accenture, Booz-Allen, Deloitte, and Palantir should be there also, because all of these companies have deep experience deploying new technologies and services into the Federal Government bureaucracy. They have real knowledge of the technology problems in the government (information not often shared with the private sector), know how the government operates, and have proven capable of implementing change.
Hire an Execution Team: Ideas are nice, but change is better. And change only happens if you execute on those ideas. As of last week, Trump had yet to nominate or fill many hundreds of federal jobs, including almost 1,200 positions that require Senate confirmation. Without a team of executors in place, even the best policy or operational changes will stall. Having admitted that he is unfamiliar with the quirks of the federal bureaucracy, it would be wise for Trump to engage people with proven track records in the space. And fast!
Mobilize and Respect the Civil Servants: Even the best products can fail when rank-and-file stakeholders aren’t shown respect and properly incentivized to use them, and since government does not have the same motivational carrots and sticks at its disposal as businesses do, the Trump Administration needs to learn how to steer and motivate the bureaucracy. Anyone who has ever rubbed a government employee the wrong way knows that, if they so choose, they can make anything impossible. Transformational leadership requires not just the imposition of new rules and initiatives, but also the mobilization of the crowd. This should be a promising sign for Trump. Many would say that he won the election because he was a better master at reading and motivating the crowd. His ability to motivate the government “crowd” to address technical debt will be the greatest test of those skills yet.
Let’s hope President Trump succeeds in deploying these 5 hacking techniques, because the stakes are high. Our veterans are suffering as they wait for the coordination of medical care, our critical infrastructure is at risk of devastating cyber attacks as we delay upgrading our technological infrastructure, and our ICBM nuclear missile system’s error-prone outdated technology is an existential threat to us all.
For a country at the forefront of technology innovation, our government has too much technical debt and needs to keep pace with private sector advancements. The idea that Washington should learn a few things from the world of technology, innovation, and entrepreneurship is not new. The problem is not in the thesis, but in its execution.