First of Kind › The Joe Gebbia interview

So, good to see you, man.

Yeah.

We're in the heart of Washington DC. You could say it one of the most prestigious addresses in the planet. We are in the White House grounds at the Eisenhower Executive Office Building. This is a building from turn of the century 1800s that I think is one of the most architectural masterpiece of Washington DC and I thought it would be a great backdrop for our conversation today. I love that I learned yesterday this room used to occupy the founding documents of our nation. That the Constitution, the Declaration of Independence, resided here for decades as part of the State Department Library and is now the epicenter of this new initiative that you're overseeing.

It wasn't a straight line, that's for sure. I remember when I first heard of the government doing things with Silicon Valley was during the Obama administration back in the day. And I was deep deep in Airbnb at the time building the company with my co-founders Brian and Nate. And I remember thinking, 'Wow, what a cool application of design at such a grand level, you know, design for society.' I remember it planted a seed in the very back part of my mind. You know, years went by and you kind of see what was going on and kind of look for those moments of like can the government adopt, you know, high standards around design and actually push through to the other side, which would be you go to a website or you have an experience with the government and it doesn't feel like the one that you're used to.

It's a lot higher than I ever thought as I've gone deep into this. It turns out that there's around 7 to 8,000 .gov domains. There's about 27,000 web portals with all the subdomains and everything. And within that there's nine .gov websites that 160 million Americans go to every month. That's practically half the nation going to nine websites. And what's occurred to me, especially since being here the last 8 months or so, is that at one point in history, Americans would access the government by walking into buildings to get stuff done. And I have to imagine at that time the lobbies would be impeccably designed, great lighting, nice art on the wall, clean floors every morning when people walk in, nice reception desk, good wayfinding and signage so people know how to navigate the building. And the reality is the front door to government is websites. And so why wouldn't we treat at least those nine .gov domains, if not the whole surface area of the government with the same care and attention that those lobbies got all those years ago?

So, in February, I got an invitation to come fix a digital problem in the government, which is this. For decades, going back to the 1960s, federal employees have retired using paper, paper forms, paper applications, the process hasn't changed since the 1970s or so. Also, the time with which it takes someone to retire in the government also hasn't changed in about four decades. It can take up to six months.

Six months. In fact, it's so complicated. Federal employees are required to do online trainings just to figure out how they retire from the government. So, you've served for decades of your life, 30 plus years, 40 plus years, given so much effort to our nation as a civil servant, and then you're met with this complete obstacle course to try to exit the government and receive the benefit that you deserve.

That's what happened to me. That's why I said, 'This is an interesting problem. I'll give it six months and I'll see what we can do here.' And so I put a small team together of software folks out of Silicon Valley, designers out of industry, and we set off on this adventure, that's what it was, to figure how do you turn a paper-based process to something that's fully digital, completely online in less than 6 months. Now granted, you just have to understand the context here. The government's been trying to digitize this since the Bush administration in 2002. It's been 20 plus years. Reports are that it's been over a hundred million on contractors that kind of fizzled out and never really solved the problem. And so, most presidents have tried to solve the speed problem by putting more headcount on it. And so, now there's close to a thousand people or so that just process paper for retirement in the government. So this is the backdrop with which I came in to help solve this. I got paired up with the career folks at OPM which is kind of the HR department of the government and started to realize the problem at hand and very quickly discovered that it involves a mine in the side of a mountain in Pennsylvania.

A mine. Yes.

Correct. And I had a similar impression. I go, 'What's in that tunnel? What is the government doing with a mine?' It turns out there's this service called Iron Mountain which has this mine underneath this mountain in Pennsylvania and the government's been a tenant ever since the 1960s as a place to store paper. If you're in the business of storing physical media, this is the perfect place. It's highly secure, climate controlled, and no natural disaster is ever going to touch anything in there. So, it makes total sense if you have a lot of paper, put it in such a safe place like the mine. That was great in 1960. It's 2025. What are we doing?

Oh, this on the order of like hundreds of miles high-speed train in motion, right? You can't stop the train, right? However, it is ironic that there is a bit of a limit to how many people can retire from the government because of limitations on how much can get processed because it's by hand, right? It's moving physical goods around. There's somewhat of a limitation on that unless we add more people to it. I always thought it's a little unusual that shouldn't anyone just want to be able to retire when they want to retire, not get stuck in a backlog, right? And so, it's 2025. We have modern tools at our fingertips that are so easy to access and costs keep going down year-over-year. So, what would it look like to bring in the entire retirement process completely online without any new paper created? That's what we worked on.

So to continue with the mine story and share how we got over this Rubicon of paper, we with a team of four started to really understand how the whole process works end to end. And we did something very very early on which was we sat down with the career teams meaning the people who are in government. They're not political appointees. They're with whatever administration comes through. They've been trying to solve this for decades. And they were very happy when we showed up to say maybe we can try some new ideas, provide some air cover to do things a little bit faster than they've been done in the past. I remember on the very first day, my very first day back in February, we're in the room with a couple of the retirement services team who's been trying to solve this problem and we talk about what would it look like if we took one retiree and we handheld their application to be an all digital process along the way. This is the roughest prototype you could possibly imagine just as a proof of concept that there is at least one way to keep an application all digital because there are some that start digitally but then they end up as you move through the weeks to months long process. And so remember we polled the room how long do you think it would take if we were just do this all digitally and someone raised their hand say about four months. Somebody raised their hand and go, 'That's at least six months.' Somebody on the screen goes, 'Ah, that's probably a year.' And my colleague and I, we look at each other, look at the room, and said, 'What if we did it in a week?' And everyone was like and we started to say, well, if you had support from leadership on the top and we actually just took this by hand and anywhere it got blocked or stuck, we just figure out where is it stuck and we go call the person or go actually down to the floor and talk to them at their desk say, 'Hey, we're running this experiment. Can you help us out? We're trying to move this along as quick as we can.' So the atmosphere turned from fear to excitement pretty quickly where people said like okay wow this is we've never had quite a challenge like this before one week to move one retirement digitally without any new paper and so we eventually found one retiree and we called her Julie and the team went off and they didn't do it in a week they did it in four days.

Yeah. Do things that don't scale. This is an early lesson from Airbnb. We would run around the country renting cameras, taking pictures of host apartments, meaning we, me, and Brian. Like, that doesn't scale. You can't be running a company while you're going to take pictures of host apartments in cities around the United States. But in doing so, it taught us so much about what was working and not working with our website because we meet our hosts and they'd tell us, 'Hey, Joe, love your site, but I needed to do these four things. Your calendar needs to do this. The messaging app needs to do this.' And so we were getting feedback and improving the quality of listings by having nicer photos. We eventually scaled out the photography program where we had thousands of photographers all over the planet providing free photography to hosts who signed up for the site. And so, back in February, as this is unfolding, I'm thinking back to those days in my Y Combinator. I'm thinking, wow, like the ethos of YC has actually found its way into the government. Like we're just employing the same thing that Paul Graham would have said in our office hours but in a place that has probably never had too many of the YC sort of methodologies applied before. And so with this newfound excitement and prospect that okay, maybe this is possible. Maybe we can conquer this 20-year quest to make retirements digital and make them faster and let people retire with dignity instead of going through this terrible obstacle course of paper where you're filling out forms. You're not really sure what you're checking boxes for things because you don't really know how it's going to impact your annuity payment that you get every month. Most people, they'll miss a signature field here or miss a checkbox there. They don't find out about that until the very end of the process months later. Then they have to go back to go.

I mean, the whole thing, I joke, is like a retention program.

So, we took the Julie prototype, we brought it to let's find 20 more people that have basic retirements, meaning they worked for one agency, they don't have a lot of the complications around it, and let's just start to move people through the next version of that prototype. And we successfully got the next 20 people through. Their processing times were remarkably faster than on paper, obviously. And so then we took the next 20, we said, 'Okay, let's take work that OPM's been doing and let's blow it out of the water. Let's completely rewrite the software that they had been kind of working through to try to do this.' And so we rewrote the entire software stack and were able to develop the product called Online Retirement Application, ORA, through a ton of iteration and tons of user testing, lots of user feedback. Talking to HR, payroll, the applicants, OPM, the whole ecosystem of stakeholders like you do with any good design, who you actually designing for, understand their needs, translate that to digital products. And so employing those basic methodologies into something that's highly usable, it's intuitive, how do we reduce customer service calls through better product. On August 1st we launched ORA which for the first time all new retirements in the entire federal government which has 2.3 million employees and has 100,000 that retire every year. Every new retirement of August 1st flows through our software. No new paper. And what took 6 months before for the basic retirements now takes 60 seconds.

Yes. All agencies and all retire employees now submit through software.

And presumably the political class has been really receptive to these lessons, these tried-and-true approaches from Silicon Valley.

Well, my takeaway from the retirement product was people are very receptive to good design, better usability, and I think it's actually one of the most bipartisan things one could work on in the government. Everybody can agree, doesn't matter who you voted for, that they would love government websites to be better designed, have great usability, and load faster.

Yeah. Yeah, it's the least controversial thing you could be working on. As I've been here, seeing the opportunities and actually getting reachouts from cabinet members and cabinet secretaries, directors from agencies across the government saying, 'Hey, Joe, I heard you're on campus. Can you please redesign my stack, my experience of my agency?' And earlier on, I was like, 'Yeah, maybe, I only got six months here and then I'm going to roll back to my regular life.' But as these began to pile up, that's when the idea began to form. Well, what would it look like if you structured something that could really go after the broad swath of digital services across the government? What would that be? As I was contemplating this and thinking it through, somebody took me off to President Nixon and an initiative that he had during his administration in the early 1970s to improve America through better design because it was a presidential initiative. It attracted the best and brightest of the time.

Well, one of my all-time favorites, Charles Eames, contributed to that. Massimo Vignelli contributed and all of their peers, the best of the best designers came to DC to contribute to Nixon's presidential initiative. It's called the Federal Beautification Project.

Well, they left us a lot. They left us literal icons. So, the NASA logo came out of this and anyone in the design world knows the worm and the brand guidelines behind NASA. I mean, it's such a classic. The highway signage system that we know and love today came out of this initiative. The agencies across the government got visual upgrades to the design standards of the 1970s. So the EPA logo which is a fan favorite in the graphic design world. Also the national parks brand system which is also a beloved brand to this day. These are robust icons. Like the output that came out of this group from that initiative has withstood the test of time. Here we are five decades later and they're like heralded as the gold standard for graphics.

Yeah. So with that as inspiration, how do you imagine the work that National Design Studio will get to work on? Well, the question for me was how do we continue what started back in the 70s under this president and for the digital age, right? To attract the best talent of our time to come upgrade today's government to be as beautiful as it is usable. The analogy I give is an Apple store-like experience. What if the government had great design, amazing usability, running on modern software, the same that any great consumer app would have? And in much the way that the NASA logo became emblematic of American excellence, technical excellence, design excellence, you would think that an Apple store-like experience for the federal government would also be an incredible emblem of today's American excellence. Yes, it's disheartening to me that we have a nation that has such technical excellence that we can produce the XAIs and OpenAIs of the world and at the same time we have government websites that still have operating hours, meaning they turn off at a certain time of the day because they have to update in the background. How is that possible?

I'm not okay with that. It's like where are these servers located? It's like this is crazy to me. Like I want to be proud of how our nation presents itself to the world and certainly to the people who live here. We can do better. We have the talent. We have the technology. It's just a matter of joining forces in this case under the presidential initiative that President Trump has created called America by Design. So this is his version coming bigger than President Nixon to really reimagine the digital surface of the US government and really upgrade them to 2025 standards.