Emily Tavoulareas
‘It’s Knowing What to Do When the Opportunity Arises’: Emily Tavoulareas on the Importance of Timing to Making Change in Government
| Emily Tavoulareas was a Founding Member of the U.S. Digital Service at the Department of Veterans Affairs (VA) and Senior Policy Advisor at the White House Office of Science and Technology Policy. Prior, she was Senior Digital Advisor at the VA. |
Emily Tavoulareas was a founding member of the USDS’s first-ever agency team at the Department of Veterans Affairs, and stayed with the team through the 2017 Presidential transition. She and her colleagues prioritized product, design, and human-centered approaches within the agency, unlocking better communication with — and services for — veterans.
During her time at the VA, Emily saw USDS grow from an idea, to a small team of fewer than 10 staff, to a larger team working across multiple agencies. Below, Emily talks about the foundational work that occurred in the years before USDS was even a concrete idea, how to sustain momentum after a crisis, and the dangers of relying on quantitative data as the only source of truth.
May 2, 2024
Kathy Pham:
Emily, tell me about your journey to USDS.
Emily Tavoulareas:
A colleague of mine, Bonnie Shaw, had several friends in the Obama Administration who were working on digital media and technology. Bonnie was tapped to help run a couple of hackathons at the White House, and I joined one. Well, the event was called a “codeathon,” not a hackathon, because hacking was considered a scary word at that time. This was 2012, the first year of the Presidential Innovations Fellows (PIF) program.
We went to the White House and I remember thinking: “How cool is this? This is super interesting. There’s lots of energy. And, look at us applying all this stuff to government — it actually seems useful. Also, where are all the ladies in the room?” The PIF program at that point had only two women: Kara DeFrias (now Fitzpatrick) and Marina Martin (now Nitze). I did not know them at the time, but would later work with both of them.
A year later, Brian Ford at the Office of Science and Technology Policy (OSTP) reached out and said, “Hey, would you like to bring your experience with technology to government? Specifically, the Department of Veterans Affairs (VA) could use your help.” He introduced me to Tommy Sowers, the VA’s Assistant Secretary in the Office of Public Affairs. Tommy’s office ran the VA website, and he was personally very enthusiastic about the potential of technology to transform public services — and specifically, to completely transform the way that the VA engages with veterans.
The role was pretty undefined, had no staff or hard authority, and it was temporary, but I was intrigued. So I originally came in for a year to help this Assistant Secretary reimagine the way that the VA engaged with veterans. I had no idea what was ahead.
On my second day, a VA employee (and combat veteran) dropped this very thick binder on my desk and said, “They’re going to build another website. Make them stop.” I don’t know if that’s exactly how it happened, but that’s how it stuck in my brain.
He pointed me to a section in a very lengthy contract, and I thought “This sounds like a marketing landing page. How much are they paying for this?” I don’t remember the exact number, but it was tens of millions. Having come from a creative agency that was building websites for private companies, I saw this number and thought, “What the hell are they making? Does it build houses? Does it make insulin? What does this website do?” So I started digging into this contract and raising flags with all the right people inside the organization.
There was only so much we could change at that point, but with the help of my colleagues across that office and the contractor (who by the way seemed to welcome the changes and was great to work with) we made the most of it and reoriented the website to test out plain language pages describing benefits, eligibility, and processes. That website, which was called explore.va.gov, became an A-B test of plain language content for VA benefits. In some ways, this was the prequel to vets.gov.
Kathy:
You’ve mentioned a lot of things that laid the foundation for what is to come: the PIFs, the codeathon, being called into the VA. Because nothing really happens overnight — this was a year before USDS started and before HealthCare.gov went down. How do you see all this leading up to USDS and your official start there?
Emily:
It’s an interesting journey with at least one more crisis. A couple of weeks after I arrived at the VA, Marina Martin arrived as the new Chief Technology Officer (CTO). She had already been at OSTP, working on veterans issues from the White House. She arrived with three PIFs: Mollie Ruskin, Ben Wilman, and Tom Black. I became fast friends with them, and we quickly realized we were pointed in the same direction.
Then I met Amber Schleuning, who was the Deputy Director at the VA Center for Innovation (VACI) and had recently joined from the DoD. She was very excited about the potential of human-centered design in transforming the entire agency. At some point early on, it became clear that Mollie was working in an extraordinarily toxic office, and we were able to move her to VACI with Amber. The two of them started thinking about how we could bring Human-Centered Design to the VA, which would later turn into a pilot that created a tidal wave of work around design at the VA.
Marina, in the meantime, wanted to codify her vision for the VA. She brought a group of us together and we whiteboarded this thing that turned into “the VA glossy” — which she spent her own money to print and handed out at a VA executive retreat. I don’t think she anticipated how much it would stick. When she got back to the VA building, people still had it on their desks and in their binders.
Another important moment in 2013 (at least in my version of the events) was the open data executive order. Agencies were scrambling to implement it so they wouldn’t get yelled at by the Office of Management and Budget (OMB). I got pulled into a meeting that was portrayed to me by colleagues as a scary/annoying meeting with OMB and the General Services Administration (GSA). Gray Brooks was in that meeting. I guess I said something sensible, but whatever it was… after the meeting he said, “Do you know Erie Meyer? Because you should meet her.” Gray then introduced me to Erie, who I instantly adored, and still do all these years later. She is a force of nature, and at least in my view the spirit of so much of our work at that time. It’s like her energy just seeped into everything.
So back to the open data push… the open data executive order required that a webpage dedicated to open data needed to be created at the VA. People at the VA were not excited about this, so we offered to do it. Mollie did the mock-up and tapped Charles Worthington — who was a PIF at the time — to scrub in and develop it. By the way, raising your hand to do something you care about, but that no one else wants to do, is a pretty effective tactic.
Kathy:
I love the PIF thread: They were assigned to other things, but because they’re already in government, they’re able to help on projects in dire need. We didn’t know that all of these things would eventually lead to 18F and USDS.
Emily:
Totally.
Also I have to say that the VA employees who I worked with in those first few months are a critical part of my story. They kicked off a bunch of these threads by just knowing who had information and experience that didn’t otherwise exist in this office. There are people who have spent their careers in these organizations and recognize the gaps, how to fill them, and where the levers are. Everything I learned about the VA, I learned from them. What could have been a contentious relationship was a very constructive one. I had no budget or staff, all I could lean on was trusting relationships. I went out of my way to listen to them and be helpful everywhere I could.
So in 2013, all these seeds had been planted: Marina Martin brought in three PIFs. Charles Worthington — who eventually became (and still is) the VA’s CTO — built the first open data website. And then the scheduling crisis happened. I think that was February 2014. Veterans were dying waiting for healthcare and appointments, and it was discovered that the VA kind of cooked the books on how they were counting time for appointments. And this just blew up. Secretary Shinseki resigned. Somewhere in there Tommy Sowers also left, but for different reasons. There was a lot of turnover. And by the way… this scandal with the wait times is such a tragic example of how arbitrary requirements can warp incentives in a manner that is ultimately harmful to the very people they are there to serve.
I’m not sure who decided that quantitative data is the only information that we should value as true, but I hate it. I think this is the root of a lot of our problems in government. There is no room for observable reality — only what is measurable. Information that is measurable is only a fraction of the picture, and provides little to no insight into how or why things are the way they are. The wait times scandal is such a painful example of this. Someone somewhere created an arbitrary requirement that was not achievable for reasons that are very likely beyond their control. The people on the front lines who were responsible for meeting that requirement probably knew that it was not achievable — did anyone ask them? I’m going to go out on a limb and guess that someone probably raised red flags and was shut down. This happens constantly. There’s a problem (long wait times), and someone decides that this problem would go away if people could get an appointment in X number of days. Then a new policy is announced, and the people responsible for implementing are held to a standard that doesn’t match their reality for any number of reasons, and are put into a position where they will be failing constantly. If someone ACTUALLY wanted to decrease wait times, they would spend time understanding why are wait times long? what are the obstacles to decreasing wait times? … and then focus on those. But I guess it’s easier to just loudly announce a requirement in a press release and call it a day.
Anyway, as this political crisis was unfolding, Mollie Ruskin and Amber Schleuning were leading a human-centered design sprint in various locations across the country. I was part of this sprint team along with a VA employee who is exactly the type of civil servant who you want working in government. She’s a veteran and spent her entire career at the VA engaging with veterans, on the ground. She was the kind of employee we would meet and think, “Why isn’t everyone listening to you? Everything you’re saying makes sense.”
After Secretary Shinseki resigned, Sloan Gibson was the Acting Secretary for some time, and then at some point in the early summer Bob McDonald came in as VA Secretary. Lucky for us, Secretary Bob (he preferred to be called Bob) was coming from Procter & Gamble, a company very well-versed in design methods.
So when Bob came in, Mollie delivered that first report titled “Toward a Veteran-Centered VA.” We intentionally decided to not make the report digital, and instead to print it, because Marina had such an amazing experience with her past printing. And people still to this day have it on their desks.
This became a ticket for a conversation with leadership. The report made it all the way to the White House. Lore has it but it made it into the Oval Office. I don’t know if that’s true or not, but it definitely made it into the Secretary’s Suite. Bob McDonald and Sloan Gibson read it and were in. This unlocked a whole bunch of conversations, resources, and work. Several people were then pulled into a customer experience office that the VA set up. A new round of PIFs came in, including Sarah Brooks, who took the baton on the human-centered design work.
As it turns out, Mollie and Amber started the Human-Centered Design work at exactly the right time and then played their hand strategically. This is something we see consistently throughout all of these USDS stories: Someone sees a need and does what is within their control — and then a crisis happens that allows it to balloon into something more. But it’s not just the combo of the seeds and the crisis. It’s also knowing what to do when the opportunity arises and knowing how to connect those dots really quickly. And that’s what happened here: We had people at the White House, OSTP, OMB, and GSA who were helpful in amplifying it.
Kathy:
It’s about all that relationship building and hard, on-the-ground work to build trust. Emily, you talked a lot about USDS history. Do you recall there being an official launch or first day of USDS or the VA Digital Service?
Emily:
Yes, but first I need to share a bit more context, sorry! As all this human-centered design stuff was picking up speed in spring 2014, there had also been a series of conversations about what to do with the momentum coming out of HealthCare.gov. Mollie pulled me into a couple of early meetings basically talking about USDS: “What are we pitching and how do we pitch it?”
Meanwhile, Jen Pahlka was at the White House with a huge amount of experience from Code for America at the state and local level. She knew the potential of all of this and played a really strong hand in planting a bunch of these seeds to date. The Government Digital Service (GDS) in the UK had also just started, and Jen and a couple other people had been paying close attention. They seemed to draw a lot of inspiration from that, both in terms of how to set things up within the institutions, but also in terms of priorities.
At some point, it became clear that USDS was happening and they were going to start an agency-level team at the VA. Marina asked, “Do you want to help me set this up and join the team?” My answer, of course, was “yes.” At the end of summer 2014, as USDS was getting off the ground, we were laying the groundwork for hiring the VA team. I was not in those rooms, but I remember it wasn’t easy. Marina eventually got an agreement to hire 75 people. I am pretty sure this involved a few leaders being locked in a room until a solution was identified for hiring.
Through the fall and winter of 2014, interviews were happening and the first six people started on January 11, 2015.
We had an orientation in the basement of Jackson Place. I remember a couple of things very clearly: No working heat, welcoming people as they walked in, sometimes with really big hugs, and confusion. People were like, “Where am I and why am I here?” I don’t know, but I can offer hugs and moral support. That was the vibe.
We did our orientation in the basement, and the part I most distinctly remember was Weaver sitting us down and saying, “There are three things you need to understand. Keep these in your brain at all times.” These things have turned out to be 100% true, and I still think about them. Number one: There is no bottom. Number two: Someone tried that already. And number three: We are the problem. When I left USDS, I repeated those in my departure letter and added a fourth one: There is no top.
Kathy:
What do you mean when you say there is no top?
Emily:
You think people at the top of an organization have the power to change things, but they don’t have unlimited power. There’s only so much about what they say or do that affects the actual work — the implementation. The organization is the organization. The system is the system. And in order to make things move, you need to understand that system, its incentives, and how things work. To make change that sticks, the top is necessary, but not sufficient.
Somewhere in there we visited Deputy Chief of Staff Kristie Canegallo, who was the real-life Josh Lyman from the “West Wing.” Kristie was the person who USDS directly reported to, and was our point person at the White House. I was terrified. Kristie’s intimidating — but the good kind of intimidating, where you’re like, “I know we’re on the same team, but I’m a little bit scared.”
Kathy:
Same team… but you better come prepared as hell to your meeting.
Emily:
That’s right. So Kristie says how important the VA is and gets us pumped up. At one point someone (Jeff, I think?) asks her, “Do you have any advice for us?” Without skipping a beat, totally deadpan, she says “Don’t fuck up.” She started laughing, and as she’s laughing, she leans back and picks up a red phone. We’re like, “What in the hell is happening?” And she calls the Situation Room and asks if we can go check it out. We did, and it was surreal. But I think that “don’t fuck up” really stuck with us. The stakes were high.
Kathy:
How did things go from there?
Emily:
There’s so much, and it’s honestly such a blur. There were a few initial stated priorities, but as everyone jumped into the work, those had to change over time. One constant was that success at the VA was a huge priority for the White House. There were a lot of ups and downs, and I can now see how unbelievably challenging, and maybe crazy, it was to try to execute high-priority and high-stakes work while simultaneously creating an organizational structure that allowed all of these extraordinarily talented people to be effective. It was a lot, and it was a lot at the same time.
Eventually we found our footing, and the team was able to identify a few really critical projects where we could have an outsized impact. By the time I left in 2018, vets.gov was in a position to merge with va.gov, allowing us to transform the entire online ecosystem using the tried and tested methods we had used in the previous years. I think helping craft this merger was one of my last actions before leaving, which was oddly poetic because my time at the VA began in 2013, pre-USDS, on the team that ran va.gov.
Kathy:
Emily, what are you most proud of?
Emily:
That we built a team that has not only survived over three administrations, but has actually changed the VA and improved the experience of Veterans. What the VA is doing right now in terms of technology was unimaginable 10+ years ago… I mean I guess we did imagine it, since we wrote our vision down in the “glossy”… but did we actually believe it would happen within a decade? In 2014 there were thousands of websites, and countless mobile apps, phone numbers, and log-ins. Today the VA has among the most popular apps on the app store.
I’m also really proud of the work on va.gov. My entry into the VA was on the team that ran VA.gov, and the initial work that I did was to A/B test plain language on that website. Over nearly a decade, the VA’s online ecosystem transformed from thousands of websites, hundreds of phone numbers, and a literally unknown number of mobile apps into… ONE that is designed with veterans.
On a more personal note… I’m glad I stayed through the transition in 2017. We were able to get solid leadership into the agency, and I think the work that has occurred since then really speaks for itself. But I know that’s a different part of the history of USDS, so I’ll just say: I feel so unbelievably grateful for playing even a small part in this work. It was the privilege of a lifetime, and changed me forever.
Kathy:
Thank you Emily