How could improving customer experience boost trust in government?
It doesn't stand still, but the government's level of service always seems to lag that of the private sector. According to Brookings Institution fellow Annelies...
Best listening experience is on Chrome, Firefox or Safari. Subscribe to Federal Drive’s daily audio interviews on Apple Podcasts or PodcastOne.
It doesn’t stand still, but the government’s level of service always seems to lag that of the private sector. According to Brookings Institution fellow Annelies Goger and Sha Hwang, co-founder of public agency contractor Nava, improving customer experience is essential to improving trust in government itself. And therefore the Biden administration’s latest executive order on customer experience can help. They joined the Federal Drive with Tom Temin for more.
Interview transcript:
Tom Temin: Ms. Goger, good to have you on.
Annelies Goger: Great to be here.
Tom Temin: And from the public agency contractor, Nava, co-founder Sha Hwang. Mr. Hwang, good to have you on.
Sha Hwang: Wonderful to be here.
Tom Temin: All right. So let’s begin with this thesis. The idea of the trust is, I guess we knew this intuitively. But a time when programs are getting bigger and more complicated. And some of the reports about pandemic spending are not all that favorable in terms of efficiency, there’s a need to at least restore some of that trust in government. And you feel that customer experience is foundational to that. Tell us more. Annelies?
Annelies Goger: Sure, and I think most people can relate to this in some way, whether it’s by going to the DMV or filing their taxes. But I think when the pandemic hit a huge amount of Americans that never before had interacted with safety net programs, like unemployment insurance or eviction assistance, were accessing those for the first time. So they were pretty shocked to see what it was like, and how much some of our systems have degraded with long term disinvestment. And I’m not just talking about technology, but also just processes that haven’t really been revisited for 20, 30, 40 years. And so, I could give one example of a restaurant worker that I interviewed, and she had lost her job with the lockdowns and she didn’t have a computer because she had cut off her internet to save money. So she tried calling the unemployment office here in D.C. And she had to wait eight or nine hours for them to pick up the phone. She tried multiple times over the course of a week, she finally got through. And what happened was the person who took her name spelled her name wrong.
Tom Temin: Oh boy.
Annelies Goger: And because of that, it put her in the queue for fraud detection, which took 12 weeks to get through. So she didn’t get her benefits for 12 weeks, you know, she’s worried about paying her rent, whether she has to move back to her parents, and she’s been working for years trying to get through college. And it was just mortifying really horrible for her to go through this experience. And it really undermines your trust that the government will be there when you are in a crisis.
Tom Temin: And that story underscores lots of technical issues that lead up to such a situation that are underneath the surface there, and Sha, your company has done a lot of work with agencies on this, maybe just take a quick minute to describe the company, a public benefit corporation, and a couple of the projects you’ve done that relate to this type of encounter that we don’t want people to have.
Sha Hwang: Well, thank you, Tom. So Nava is a public benefit corporation. So we work as a contractor to federal and state agencies. We got our start during the Healthcare.gov rescue effort in 2013. During that first open enrollment, where we saw a lot of the same issues that frankly, we still see today that other agencies and other programs still struggle with the delivery of critical public services under overwhelming demand. The challenge of integrating across multiple different legacy systems and things like that. So we partner with state and federal government agencies focused mainly on safety net programs to do so.
Tom Temin: So with respect to building a public facing application or adapting one, say in the case of the delivery of test kits by the post office, they simply adapted an application that existed, the application worked great. We don’t know how well the post office is going to work yet. But we know the application worked. And so what elements does an agency need to consider in architecting a public facing application they’re such that problems of spelling or name or identity and all of this can be avoided to the most extent possible? Sha?
Sha Hwang: I think Annelies and I touched on this a bit in our piece, the Brookings blog responding to the Biden’s executive order is, you can cut through so much of the noise and set things up for success. If you just do the simple thing of talking to the people who will be impacted by the services you’re designing and building that just radically reduces the kind of level of complexity that may be imagined when you’re kind of designing against requirements or designing against things without testing them in the field.
So I think some of the successes of COVIDtest.gov build on some of the practices that USDS who is involved in this effort has kind of been advocating for a while around kind of starting with users and starting where people are. Some of the other things I think Annelies and I argue are considering steps holistically. I think this also helps set up new efforts like COVIDtest.gov for success. Being able to consider the technology within the context of the broader kind of customer experience and thinking holistically. Understanding that in this case, there didn’t need to be an identity management step, you didn’t need to create an account to request these tests. And so then immediately you remove both a user experience bottleneck where people might fall out of the flow, and you also reduce a lot of load, you reduce the technical need to be able to support all of these workflows, or account management problems or things like that, that oftentimes come up. You can design something more like a grocery store experience than passing through security to board a flight type of experience. And I think that sensitivity is a customer experience perspective that has many kind of tech and architectural implications as well.
Tom Temin: Sure you can’t even buy razor blades nowadays without having an account somewhere. And sometimes you just don’t want an account. You just want the razor blades. We’re speaking with Sha Hwang, he’s co-founder of public services contractor Nava, and with Annelies Goger. She’s a fellow at Brookings, and Annelies, why do you think the government is always a step behind on these things? I mean, customer experience, and using the human centered design approach that Sha just described is almost intuitive or should be. So what takes them so long to get there?
Annelies Goger: Well, one of the issues is that government has a lot of different priorities to balance when they’re working with programs or service delivery. And so obviously, one of them is things like fraud, or maybe at a lesser scale of severity, somebody getting access to a program that really it wasn’t designed for. So a lot of our programs since the 80s, have really been framed around eligibility processes that are sort of measured and rewarded for letting the right people in and not letting the wrong people in. Because if you’re the public servant, and your program goes to the wrong person, your whole career could be in jeopardy, it could be a public scandal. And so I think there’s a culture of fear as a result of this and a culture of compliance-oriented thinking. And that undermines the idea of innovative thinking, taking risks, trying new things.
That’s one thing. And then I think a second factor that’s really important is the structure of how budgets are set up how procurement happens, it’s often kind of built for a previous era, when you did waterfall, big long term projects and not little bit by little bit by little bit, which is how the private sector generally develops tech right now. And so there’s this idea that technology is an afterthought that supports the program instead of you’re integrating the technology, the process and the user input all the way along the way.
And then the third factor is really just about a massive disinvestment over 30, you know, similar to our physical infrastructure, or our digital infrastructure, our programs have been not invested in and not maintained. So in many states, the underlying data systems are 30, 40 years old. They’re not designed for the internet, let alone for apps. And so it’s a question of staff capacity. Are there pay scales where you can recruit the right people? It’s also a question of things like rules from the Office of Management and Budget, for example, they have a law under the Paperwork Reduction Act that says you can’t interview more than nine people in a federally funded project, which is totally counter to the idea of getting user input, right? You really want to be able to interview more than nine people for one particular service. So it’s built for a different era. And I think part of the problem and I’m glad, one of the reasons I’m glad about this executive order is that maybe we can start to revisit some of those rules and think about them with the priority of the end user and reducing the burden on them. Instead of like, just thinking about compliance.
Tom Temin: And Sha, where you have encountered some of these monolithic systems, the VA operates a lot of them, HHS has them, the IRS, we all know about that one. They’ve been trying to modernize that one for 30 years, with assembly code actually running for their essential systems. Is there a way from your experience that those systems can be incrementally over time adopted to the new ways that are needed, such that you don’t have this sudden, waterfall lift and shift that we just, I don’t know how many times they try it, it never works?
Sha Hwang: Yeah, I think about it. Oftentimes, we talk at Nava. We’ve jokingly like to call a bit of our practice software archaeology, like there’s a bit of that spelunking that needs to happen to be able to understand not just the technical history or the technical kind of substrate, but also the organizational history. What were the decisions that led to systems being built in this way? What are the processes behind it? How much of these are regulatory or like hard requirements? And how much of these were to Annelies’ earlier point, workarounds for an era that we don’t exist in anymore? That we can kind of revisit a good example maybe is we did some work with the VA, Department of Veterans Affairs, on the veterans appeals process. And part of that work was about digitizing a form that VA caseworkers and attorneys were using and filling out manually. But when we did that user research and started interviewing stakeholders, we started to understand that one, all the VA caseworkers and stakeholders we were interviewing could find that data elsewhere. So the form, and we traced that form all the way back to regulations introduced in 1933. And we’ve found that the form did not need to exist anymore.
So that I think was a perspective where if we had purely taken a technology as the only solution lens, or using tech as a hammer or app development as a hammer, we could have created a digital process for a workflow that didn’t need to exist anymore. And instead, it became working with stakeholders at the VA and figuring out what were the kind of business process changes to simplify the experience and not have to actually build this in code. So I think that’s been a useful example of I think, the value or the investment of research and investigation and this kind of archaeology, that allows us to not just allow projects to get bigger and bigger so that they become too hard to handle or too hard to comprehend. And then they kind of are set up for failure, but kind of starting small and kind of pursuing a line of inquiry that allows us to design much more specific interventions that leave the agency in a better and more sustainable place long term.
Tom Temin: All right, so as we close, are there any examples you can point to where, hey, they did this really well in the federal government? Annelies?
Annelies Goger: Well, I wouldn’t say it’s federal government. But it’s local implementation of federal funding, which is the SNAP program. And so Code for America developed an app called GetCalFresh, which I think is a really good example of making a program application process easier over time. And so this is an app where you can apply for food assistance. And what they did were things like, let’s take the question of what’s your income, right? That’s actually really tricky to ask in a way that people with different types of income can understand and answer in an accurate and consistent way. And so they did a lot of testing, if I ask it this way, or this way. And so eventually, they get it at a way of asking the question that gets better data from it that’s more consistent, and then they can give the person a more accurate sense of whether their application is likely to be eligible or not for SNAP. And that kind of takes away some of the mystery and stress of trying to apply and figuring out like, how do I respond to this question that seems like Greek to me, which I think, is, unfortunately, a way a lot of our forms and applications are. And so this idea of having something you can do on a phone, you don’t need even necessarily internet, and over time, all the questions are getting better and better at really gauging what is comprehensible to like different types of people.
Tom Temin: Yeah, that’s a key point you bring out I was going to say, there is a large or a substantial portion of the population that does not have great internet access, or may not be able to afford the best iMac or tablet there is. And also, there’s a large percentage of the population that is unbanked. And somehow these applications, as you say, have to be able to take those people into account, who might be the most vulnerable of all and need these benefits more than anyone so that it can be done by telephone and Postal Service.
Annelies Goger: Mm hmm. Or just making a phone call and not having to wait nine hours to talk to somebody. Yeah.
Tom Temin: So that ties in the call center and the rules engine as well as the application, Sha?
Sha Hwang: Yeah, I agree. I think that’s an excellent example. And I think also, some of the consolidation around the Center for Medicare and Medicaid Services has been upgrading and modernizing Medicare.gov. And we’re actually called out in Biden’s executive order as well to just continue the work that they’ve been doing, kind of personalizing the experience. Where I think before, if you were on Medicare, you were managing up to I think 11 different accounts on different Medicare.gov type properties. But now consolidating that so that if you are on Medicare, you have one account, one experience, and that’s a place where you can actually see more personalized information about plans or prescription information that changes. I think that’s been pretty successful because they’ve thought about it from not the organization structure or the various kind of departments and offices within Medicare, but from the Medicare beneficiaries’ perspective about what are their needs and what are they trying to solve.
Tom Temin: Sha Hwang is co-founder of public services contractor, Nava. Annelies Goger is a fellow at Brookings. Thank you both for being with me.
Annelies Goger: Thanks for having me.
Sha Hwang: Thank you so much.
Copyright © 2024 Federal News Network. All rights reserved. This website is not intended for users located within the European Economic Area.
Tom Temin
Tom Temin is host of the Federal Drive and has been providing insight on federal technology and management issues for more than 30 years.
Follow @tteminWFED
