Hi I’m Matt Cutts, Administrator for USDS, AMA 🔥

@rrhoover I'd hate to pick between projects; I'm incredibly proud of all the work that USDSers have done and are doing, from changing federal hiring of experts to improving procurement to assisting doctors and patients to helping immigrants become citizens using better online forms. There's been so much work, from small businesses to the College Scorecard, which helps searchers get a better idea of which colleges will serve them well. But I will mention one project, just because it's the product of years of effort. A little over a year ago, the U.S. Department of Veteran Affairs launched an entirely new front door for Veterans at VA.gov. On the new site, you can do things like change an address and have that address change throughout dozens of systems at the VA. You can message with your doctor or refill a prescription. You can check the status of a disability claim or appeal. It's the kind of useful, friendly, user-centered site that reflects the best practices of the tech industry. But it can significantly improve life for the 20M+ Veterans in our country. The team that worked on that project highlights so many of the values that we want to reflect, like designing with users, not for users. I'm so proud of every project that USDS has done, but VA.gov is a perfect example that everyone can see of how government can work well. I'm incredibly thankful for the partnership we've had with the VA and that they've let us assist with their mission to serve Veterans. One lesson that I learned at Google was to avoid announcing new work before it's completely done. But we have been helping an agency with a bug bounty program, which is really exciting. We've also been assisting with a school safety resource, and I'm excited to see that launch (I hope within a couple weeks or so).

@ayyar government can definitely learn how to build things faster: think of minimum viable products instead of specifying all the requirements upfront in a waterfall-like way. Or the ability to move in an agile, iterative fashion. Or listening more to users to prioritize which features really matter the most and which aren't as critical. It's amazing how much impact you can create even just with a recording of a person trying to figure out how to use a government product or service. I also think some folks in government can worry about technology risk, even though many techniques like putting things in the cloud have been proven to be pretty safe and a wiser use of taxpayer dollars. Meanwhile, the tech industry can learn more about how to design for everyone, be more inclusive, and represent the entire diversity of the US.

@mattcutts thanks, very insightful. Regarding minimum viable products: any thoughts on how the government does / can adopt A/B testing and experimentation ? This seems like a key pillar of how the private sector learns and iterates. Also curious to what extent the government takes good software practices like unit testing and code reviews seriously?

@ayyar it is possible to do A/B testing and user experience testing. I love this blog post by Erie Meyer, for example: https://medium.com/@ErieMeyer/us... It's also true that the government lags behind in some areas like testing and code reviews though. I think one of the best ways to improve things is just to get more technical people into government, even if it's for shorter stints at first.

@ayyar @mattcutts Matt, your comment about the private sector is exactly why I just applied to the USDS. I'd love to talk shop and inclusion.

@hasan_diwan that's an interesting question. I view the goal of the US Digital Service as non-partisan and trying to make government work better for the American public through technology and design. That can be helping a Veteran apply for her health benefits or assisting a small business become certified to participate for contracts. It's vital work and we need more engineers, designers, and product managers to come help. For larger questions like this, it's better to look toward elected officials.

@mattcutts It's not a question of who to look to to sort it, it has not been done. And the fact that it has not been done is the primary reason my wife and I no longer live in the United States.

@abadesi well, I was fortunate to inherit an amazing team of fantastic individuals. The sort of people who apply to USDS tend to be the sort of people that want to work on important, impactful projects, and that's really inspiring. As far as concrete steps, we try to make our team culture accessible. So we try to have a range of events, from board games to bocce ball teams to pumpkin carving to movie nights to scavenger hunts to escape rooms to lasertag to activities that don't require alcohol, and we try to be mindful of parents with children. I hope that as a team we create a supportive, inclusive environment so people feel comfortable asking questions or voicing objections. We do a survey to assess our community health each quarter and people can provide feedback anonymously on that. We're still learning ways that we can be better. But it really helps to have impactful work. The day-to-day hours can be difficult, but the long-term sense of accomplishment tends to be really meaningful.

@emersondameron absolutely. One thing I've learned (but continue to keep working on) is listening. At Google, when someone had feedback--even angry feedback--it was important to listen for the kernel regarding what could be done better. Listening first before proposing solutions is also critical in public service. That can be listening to our partners in federal agencies, or trying to make sure that we bring in the voice of the user to give feedback on services. Another thing (which again, I'm still working on) is trying to give credit where it's due. In government, it's important to realize that most of the best ideas will come from civil servants who have been steeped in issues long before we arrived, and who will still be serving the public even after an engagement is finished. One big difference between the private sector and the government is that government serves everyone; you can't cherry pick a certain segment of the population like you can in the tech world or at a startup.

@jonathan_senin these are great questions! Typically I think of service design closer to design than product management, but it also feels like it sits at the intersection of those disciplines when it's done well. For me, it's mostly about mapping the user journey and those pain points, and then thinking about how to address those pain points, so that feels a little closer to design for me. But I think it could fall in either field depending on how you wanted to approach it. For your second question, there's a ton of newer companies acting as contractors for the federal government now. For example, there's one group at https://digitalservicescoalition... that you might check out. I'm not endorsing any companies, but that might be a good starting point.

@barneda I've been surprised how much of a difference it makes having a technical person at the table or when writing the terms of a contract (or when evaluating potential contractors). It's much easier to fix things upstream, like getting a good contractor on board, than it is to resolve things downstream. One of the harder things to solve in government might be hiring. For example, hiring technologists is really tricky, and we've met a few people who were skeptical of a particular proposal for how to hire differently. In that case, we not only brought a memo from an agency to say that it was okay, we actually brought a couple representatives from that agency in person! Once the relevant folks were assured that they would be allowed to hire in a new way, they signed up the same day!

@matthew_snyder That's a great question. America is a very diverse place, with many different cultures and perspectives, so it's important to us to include folks with lots of different backgrounds. We build better products when we better represent America. We try to recruit not just from Silicon Valley for example, but from across the country. I wish I had more resources at my fingertips that I could recommend.

@aleksandr_lanin you have to bear in mind with Medicare payments is that system is 40+ years old, running on a mainframe, and written in COBOL and assembly. So from that perspective, most of the technologies are an improvement! :) Almost 60 million Americans, which is roughly 1 in 4 people in the US, rely on Medicare to work well, so that's a huge impact. For our project on Medicare payment system modernization, our team at the Center for Medicare and Medicaid Services (CMS) worked with partners at CMS. We started by taking one chunk of the monolithic code (a pricing module), breaking that out from the mainframe, and actually putting that code in the cloud. Then we added a mechanism so the mainframe code could call that code in the cloud. The new code runs significantly faster. Modernizing a system like this will require multiple years of effort, but it's a pretty robust process to refactor or separate out a piece of functionality, rewrite it as clean code, and host it in the cloud. And as more of that functionality gets pulled out, it also provides an opportunity to collaborate with our federal partners so they build those skills/muscles as well.

@stewdog79 hey, good to hear from you! You were based out of the UK, right? Hope you're doing well!

@mattcutts thats the one... from the DaveN, EvilGreen monkey, phantomaster days etc. I changed from search... now building software. Best of luck to you sir with you new projects

@iamsarahcoloma these are great questions! In terms of criteria we use when doing project selection, we absolutely look at impact, but also things like executive support. We do ask whether a project is a good match for USDS' skills. We also ask ourselves "If USDS doesn't get involved, what's the likely outcome? Will this project fail if we don't get involved, or will it simply take a little while longer to be delivered?" Often technology is a common theme, but when you've got great user researchers and designers, it's perfectly fine to tackle a backlog or a process that might not have technology as the core problem. One of USDS' strengths is navigating between silos in government and getting different silos to talk to each other. We do try to work with our federal partners to make sure that we've got people who can ramp up with us, and then maintain and improve a project after our engagement ends. I'll be the first to admit that the hand-off of a project to an agency is often the most difficult step. It can be tricky to find the right balance so that you're not too load bearing on a particular engagement.

@amirkhawaja just jump in! Groups like Coding it Forward at https://www.codingitforward.com/ or Code for America brigades at https://brigade.codeforamerica.org/ provide a chance for college students and other folks to experience civic tech. I've been incredibly inspired by the people I've met from these areas, and often it feels like people come in expecting to do a short tour and it changes their goals for their entire career. Let me just say, civic tech is just at the beginning of this story. Most people are aware of their options to join academia or industry, but joining government is an option that more and more people are exploring. There's ways to help at the municipal and state level--including digital service teams in places like Colorado and California--and many countries are starting their own digital service as well! Lots of people have thought about money or compensation and how they want that in their life, but civic tech provides the chance to get a lot more meaning and purpose in your life.

@mattcutts Thanks for the response, I'll give those a look! Keep up the great work, you're a hero to us all!

@amirkhawaja Thanks! I'm just grateful that I get to work with a wonderful group of folks, including the federal partners that we collaborate with!