6

Trisector Entrepreneurs

A senior official in President Donald Trump’s administration told me, “If it can survive the Trump–Obama transition, it can survive anything.”1 I don’t think he meant it as an exaggeration. He told me this at around the same time the Washington Post reported on Trump’s efforts to “review, revoke and overwrite key parts of his predecessor’s legacy.”² After year one, the country’s forty-fifth president had issued seventeen executive actions to roll back rules and regulations championed by its forty-fourth. Ninety-six cabinet-level agency decisions had been promulgated to push back on the Obama era, and more were in the works. Trade. Environmental policy. Foreign policy. Where President Obama had zigged, President Trump now zagged. And yet somehow, this new piece of the bureaucracy, this part of the Obama agenda, had managed to survive. It still does as of this writing.

It was the United States Digital Service (USDS), and I had become curious about the new operation’s survival before Trump even announced his candidacy.3 The USDS had been born out of HealthCare.gov’s debacle of a rollout; a tech mistake that had threatened one of Obama’s signature policy achievements. President Obama and the people he brought in to rescue the downed site decided to start up a new agency—one that could draw into government the country’s best software engineers, designers, and related experts—with a goal of making sure something like that never happened again.

The US leaders had modeled it on a similar effort in the United Kingdom, the Government Digital Service, which had likewise been born out of a giant technology failure in the government health care space. I wondered if the USDS could survive past the burst of energy delivered by its founders in its early, heady days. Would it last? I had visited the USDS as it finished its first year, to ask its founders what they planned to do for its second, while Obama still had eighty weeks to go. What could they do to make sure their efforts to build more-agile government outlasted them? Three years later, the USDS remained. And the administration official was telling me that this was all very encouraging. Because the effort to modernize government would require a “generational” effort. And he is right that it will.

Will possibility last? Will the efforts that need to be rallied be rallied? The methods tried? The chances taken? I am hopeful. But the ties that have been built over the last decade between private entrepreneurs and public officials are threatening to fray during the next one. The reasons for that are complicated. One solution, I think, lies in the USDS and agencies like it around the world, for reasons that go beyond the improved services these agencies will deliver.

What most threatens the possibility project in the coming years is a brewing animus between techie entrepreneurs and government types. Some of that stems from a fear of what things like artificial intelligence can do and what government can do with such things. The protest of thousands of Google employees about their company’s involvement in AI on behalf of the US Department of Defense was but one episode that foreshadowed the growing tension. So were the Facebook hearings in the spring of 2018—a real Rorschach test for possibility if ever there was one. Techies thought it was government ignorance that was on full display; government types thought it was techie hubris. Things have not gotten better since.

What the USDS and agencies like it do in the face of all this is subtler than their main mission. Their main mission is to help their governments deliver software that works for citizens and under contracts that don’t swell to the hundreds of millions of dollars. But what they might do that’s as important, and increasingly so, is generate leaders with experience starting things in both the public and private domains. Techies come into government, and they leave, carrying the sensibilities of each sector to the other. There is a version of this that we should be wary of, yet another revolving door that enriches a few and fosters distrust. But there is another version, one that I am cautiously optimistic about, where what’s created with this kind of movement back and forth is a generation of entrepreneurs, of possibility artists, who have experience working in the private, public, and not-for-profit domains. If we can pull that version off and cultivate a new network of trisector entrepreneurs, we might get possibility that lasts.

Coming Together

The launch of HealthCare.gov was one of possibility’s dark moments. President Obama had swept into office under a banner of optimism. The audacity of hope. Change we can believe in. Yes, we can. And the president in his second term was indeed going to deliver what had seemed improbable—what had been impossible at least since the 1960s—and that was a sweeping expansion of health coverage that would bring the United States as close to universal coverage as it had ever been. On the day the site launched, Obama announced that HealthCare.gov would allow Americans to shop for health insurance with the ease of buying a book on Amazon.4 As it turned out, only six Americans were able to sign up for health insurance on the site that day, while tens of thousands were unable to access it.⁵ In the early weeks of October 2013, as the site was crashing around them, no one in what had been regarded as a tech-savvy White House knew the full extent or cause of the problems. In high-level discussions, leaders considered scrapping the original site (which had taken years to build) and starting over from scratch.⁶

The effort to rescue the site fell in part to a small group of people who hailed from the startup world, and the tech world more broadly. President Obama tapped Jeffrey Zients, the country’s first ever chief performance officer, to lead the operation to fix the website, along with Todd Park, the country’s chief technology officer. Park had been a health-technology entrepreneur (cofounding his first company at twenty-four) and then was persuaded to become the CTO at the US Department of Health and Human Services before serving in the same role on a government-wide basis. The two quickly pulled together a small, ad hoc team of highly skilled tech managers from inside and outside of government. These included Mikey Dickerson, a Google site-reliability engineer; Ryan Panchadsaram, who’d worked at Microsoft and Salesforce before becoming a Presidential Innovation Fellow; Jini Kim, a Google product manager; and a few others.⁷ Over the course of approximately two months, the team led the repair of the website (and its underlying IT systems) to the point where the vast majority of users could shop for health insurance online. The site was ultimately instrumental in allowing eight million Americans to sign up for health insurance in the program’s debut year.

HealthCare.gov demonstrated that a small group of highly talented entrepreneurs and tech workers from outside the normal government-tech ecosystem had the ability to fix problems and also revealed the misaligned incentives in the government’s IT-procurement process that had created HealthCare.gov-style implosions.8 Park had seen the potential for what this kind of talent could do in two other programs that had invited technologists and entrepreneurs into government: the Presidential Innovation Fellows program, which he had cofounded in 2012, and 18F, which some of those fellows had gone on to create in 2014 to extend its work. “The most important rule of startups is if you get the best people, you win. Full stop. These early programs were alphas for getting lean-startup talent and thinking into government,” he said. Park, with Obama’s support, refocused his time and energy on recruiting more of that kind of talent into government. Panchadsaram began coordinating teams to create space for a new bureau that could house this talent. Park persuaded Dickerson to lead the new bureau. Dickerson recalled, “I returned to California after working on the team to fix HealthCare.gov. I slept for a couple of weeks, and I began the task of processing what I had seen and done. I knew that HealthCare.gov was the most important work I had been a part of. I saw that technology in parts of government was in bad shape.” He went on, “But there was hope. When asked, some of the very best engineers and troubleshooters in the world willingly put their lives on hold to dedicate their time to this very difficult problem. When they got there, they found government officials and contractors who also wanted nothing more than to fix the site and who were ready and willing to work together to make it happen. There was limitless opportunity to do more.”⁹ Dickerson moved back to Washington to lead the new USDS. Erie Meyer joined him to get it started. She’d been the founding member of the tech and innovation team at the US Consumer Financial Protection Bureau, which was itself a startup within government, having officially been created in 2011. Haley Van Dyck was named deputy administrator of the USDS. Van Dyck was optimistic about what it would mean to bring new perspectives into government; both for the people that were brought in and for the delivery of public services. “The work of the engineers and designers that we’re pulling in from across the country will actually have a chance to influence and change the way government is operated,” she’d told tech journalist Steven Levy.10 The subheadline to the article, “Stock Options? Don’t Need ’Em! I’m Coding for Uncle Sam!” captured the brewing esprit de corps. By 2015, the new service counted more than a hundred employees on its team. It also began standing up digital-service teams within the Department of Homeland Security, Veterans Affairs, and other parts of the government.

This new government agency followed on models that had been tried elsewhere, and it has been emulated since. The United Kingdom formally opened its Government Digital Service (GDS) earlier, in 2011. It was initially led by Mike Bracken, who before that had been the director of digital development at the Guardian, in charge of its digital transformation. He’d been involved in several other technology ventures, too. One of the first projects the GDS took on was GOV .UK, a single landing page for all online citizen-government inter actions in the United Kingdom. GDS used lean-startup methodologies to iteratively build the site. In 2019, Senator Kamala Harris proposed a bill in the US Senate that would give state and local governments grants for standing up digital-service teams. California, Kansas, Colorado, Georgia, New Jersey, and Massachusetts were all said to be working to launch and sustain their own digital-service teams.¹¹

Other countries, cities, and states have adopted different models, but also with related goals—attracting entrepreneurial talent into government and inviting entrepreneurship from the talented people already serving there who may have felt it wasn’t welcomed or even allowed. John Paul Farmer, who had cofounded the Presidential Innovation Fellows program with Todd Park, was appointed CTO in New York City in 2019 and started recruiting for a variety of start-uppy tech roles: deputy CTO for digital, product manager, deputy CTO for innovation, innovation coordinator, director of NYC AI lab, director of NYC design lab, UX designer, director of research and future planning, and more. Shireen Santosham led a very active group of innovative thinkers as chief innovation officer of San Jose, California. In 2014, Rudi Borrmann became director of innovation and open government in Buenos Aires. Two years later he became undersecretary of public innovation and open government for all of Argentina. David Moinina Sengeh was appointed chief innovation officer in Sierra Leone—the first in his country—leaving his role working on AI research in Nairobi for IBM.12 All around the world and at every level, governments were bringing themselves closer to the skills and tools of modern entrepreneurship and to contemporary entrepreneurs.

And all around the world, great leaders were helping usher entrepreneurial and civically minded talent in their direction. Code for America, which Jen Pahlka had started in 2009, created fellowships that drew designers, software engineers, and their ilk into city governments. And before the organization had branched out in more directions, it had started a global trend. Soon enough there were Codes for Canada, Mexico, Ghana, Kenya, Japan, Germany, India, Australia, and others.¹³ And it had also started a generational cascade. Rachel Dodell and Chris Kuang, two college students, launched Coding It Forward to create pathways for young people to do something “mission-driven” with their technology skills and to make sure many of those pathways led into government. In the summer of 2019, fifty-five students from thirty-four colleges and universities in the United States spent ten weeks of their summer at six federal agencies “putting their software engineering, data science, product management, and design skills to use serving the American people.”14 The Ford, New America, and Hewlett foundations set out to cultivate a field of public-interest technology and to help universities work together to “equip tomorrow’s computer scientists, information architects, engineers, data scientists, designers, lawyers, policy experts, and social scientists with the skills to create public policy that centers the needs of people and their communities.”¹⁵

It was a bull market on “Tech for Good,” and it seemed likely that #techforgovernment was set to be a big part of that.

This rise of interest and activity was mirrored in the private sector. The preceding decade had seen a boom in private companies coming to market with “govtech” and “civictech” solutions. There were new private ventures trying to solve housing problems and security problems and climate problems and health problems. Two collaborators at MIT, architect Newsha Ghaeli and biologist Mariana Matus, founded Biobot Analytics. They figured out a way to measure opioids in sewage (excreted by humans) and map this data for cities looking to tackle the opioid epidemic in real time.¹⁶ And they figured that would be on the way to transforming wastewater infrastructure into “public health observatories.” They were accepted into the highly competitive accelerator Y Combinator. They got their first pilot with Cary, North Carolina.¹⁷ Seasoned entrepreneurs were flocking to the space as well. Ralph Clark took over as CEO of ShotSpotter in 2010.¹⁸ He’d spent his career in finance and high tech and had sold his cyber-security company. Why did he step into the role of leading a gun-fire detection and location provider? Why wade into a morass of new technical hurdles and government selling? “I liked the mission and saw that I could make a difference.”¹⁹ Many others were thinking and acting similarly.

Moreover, they found willing investors. Ron Bouganim, a serial entrepreneur who’d also been civically active, made the case to big venture-capital firms that they should be investing more fully in this market, and when they were slow to take him up on his suggestion, he decided to launch his own fund. Bouganim did 562 pitches in 2013–2014 and raised $23 million for “the first-ever venture fund focused on startups that are transforming the trillion-dollar global government-technology market.”20 He’s gone on to raise more. Shaun Abrahamson and Stonly Baptiste, two reformed govtech entrepreneurs, started their own firm in this direction as well. They launched Urban Us in 2012 and aimed to invest in hardware and software companies solving city problems.²¹ Julie Lein and Clara Brenner had created Tumml, a startup hub for urban tech, and then launched their Urban Innovation Fund. The Wall Street Journal observed, “More venture capitalists are seeking ways to support startups in sectors with heavy red tape.”²² And it wasn’t just the funds focused in these areas anymore: Andreessen Horowitz, Highland Capital, and General Catalyst were each in on the solve-public-problems action in some fashion. Vista, a private equity firm, scooped up GovDelivery and Granicus, two companies that helped governments inform their citizens. Berkshire Partners acquired Accela, a government permitting and licensing company. Government Technology Magazine catalogued more than six hundred unique investors—“from individual angel investors to large private equity companies”—investing in and shaping the govtech market.²³ I remember one thirty-day span in 2018 when almost $50 million was invested by venture capitalists in just three firms selling technology to government in the United States: $3.5 million for Seneca, a company that intended to help governments manage citizen complaints; $7.5 million for SeamlessDocs, a company that was helping government create web forms; and $38 million for Mark43, a company I had come to know that built police-records management software. It was a lot of money being invested in not particularly flashy stuff. All of this might have been kind of shocking, except that that spring looked a lot like the spring before it, when $25 million had been invested in Neighborly to build technology to help governments raise municipal financing, and $30 million was poured into OpenGov, a company that built software that helps with public budgeting and planning. These trends were evolving elsewhere, too. Daniel Korski had been deputy head of policy at 10 Downing Street and afterward launched PUBLIC with Alexander de Carvalho, a venture investor. They put together a suite of programs to help startups transform the public sector, which included advice and also capital, with an early focus on Europe and an eye toward the global govtech market.

I had set out in 2010 to chase possibility myself in Boston’s city government. I had joined Harvard Business School to teach it in 2014. I set out to write this book about it as the decade came to a close, and as I did, possibility seemed within our grasp. The image that all the activities—the digital services and innovation offices; the incubators, accelerators, and fellowships; even the capital investments—conjured up in my mind was of a wildlife crossing, those green underpasses or overpasses that allow animals to cross safely from one area to another. Habitat fragmentation, it is called when humans build roads and canals and power lines that divide wildlife habitats, and these bridges and tunnels are meant to alleviate the division. And I thought that whatever it was that had fragmented our own habitats—hiving off government from entrepreneurs—whether it was political philosophy or the march of technology or just the burdens of bureaucratic buildup over time, that all this work individuals were doing all over the world was giving us a means of finding our way back to each other.

Coming Apart?

Then I wondered whether it would be enough. Would it last? That had been my question of the USDS and, by proxy, of the movement, and there were worrying signs. Some of the reasons seemed more mundane and even fixable, and others were more profound.

On the more banal side, govtech-startup enthusiasm did begin to meet with the reality of starting up at all (the most likely outcome being failure) and of starting up to sell to government. One CEO of one of these types of companies told me, several years into the job, “You would have to be an idiot to start a company that sells to governments.” Another CEO was more generous, but clearly felt burdened by the high expectations that came with raising venture capital in this space (expectations for growth at high speed) and by the vagaries and vicissitudes of the government sales cycle. I wanted to beg both not to throw water on the whole enterprise in front of my students. Not because I wanted to hide the truth from my would-be public entrepreneurs, but because there were two truths. Selling to the government or starting up to solve public problems was challenging. And some companies were able to get quite good at it.

There are strategies for overcoming these more practical challenges and the frustrations that come with them. I have a few simple ones, and they come from my time observing startups in this space and especially from my time watching one of them: Shield AI, which was founded by Brandon Tseng (once a student of mine), his brother Ryan (a serial entrepreneur), and Andrew Reiter (a technologist).24 Brandon had been a Navy SEAL and, at one point, a human building-clearer. He’d seen one of the most dangerous operations in modern warfare—entering a building—and imagined a better way. He, his brother, and Reiter, along with the team they built, developed autonomous flying robots that could enter buildings first, self-navigate and map buildings, and provide video intelligence to protect service members and civilians inside. More than twenty venture capitalists at first declined to fund the startup. Several said they would if only Ryan and Brandon would pivot to oil and gas companies or to making selfie drones. Eventually the two found investors who bought into their plan to sell to the military, and then the task was figuring out how to do so. Ryan understood the challenges of selling to programs within the US Department of Defense and similar customers, but felt it was more doable than selling-to-government pessimists thought, and I feel the same way. He said it was a matter of getting down to the building blocks, and I agree. Almost all government buyers are undertaking three (often separate . . . yikes!) processes: planning, funding, and acquiring. Governments assess their needs. They source and appropriate money. And they choose among possible providers and ultimately deploy the new products or services. In each of these phases there are biases for the status quo that good providers can overcome. The status quo of endowments affects governments assessing their needs: they are likely to overvalue their incumbent providers relative to their merits.25 Therefore, new providers should work to deliver benefits well in excess of those of their often entrenched competitors, and those benefits should go well beyond just saving money. The status quo of the budget (it mostly rolls over from prior years, in many places) means that, among other techniques, strong providers must work to make sure the benefits they do offer get the attention of people in positions of budgetary authority. The status quo of behavior means that deploying any new product or service will run into serious obstacles if it requires public workers to alter very much how they do their work. New providers can seek to find ways to still deliver their new innovations, but should minimize the required behavioral change for public workforces. There are, of course, many other techniques that sophisticated government sellers are deploying, and I hope companies in this arena will get better at them. I do believe that one of the ways to scale possibility is to have companies build great products and services for governments and then to sell them skillfully.

Then there is, of course, frustration running in the other direction. I often advise companies selling to governments to realize that before you walked in with your really well-meant and well-thought-out and well-produced innovation, the official you are meeting with had a meeting with someone selling snake oil (maybe because the snake-oil salesman was related to someone “important”), and she has to spend all day keeping snake oil out of government, and she’s not wrong to think you might be selling it, too. Frustration was born of fraud. It was also born of instability. Among the companies started up in this new wave, many have come and gone. Neighborly—which was going to revolutionize municipal debt financing for localities, and which was part of the investment spree I mentioned above, and which had piloted in Cambridge, Massachusetts, in 2017 and in Madison, Wisconsin, in 2018, among other places—was on its way to dissolution by 2019.26 Some governments were being bombarded with promises of new this and new that (blockchain! virtual reality! machine learning!), and then jilted when the companies making these promises came and went. And frustration was born of hubris, too. Government officials were chastened by startups promising to solve complex problems with simple solutions. The officials were growing tired, to turn a phrase of a former US president, of hearing, “I’m from Silicon Valley, and I’m here to help.”

Above and beyond these growing pains of govtech selling and buying, an even more profound anxiety was settling in. Mass automation loomed as a potential jobs killer. Mass observation (by cookies, sensors, etc.) made an awakening public feel like it was living amidst “surveillance capitalism.”²⁷ People were worried that they were being exploited by their social-media platforms, and that their democracy was, too. And startup-driven ecosystems that had seemed a panacea a decade earlier (“we want to be more like San Francisco”) were adding to inequality in cities. Skyrocketing housing prices, rampant evictions, and rising homelessness in the Bay Area were three warning signs for cities elsewhere. There was a growing sense among some that technology was not the solution to our problems—technology was the problem. Public officials justifiably felt that entrepreneurial types who’d promised to solve the problems on their plates were filling them up with more.

If there was a growing enmity, the feeling was mutual. Tech workers were increasingly uncomfortable with the uses to which their work product was being put. Five thousand Google employees wrote to the company’s CEO that “Google should not be in the business of war.” They specifically singled out Google’s participation in Project Maven, an artificial intelligence effort. Ash Carter, a scientist himself and a former US defense secretary, published an open letter in reply. “Dear Googler,” it started, before making the case for Google’s participation: national defense is an “inescapable necessity”; AI is an “increasingly important military tool”; human involvement is still required “in any decision to use lethal force”; making sure that stays true and technically possible “takes specialists like you”; “your very survival rel[ies] on the protection of the United States. Surely you have a responsibility to contribute as best you can to the shared project of defending the country”; and so on.28 But AI for war wasn’t tech workers’ only beef. Amazon employees protested the company’s work with Palantir, a big-data company that had been working with US Immigration and Customs Enforcement (ICE). The protests came amid ICE’s detention of families at the US borders and raids in US cities.²⁹ The workers also called on the company to “stop selling facial recognition software to law enforcement.” Their 2019 letter to CEO Jeff Bezos concluded: “Our company should not be in the surveillance business; we should not be in the policing business; we should not be in the business of supporting those who monitor and oppress marginalized populations.”30 Amazon told its employees it would continue selling Rekognition to law-enforcement agencies. Then, in June 2020, the company put a one-year pause on letting police use the tool, “in a major sign of the growing concerns that the technology may lead to unfair treatment of African-Americans.”³¹ The company expressed hope that the pause “might give Congress enough time to put in place appropriate rules” for ethical use of the technology.³²

A bridge had been built between two worlds, and now it threatened to come down. There are no simple answers for how to make sure that it doesn’t or, conversely, for how to know when in fact it should. Governments will regulate technology companies in the ensuing years, but with which motives, and precisely how? Some wary techies will keep their talents from governments, but to what effect? And in many places, a circumspect public will ask them both to keep their distance from each other (facial-recognition bans are one example; so is the battle over encryption), but by how much? Some questions seem easier for me to answer. Yes, companies should help the United States protect itself from its adversaries. No, software engineers should not be compelled to help authorities separate parents from their children. No, they shouldn’t be violating civil liberties and human rights. Many questions can be answered only on a case-by-case basis. Take just one: The Waze app we came across in chapter 5 led an elderly couple into a Brazilian neighborhood where they became the victims of a deadly shooting.³³ In the wake of that tragedy, Waze could incorporate data about neighborhood safety into its app. It could give the app the technical capability to route people around “dangerous” places. But could the company do that without discriminating against particular neighborhoods (digital redlining) and their residents? Where should the company and the governments it works with strike a balance between safety and equity? Answers to that precise question and the thousands just like it raised by this modern marriage of startups and governments are mostly beyond the scope of this book, though we have addressed some strategies for beginning to tackle them: designing with users, testing and iterating, and mastering platform mechanics. And I do think there is one other strategy that can help.

Trisector Athletes

Joe Nye, a professor at Harvard’s Kennedy School of Government and its former dean, described a unique kind of leader with the ability to “engage and collaborate across the private, public, and social sectors.”34 He called them trisector athletes. What we need today more than ever are trisector entrepreneurs—leaders who can invent and build across all three sectors.

In 2013, Nick Lovegrove and Matthew Thomas, two private sector consultants who had also worked in public organizations, expanded on Nye’s definition. They pointed out that the world’s biggest problems all required government, business, and nonprofits to cooperate. They acknowledged that the challenges to developing trisector leaders were partly owing to “a widening disparity between business, government, and nonprofit incomes, the onerous confirmation process for senior government jobs, and the differing incentives and cultures of the three sectors.”³⁵ I’d say these obstacles are still present, wildlife bridges notwithstanding. Nevertheless, Lovegrove and Thomas argued that the world needed to nurture these types of leaders and that leaders could cultivate a trisector mindset in themselves. Aspiring trisector leaders could set out to balance competing motives, acquire transferable skills, develop contextual intelligence (seeing difference across the sectors and translating across them), forge an intellectual thread (I call this having an idea), build integrated networks, and maintain a prepared mind.36 Theirs is a wise list.

I shared it with my MBA students on a day when I invited them to consider their own career choices by stepping into the shoes of five students who had come before them. One of those five was Henry Tsai. He had been an early employee at a tech startup that was later sold to Yahoo. He had worked for Marissa Mayer. He then came to Harvard Business School, and I was fortunate to have him in class. He spent a summer during business school working in the San Francisco Mayor’s Office of Civic Innovation. On leaving business school, Tsai wondered whether he should return to private-sector tech or take on a more public role.³⁷ He ended up accepting a one-year fellowship in the office of San Jose Mayor Sam Liccardo. From there he went to Facebook, where he worked on the civic-engagement and elections-integrity team for several months, and then moved on to help manage the platform’s antiharassment strategies and tools. The kinds of questions Tsai wrestles with in that role are not the kinds that lend themselves to easy answers. But I’m supremely confident that he is better at working through them for the time he spent in government. Moreover, I suspect that someday he will find his way back to public service, and that if/when he does, he will be even better in that role for the time he spent at Facebook. Now, Tsai is a special person; by makeup he could never be the kind of person who simply leveraged his government role into a private-sector gig or who would reenter the public sector to do his former employer’s bidding. And I recognize that there are people like that, and that we should be on the lookout for them and even have policies that prevent them from abusing a revolving door. But I know Tsai takes questions of public value and public purpose into his private-sector work these days. And I know that if he goes back to serve publicly, he will take a sense of what is and isn’t technically possible and commercially sustainable into the public work he does, and that the people he serves will be the better for it. And I believe that if more people pursued careers like this, we all would be better off.

Trisector entrepreneurship is not a method for solving all the riddles, but of giving us at least a fighting chance at solving them well, and it’s something more than that, too. It’s a method for arriving at a life well and interestingly and meaningfully and adventurously lived. We can, by inventing and building in each sector, develop a set of skills for doing so across them. Then all we have to do is decide what to use those skills for.