Investing in Vices and Virtues

🎧 Listen to this article

This essay was first published by the Knight Foundation

Technology businesses create massive benefits for society. But when I started angel investing in the dot-com era and paying closer attention to why some websites grow quickly while others never really catch on, I soon realized that many of the most popular websites leveraged humanity’s less virtuous impulses.

Now, I’m not talking about porn or gambling websites, or multilevel marketing operations offering shady, get-rich-quick schemes. I mean mainstream consumer internet websites and platforms attracting hundreds of millions of users. As I later put it to the Wall Street Journal, in 2011, “Social networks do best when they tap into one of the Seven Deadly Sins. Facebook is [pride]. Zynga is sloth. LinkedIn is greed.”

The internet is hardly unique in this regard. At the same time the consumer internet was really starting to take off in the late 1990s, Fox News was learning how profitable it could be to not just cater to its viewers’ wrath and envy, but to inflame those passions methodically and persistently. HBO was relying heavily on sex and violence, via shows like Sex and the City and The Sopranos, to ensure its original programming appealed to its subscribers. Local TV news stations had been employing an “If it Bleeds, It Leads” ethos for decades.

What intrigued me about the new internet platforms emerging in the Web 2.0 era, however, was not just how they tapped into humanity’s innate and enduring appetites for the Seven Deadly Sins, but rather how these platforms could productively harness those appetites and transform them into more aspirational behaviors that could improve both individual lives and society as a whole.

In this regard, I was operating under the influence of Adam Smith. In The Theory of Moral Sentiments, Smith observed how human beings convert inward-directed impulses, like self-preservation and self-interest, into behaviors like sympathy, kindness and generosity, which ultimately advance the interests of society. In The Wealth of Nations, he explained how free enterprise creates structures and incentives that transform individual self-interest into widespread abundance, prosperity and other social goods.

So, apply that kind of thinking to LinkedIn. My theory was that greed could quickly attract enough users for us to develop revenue streams that would ensure long-term sustainability. But the real purpose of the platform was to inspire users to be more intentional about building the professional networks that positively impact both daily work life and long-term career development. This, in turn, would provide them with more economic security and autonomy while simultaneously enabling companies, industries and even entire geographic regions to operate more effectively. Ultimately, both individual users and society at large would benefit.

Twenty years later, I still strongly believe that internet social platforms can and should function in this way. But I also learned along the way that it’s a much, much harder and more complex proposition than I initially envisioned. In the early days of the Web 2.0 era, we may have aspired to the wisdom of the crowd. But the way things played out, we often simply got the madness of the masses. Learning how to effectively counteract such outcomes has been a learning process that continues even today.

The Challenges of Angel “Sinvesting”

Had I simply sought to profit from human weakness, I could have invested in casinos or cigarette companies. Instead, my intention was the opposite. I wanted to create something that could help humanity progress at scale.

As a graduate student in philosophy in the early 1990s, my plan was to pursue a career in academia and writing scholarly books about human nature and how we communicate with one another. But I eventually realized I could reach far more people through internet software platforms, in more substantive and enduring ways. A book that attracts a million readers, after all, is a rare breakout phenomenon. And most of that book’s readers will likely engage with it for only a week or two before putting it down forever.

In contrast, an internet software platform can attract hundreds of millions of people. Many of them will use it on a daily basis. For years. And because of its participatory nature, and how it connects users to other people, it becomes a more essential part of people’s lives compared to newspapers or TV shows.

It was this unprecedented capacity for scale – to impact huge numbers of people, in deeply engaged ways, over time – that made me believe internet social platforms had massive potential for improving humanity. And it was the same for virtually every other founder I knew in that era. Wild idealism was the lingua franca of Web 2.0. The general belief was that fewer gatekeepers, freer information flows, and peer-to-peer connectivity would inevitably create positive outcomes for individuals and society alike.

What we tended to overlook, however, was the role that basic human nature would end up playing in this process. Now, I certainly recognized that the fundamental drives and desires codified for centuries as the Seven Deadly Sins played a key role in how quickly users would adopt a given website or not. I even used them as one criterion for my investing decisions. In retrospect, though, what I didn’t sufficiently factor into my Seven Deadly Sins heuristic is that the Seven Deadly Sins are an enduring index of human behavior precisely because there’s so much gravity in the behaviors they encapsulate

The work also has costs that you, as a platform developer, could conveniently sidestep if you were catering to or even just passively benefiting from people’s negative emotions and appetites. Your economic returns will likely be lower.

All of these dynamics are especially in play on platforms that attract millions, or hundreds of millions of users. Because as platforms grow, and negative emotions and behaviors like fear, anger, greed and malice gain a foothold, the lowest common denominator often sets the tone. So, what starts out as digital Eden devolves into a digital hell. Adam and Eve, meet Mr. Serpent. Othello, meet Iago. Instead of a virtuous loop, you get a vicious one.

In addition, while exclaiming that LinkedIn is greed, Facebook is pride, and Zynga is sloth may make for the crispest aphorism, it overlooks the reality that people bring all their sins and negative appetites with them wherever they go.

For example, I initially believed that people were drawn to Twitter out of pride: You use it to let everyone know you are currently in the middle of dinner at the hot new restaurant everyone is talking about. But, over time, venom and outrage overtook humblebrags. So, wrath started to dominate as the primary sin there.

This dynamic held true for LinkedIn, as well. Before we launched, my initial theory was that greed would compel people to create highly detailed profiles. After all, the more comprehensive your user profile is, the more discoverable you are to other users – and that’s the key to maximizing your potential economic opportunities on the platform. As it turned out, though, most of our early users posted very perfunctory profiles. In fact, at one point in our first few years, the average profile length was just 90 characters – shorter than a tweet!

At the same time, people immediately exhibited a great deal of pride in their number of connections to other LinkedIn members, which we initially displayed on their pages. In an effort to convey status and authority through sheer quantity, our early users were amassing 1,000, 5,000 and even 20,000 connections. In the domain of LinkedIn, however, connections were intended as a proxy for trust and reliability, so more was decidedly not better. No one can have 20,000 professional connections whom they know well enough to credibly vouch for. That’s why, after extensive discussion and design thought, we eventually chose to cap the number we display on a user’s page at 500+. That allowed us to utilize a game dynamic that tapped into pride, but also to temper it.

What I began to learn from product evolutions like that, and can say now, with the benefit of hindsight, is that however hard I had anticipated it might be to convert emotions like pride and greed into more productive behaviors and aspirational identities, it has proven to be even harder in practice. To do so requires a deliberate, judicious, and consistently well-executed commitment to leadership and governance. And because of the crabs-in-a-bucket mentality that often gains traction as platforms evolve, the challenges of governance grow harder over time.

Even in a peer-to-peer world, leadership is important

At LinkedIn, we constantly faced trade-offs between catering to the Seven Deadly Sins in ways that would likely increase users and engagement or making choices that we believe would make the platform a venue for more aspirational behaviors.

For example, while LinkedIn was conceived from the start as a platform built on real identity, we initially decided not to let users upload portraits of themselves on their profile pages. Including that feature would have tapped into pride, lust and envy, and dramatically accelerated engagement. But we didn’t want our platform to become the workplace version of “Am I Hot or Not?,” nor did we want to make it easy for our users to engage in discriminatory behavior based on physical appearance. So, we didn’t let users upload self-portraits until much later, when this feature had become a standard element of virtually every social media platform.

We also had users asking us almost from the start for a channel or mode that would allow them to post work-related information anonymously. While such functionality might have yielded highly useful information about specific workplace cultures, salaries and even corporate wrong-doing, it would have also enabled the spread of unverified rumors and accusations, deliberate fabrications and libel. Above all else, we wanted LinkedIn to be a platform defined by trust, reliability, usefulness and civility. The best way to do that, we decided, was to only allow users to post information they were comfortable sharing under their real identities.

We also chose subscriptions rather than advertising as our primary revenue stream because it better aligned with our goal of being a time-saving platform, not a time-wasting one. Consequently, we never felt compelled to maximize time spent on site, and thus had no incentive to emphasize the kinds of controversial, provocative and high-emotion content that can dramatically increase engagement by playing to wrath, lust and the other Seven Deadly Sins.

But even as I take pride in these good decisions we made – see what I did there? – we also missed many opportunities to design for better potential outcomes.

During the Web 2.0 era, the reigning ethos was to minimize rules, institutional hierarchies and gatekeeping of any kind while enabling decentralized networks and peer-to-peer interaction. When we were designing communities then, we tended to prioritize the idea that every node in the network was equal.

And I don’t just mean the processes by which certain network participants emerge as de facto leaders in their spheres of engagement. A charismatic leader who encourages users to embrace pride or wrath, with no attempt to transform those impulses into positive behaviors, can do a lot of damage fast.

To counter this, platform developers should be extremely judicious about how they’re designing leadership roles and opportunities in a platform’s structure from the start, with well-considered processes, mechanisms and incentives for identifying, selecting and empowering the kinds of leaders who will help these platforms grow in ways that benefit the platform itself and society at large. Because when a platform attracts a million users, then 20 million users, then a billion users, it can’t function as an “electronic bulletin board” or a “digital public square” any more effectively than New York City or Tokyo can. It’s far more complex than that. It needs leaders and infrastructure that provide governance and facilitate aspirational behaviors and values.

This is something that we should have done more of earlier on at LinkedIn. As it happened, it wasn’t until well into our existence that we began to do so.

One example involves how we rolled out LinkedIn’s publishing platform, which now enables all our members to freely post long-form content that any other LinkedIn user can read. But when we first unveiled the publishing platform in October 2012, only 150 hand-picked business leaders, journalists, politicians and other noteworthy individuals could post on it. Obviously, highly regarded figures like Sir Richard Branson and Arianna Huffington, with well-established reputations for entrepreneurial innovation and success, helped attract attention to this new part of LinkedIn. Even more so, however, they helped create the norms and behaviors that we hoped all our users would adopt when we eventually opened up the platform feature to everyone.

Compared to other parts of LinkedIn, where profile templates and other interface cues provide guidance on what kinds of information users post, the publishing platform was fairly free-form. While we had terms of service and some editorial oversight, our user base numbered 178 million at that point. They would potentially be posting tens of thousands of articles and essays every day, and so the ability to abuse this new publishing capability was very real. What if our platform quickly descended into a venue that was mainly used for defaming former coworkers? Or spreading lies about competitors? Or simply posting content that had little to do with the business world?

To discourage such behavior, we only allowed our relatively small number of hand-picked influencers to post on the platform for nearly a year and a half. Their place behind the metaphorical velvet rope gave them power other users didn’t yet possess, but our belief was that they’d be highly likely to use that power responsibly. And, by and large, they did, writing in informed and civil ways about issues that were relevant to the business community. In doing so, they set the tone for the platform that others ultimately followed.

Ironically, though, “sin” was also part of this tactic’s success. To create a game dynamic among our designated influencers, we occasionally shared an internal leaderboard where they could observe who was posting most often, whose posts were getting the most views and shares, and so on. Naturally, pride and envy kicked in, which caused influencers to invest more time in writing and promoting their posts.

This had an obvious benefit for them: they could use our platform to share their ideas with millions of readers. And it benefited our platform overall, because it wasn’t just that we had designated a couple hundred prominent business figures as aspirational role models. These role models were participating on the platform in a highly visible and highly active way. So, when we opened up the platform to all our users after seventeen months of influencer-only usage, positive and healthy norms had been established.

Even if you somehow manage to get it right the first time, things will eventually change in ways that make additional adaptation necessary.

In this regard, designing digital communities at scale is no different than designing physical communities at scale. While small towns are often characterized as innately virtuous and wholesome, big cities are depicted as fundamentally wicked and sinful. But the reality is that the latter simply aggregate more people, and thus more human behavior, including those codified as the Seven Deadly Sins.

As entrepreneurs, designers, and technologists building digital platforms that significantly impact the lives of billions of people, we must take this into account. Indeed, if we’re not actively increasing our efforts to put checks on bad actors and lowest-common-denominator impulses through more effective leadership and governance, then we’re not just passively benefiting from humanity’s fundamental appetite for the Seven Deadly Sins – we’re passively optimizing for it.

Instead, we should be mindful of Adam Smith’s insights. Just as capitalism creates systems and incentives that transform individual self-interest into the wealth of nations – which can then be converted into education, healthcare and other social goods – internet platform builders have an opportunity and an obligation to convert humanity’s negative passions into more positive outcomes.

By continuously innovating and renovating the incentive structures and power structures that define these communities, we can both empower individuals and transform the madness of the masses into the prosperity and well-being of the crowd.