What the Like Button Can Teach Us About Innovation

In 2022 one of us (Bob) was going through the onerous process of moving from one house to another when he stumbled upon an old sketch in his files. It depicted a prototype for a thumbs-up like button, made when Bob was the first employee at the review site Yelp. Today the like button can be found everywhere on the internet. It has transformed digital advertising and marketing and helped fuel the growth of the $250 billion social media industry. Most people assume it was invented by Facebook, the firm that first used it at scale. But intriguingly, the date on the sketch (May 18, 2005) preceded the adoption of the like button at Facebook by nearly four years.

People tend to think that innovations spring from farsighted individuals who change history by creating new solutions to well-defined challenges. Yet here was Bob, standing amid his household detritus with a bemused look on his face and only a distant memory of his contribution to one of the most important mechanisms of the digital realm. Could it be that Bob had, in fact, invented the like button—and that he had simply forgotten about it?

The discovery of the sketch inspired us to have a day-long conversation, then start a three-year research journey, and eventually write a book, Like: The Button That Changed the World. Alas for Bob, our research was unable to identify a single clear inventor of the thumbs-up button, but that’s part of what makes its origin story so instructive. Indeed, the button’s design, development, and deployment could be considered a miniature case study on the true nature of innovation—which is distributed, unpredictable, and often far more modest in ambition than heroic innovation narratives suggest.

Many organizations like to believe that innovation can be channeled along a pathway from concept to implementation that’s been tightly defined by smart managers. That’s a reassuring perspective because it feels empowering and calls for companies to do what they do best—allocate their resources and design and control processes. But the origins of the like button reveal that innovation tends to be serendipitous and social in nature—a melding of accidents, ideas, iterations, false starts, missed opportunities, and unexpected outcomes that cannot be willed or cajoled into existence. While not entirely new, this understanding is hard for companies to internalize. But in an era when artificial intelligence and other technological advances will upend business models, the wild story of the like button holds key lessons for managers who seek a better way to encourage and harness innovation.

A Collective Discovery

When Bob sketched a thumbs-up button in 2005, Yelp was a small, scrappy startup trying to solve a very specific problem: The company needed to persuade a volunteer army of consumers that writing reviews, without pay and for the benefit of strangers, was a good use of their time. The first version of Yelp, which relied on requests for help from users’ friends, wasn’t achieving this. The founders, Jeremy Stoppelman and Russel Simmons, suspected there might be a way to motivate people to write reviews just for fun. Their hopes for this were raised by observing the unusual behavior of two users, Kevin and Stella, who were prodigiously submitting review after review on the website. Jeremy reframed the challenge as how to encourage more Kevins and Stellas. To do that the Yelp team created a feedback mechanism called “Send a compliment.” It was a way for users to express their appreciation for reviews and in turn encourage writers to submit more. The new feature worked, but the team wanted to figure out how to make it as frictionless as possible. It was one of many tactical challenges facing Yelp, and there was no sense that it was special in any way.

The thumbs-up icon that Bob suggested sparked debate. The team felt that people responding to reviews had different reasons for liking them, so they implemented a system in which users could click one of three buttons: useful, funny, and cool. When the founding group passed on Bob’s thumbs-up sketch, he filed it away and forgot about it. Yet the alternative design that Yelp settled on was still a crucial contribution to the digital economy because there was nothing quite like it yet on the internet. By allowing users to react to other users’ online content, Yelp showed that peer interactions could be a powerful driver of engagement. It proved that social validation and positive feedback could motivate people to contribute content and participate actively in a community—two things that fuel social media to this day.

But in fact, Yelp’s user-feedback innovation wasn’t really the first crucial breakthrough, as we discovered in our research. In 2000 the somewhat prurient website Hot or Not went viral, rivaling major news networks in page views. The site allowed people to upload photos of themselves in order to let users decide if they were physically attractive. Its main innovation was that it enabled users to react to content without forcing what was then a time-consuming page refresh with every click. It thus removed the friction from its rating process—a key step toward being able to simply click “like.” Does that mean that Hot or Not’s founders, James Hong and Jim Young, should take credit for creating the like button? When we interviewed Hong, he wasn’t so sure. “We definitely invented it independently,” he said, “but I could never say with certainty that we were the first.”

The button’s design and development could be a miniature case study on the true nature of innovation—which is distributed, unpredictable, and often modest in its ambitions.

Hong is right to be humble. After three years of research, neither of us could pinpoint where the credit for the invention of the like button should go. There are a number of contenders. The popular video recorder TiVo, which was launched in the late 1990s, had a remote control with physical buttons users could hit to express their content preferences, and those buttons were in the shape of little hands—on the right, a green thumbs-up, and on the left, a red thumbs-down. In the online sphere, websites from the early 2000s such as Digg.com (a news aggregation site), Everything2.com (an online user-generated encyclopedia of sorts), and the early blog platform Xanga all introduced ways for people to endorse content they enjoyed. It seems that many players had simultaneously converged on solving the common challenge of creating an individualized digital feedback mechanism, with absolutely no inkling of how central the like button would become in social media.

What is certain is that at the company that would eventually make the like button a global phenomenon, the CEO opposed its introduction for several years. In the mid-2000s, Facebook’s Mark Zuckerberg repeatedly overruled proposals to introduce the button, worrying that it would cannibalize engagement. He relented only in 2009 after the like button had proved successful at a company called FriendFeed, which Facebook later acquired. Facebook’s adoption of it was an immediate success and was soon copied in different forms by YouTube, Instagram, Twitter (now X), and virtually every other social media website.

Misplaced credit is common among game-changing innovations. When introduced, they often fail to catch on widely until a later version from another player popularizes them—often, as in the case of Facebook, with little foresight about the impact they’ll eventually have. Another classic example of this is IBM’s Simon Personal Communicator. Introduced in 1992, it combined a mobile phone with a personal digital assistant and included features like a touchscreen and email capability. However, it wasn’t until Apple’s 2007 introduction of the iPhone that smartphones took off.

The murky origin of the like button illustrates how innovation often emerges from unplanned cooperation across a distributed network of contributors. Each player implemented its own variant of the button, sometimes incorporating elements invented by the others. This story also highlights the power of clustering. In Silicon Valley the proximity of many tech startups and the culture of openness created a fertile environment for the exchange of ideas. Pioneers in the industry were keenly aware of and inspired by their competitors, sharing know-how at frequent meetups and informal discussions. And the story also suggests that sometimes innovations are simply “in the air” and ready to be discovered. Isaac Newton and Gottfried Wilhelm Leibniz, for instance, developed calculus around the same time, despite being in different countries and working independently. Similarly, Max Levchin and David Gausebeck at PayPal came up with a novel way to fend off fraudsters—one of the first commercial uses of what’s now called CAPTCHA technology—unaware of similar academic work by Luis von Ahn’s team at Carnegie Mellon University.

But perhaps most important, the like button’s story demonstrates the disorderly, evolutionary, and iterative nature of innovation. The various companies experimenting with it created a myriad of designs and functionalities. Far from being the result of a single “Eureka!” by an individual, it emerged from a series of small improvements made by different teams working separately yet influenced by one another’s progress.

Innovation happens like this all the time. And as we have noted, people don’t necessarily see the ultimate prize at the outset. Sergey Brin and Larry Page’s development of PageRank, which underpins Google’s search engine, is a good example. At the time the prevailing way to automatically assess the quality of a page of content was the number of hits—that is, how many visits had been made to it. Taking a lesson from the world of scholarly citations, Brin and Page’s insight was that the number and quality of other pages that linked to a page also mattered. The Google founders knew that finding a better way to rank page quality was important because marketers were starting to run banner ads on pages. But unexpectedly, PageRank really paid off by also providing the basis for a superior search engine and a way to monetize search. It’s no exaggeration to say it was a trillion-dollar innovation, yet Brin and Page weren’t envisioning Google’s eventual core business model when they focused on page ranking. They were simply trying to solve the next technical problem in front of them.

We’re not the first to point out the messy, unpredictable nature of innovation. Drawing on the history of technological innovation, the economist Brian Arthur, in particular, has described the incremental, cumulative, and often-unintended path that most groundbreaking inventions follow. Yet acknowledging that innovation is nonlinear has done little to diminish the appetite of many companies—particularly large established ones—for trying to force the process on to metaphorical assembly lines and tightly manage it.

There are several good reasons to revise the innovation practices at large companies to reflect the reality we describe. Growth rates are trending down in most economies, putting greater emphasis on the need for new ideas. A return to higher interest rates also means that innovations need to perform better to be financially viable. A managerial approach that addresses innovation’s true nature—social, iterative, nested, and serendipitous—will be essential to address those imperatives.

Here are five ways that companies can improve their innovation models and practices.

Be Alert for Surprises

Typically, the idea for an innovation is sparked by an anomaly that alerts companies to new opportunities or ways of doing things. It might be a fringe group of consumers who have devised their own inventive solutions or a maverick competitor that has taken an unusual approach. Unfortunately, because companies have a tendency to become more inward-looking as they grow, many firms fail to spot these anomalies.

Some tried-and-true ways to strengthen your firm’s external orientation and discover such catalysts include:

Measure it. Look at how many mentions of recent trends, competitors, and customers come up in a random sample of company meetings. If these topics aren’t being covered regularly, put them on the agenda.

Model it. Have senior leaders spend a large proportion of their time with customers. This will encourage employees to behave the same way. Some companies, like the global professional services firm Genpact, mandate that senior managers spend a minimum proportion of their time in the field.

Nourish it. It’s striking how often the pioneers of the like button referred to meetups or restaurant conversations at which people in the industry discussed common challenges and recent developments and how aware they were of the activities of others in their field. Your team members should participate in or even organize such exchanges.

Reflect on it. In any industry there are always maverick outsiders who are betting against the incumbent’s business model. Figure out who they are, why they’re doing what they’re doing, the consequences of their being right, and how those insights could fuel your own innovation efforts.

Take Things Step-by-Step

The retrospective accounts of major innovations often make them sound as if they were all great leaps toward clearly laid out goals. Close historical examination, however, reveals the many incremental moves that lead to pivotal inventions.

Ways that companies can embrace a more evolutionary approach include:

Avoid making a master plan. It’s striking that none of the pioneers of the like button had any sense of its eventual significance or even that their efforts with it were remarkable in any way. It didn’t emerge from the search for a revolutionary new monetization model for the social media industry; it resulted from efforts to find ways to get consumers to submit more reviews, express their preferences, and interact easily with content.

Conduct trials sooner. Grandly conceived innovation projects often are very expensive and therefore get subjected to lots of scrutiny and analysis. But eventually all new ideas have to be tested against reality, and the earlier that takes place, the faster they will progress. As executives at one software company told us, the best ideas are not the ones that initial analysis said would be winners but rather those that had the most opportunities to evolve.

Be receptive to unintended outcomes. We were struck by the predominance of both positive and negative unintended outcomes in all stages of the like button’s story. Capitalizing on the unexpected requires close observation and an open mind.

Persist. Innovations are rarely “one and done,” even if post hoc accounts present them that way. We can see this not only in the origins of the like button but also after it began to take off: in the development of sophisticated back-end software to analyze likes, the creation of new user-engagement metrics, the emergence of a new monetization model for the social media industry, and more. Each interim solution begets new challenges and opportunities and new risks of ceding ground to competitors who take them up, requiring continual efforts over time.

Lay the Myth of the Solo Inventor to Rest

Some individuals, like Steve Jobs, have played an outsize role in innovation, but it’s often exaggerated in the retelling to satisfy our appetite for hero’s journey narratives. In our deep dive into the history of the like button, we found that many unsung heroes contributed significantly to its development. Innovation is a social process within companies, requiring visionaries, contrarians, scientists, engineers, systematizers, communicators, champions, and more. It’s also a social process across companies, which build on and extend one another’s ideas.

Here are some ways in which companies can better tap into the social aspect of innovation:

Recognize multiple heroes. Be sure to celebrate the contributions of all the individuals involved in innovation. Create a culture that encourages everyone to participate in it, as the Japanese business services company Recruit does. Recruit has established a contest called Ring, which is open to any employee. Participants come up with out-of-the-box ideas for new lines of business and then pitch them to the company’s executive team. The program has evolved to include workshops and tools that help people search for teammates, draft proposals, and solicit executive cooperation.

Promote cognitive diversity. Teams that encompass a variety of mindsets generate better innovations. So hire, know, protect, and encourage your mavericks, visionaries, critics, theorizers, tinkerers, testers, system builders, storytellers, and innovation evangelists. Examine whether your firm recruits, rewards, and promotes one type of personality or cognitive talent predominantly. Create a culture that recognizes the necessity of multiple types of talent, and equip people with the awareness and skills to work with colleagues who are different from them.

Practice humility. Much of the innovation necessary for the future growth of your company will probably occur beyond its walls. Recognizing this, especially after experiencing some measure of success, is the first step toward tapping into broader pools of insight and talent.

Calibrate Managerialism

As firms grow they have to put in place structures and processes that both exploit their early entrepreneurial success and manage their increasing scale. Though much decried, this bureaucracy serves an essential function. But in some places the effort to systematize, standardize, and streamline can go too far.

That would be less of a problem if entrepreneurial success were long-lived. But that’s currently not the case, given the rapid pace of technological change. We were struck in our research by how often firms achieved innovation breakthroughs in spite of their declared strategies, policies, and intentions at the time. In predictable, “engineered” environments like factories and HR departments, clear goals, plans, and policies and their decisive, consistent implementation make perfect sense. But how can companies leave enough wiggle room for the serendipity of innovation?

Here are some ways to achieve the right level of managerialism in innovative activities and balance it with entrepreneurship and flexibility:

Foster curiosity. Create a culture that is bottom-up and outside-in. A good way for leaders to do this is by demonstrating curiosity to employees and asking questions rather than giving instructions.

Avoid the tyranny of averages. If only cumulative performance totals or averages are reported, then current beliefs and policies are likely to be reinforced, and initiative is likely to be discouraged. A focus on anomalies and individual cases, by contrast, will raise questions and possibilities that may lead to innovations that change the status quo.

Legitimize inconsistency. What if instead of thinking of your current goals and policies as fixed and unchangeable, you saw them as your best guess at the time—as temporary plans that are open to revision? With such an attitude, your organization is more likely to revise and pivot when faced with novelty or unexpected findings. Innovation is not just about new things but also about new ways of doing and thinking about things. As the Chinese e-commerce giant Alibaba has asserted, everything in the company should evolve in the face of the changing external market reality.

Moreover, as Tom Ward, a Walmart and Sam’s Club executive, told us, “If you’re trying to build something that’s perfect, you’re going to have 10 different versions of what people think perfect means, and you’re not going to deploy anything. Instead, we get something out there and iterate.”

Shift Your Mental Model

Most fundamentally, effectively navigating the messy reality of innovation requires a new kind of business logic. In mature companies the making and execution of plans drives most operations, and progress is measured by achievements toward relatively unchanging goals. This approach often spills into unpredictable and creative activities like innovation, where it isn’t that effective. The alternative is not to replace planning with passivity and chaos, however, but to supplement it with evolutionary and narrative thinking.

You can facilitate that shift if you:

Learn from the past. The theories that inform business are often didactic, prescribing how innovation efforts or other aspects of business should operate, with the assumption that business reality can be molded to our will. When this is not the case, detailed business histories, such as that of the like button, can highlight the serendipitous and unpredictable patterns that companies need to understand and adapt to.

Embrace storytelling. For a predictable, quantifiable problem like factory management, managers can use a deductive approach to figure out solutions. With innovation, however, it’s often unclear why consumers do what they do, what they would prefer, or why they have certain reactions. To make sense of such unfamiliar and complex situations, managers can construct a variety of narratives that explain what might be going on, gather new information, and test the narratives out to determine which work.

Model and train people in the new approaches. The leader of one tech company asks the people who propose new ideas the following questions: What will the innovation improve, and what formula will they use to measure its success? What changes will it require? What A/B tests could they run on it? What might the data say about whether the innovation should be adopted or developed further? He is modeling evolutionary thinking and structuring the decision-making behaviors of the firm around it. Similarly, the R&D leader of a major pharmaceutical company asks pioneers of novel drug candidates to explain a plausible vision of what success could look like. He is modeling the narrative approach for upstream innovation, where deduction would be artificial.

. . .

The messy invention process revealed by the history of the like button turns out to be not an exception but rather the dominant pattern of innovation. As businesses contemplate the new possibilities that can be unlocked by artificial intelligence, they should think about how they’ll adapt their own approaches to innovation to reflect this reality. And now that you know about the origins of the like button, perhaps your day-to-day use of it will serve as a reminder of the mental model of innovation you should embrace for yourself and your organization.

This article is adapted from Like: The Button That Changed the World (Harvard Business Review Press, 2025), by Martin Reeves and Bob Goodson.

Copyright 2025 Harvard Business School Publishing Corporation. Distributed by The New York Times Syndicate.

Explore AAPL Membership benefits.