Are you trying too hard to be original or the first and missing something glaringly obvious in the familiar or the mundane? Don’t gold-plate or make things unnecessarily futuristic. Pare it back. Simple is often the best.
Despite three years of research, it’s hard to say who should take credit. Facebook popularized the button in 2009, but its origins go back far earlier. Many people in Silicon Valley were working on a related set of challenges in parallel, like how to rank content (Everything2, 2000) or encourage user-generated content such as restaurant reviews (Yelp, 2005). They did so in an environment where informal sharing and learning across companies was common.
In solving these tactical challenges, the like button created several positive feedback loops that magnified its utility and inadvertently solved some bigger issues beyond the imagination of the early pioneers.
More recognition for content creators led to more user content. More feedback on who liked what content led to more targeted content feeds and increased user engagement. More “like” data meant better predictions. Eventually, the tech community recognized that this ability to attract attention and finely segment audiences according to their preferences and affiliations could create a better value proposition for advertising. It addressed a well-known quip about the industry: half of advertising was wasted, but it was impossible to know which half. So was born the advertising-based business model, fueling the rise of social media and digital advertising.
The impact of simple, inexpensive, analyzable, predictive, real-time feedback didn’t remain confined to social media. Online and offline retail businesses started to use like-type data to refine product portfolios and optimize promotional spend. Even industrial business began to understand their customers’ experiences more effectively.
Increased internet bandwidth supercharged the positive feedback, permitting engaging video content, and cheap, powerful consumer electronics that democratized content creation and created the fast-growing influencer marketing phenomenon.
The entire advertising and marketing industry was forced to reinvent itself due to the spread of the “like” button.
One reason for the meteoric rise of the like button was that it mirrors our neurological wiring. Studies of brain chemistry and anatomy confirm that the frisson we experience when liking or being liked – the release of dopamine in the nucleus accumbens – is the same mechanism that rewards real-world pleasures like social interaction or sex. The like button stimulates ancient evolutionary circuits tied to our survival and sociality.
Yale sociologist and physician Nicholas Christakis told us that what’s being rewarded and reinforced is a “social suite” of behaviors underpinning the human superpower of social learning – our ability to learn not only from direct experience but also by watching or listening to others. One key behavior in this suite is homophily, our preference for associating with and learning from people like ourselves. The like button encapsulates this perfectly, capturing both meanings of “like” – preference and similarity.
Another element of the social suite is our preference for mild hierarchy. Unlike dominance hierarchies in the animal world, which are maintained by violence or the threat of violence, humans tend to admire and affiliate with those others like to learn from. The counter beneath the thumbs-up symbol supports this tendency, providing a visible measure of social capital.
The familiarity of the pre-existing thumbs-up gesture also facilitated the spread of the like button. From gladiatorial depictions to paintings of hands in ancient caves, the human hand and our opposable thumb have been embedded in gestural language for centuries, albeit with shifts in meaning over time. No surprise then that the like button pioneers reflexively repurposed this.
We traced a web of such influences underpinning the development of the like button.
We were struck by how well the notion of social learning mapped onto Silicon Valley, which at the time was a giant collaborative learning lab overflowing with diverse examples of wins and losses to learn from.
Yelp was a pioneer of one-click reaction buttons in 2005. It was a watershed moment, removing the need for users to endure painfully slow page refreshes and URL redirects when expressing their preferences. It ensured that users could continue enjoying content without being disengaged. Russ Simmons, Yelp co-founder and the CTO at the time, coded the buttons using JavaScript. He only thought it possible because he had observed engineers at Gmail using JavaScript to embed different user features. We may never have had the joy of “liking” or being liked on the web had such knowledge been hidden in organizational silos. We traced a web of such influences underpinning the development of the like button.
There is a lesson here for all innovators: do not isolate yourselves – share your experiences and be curious about the experiences of your peers and rivals. Don’t assume you work in a unique context; it’s almost certain others are working on similar problems with similar solutions.
We conducted over 100 interviews with the pioneers of the like button. Not one of them foresaw its eventual consequences, or regarded it as more than a solution to one of the countless tactical design challenges they faced daily. It is surprising that something so world-changing could be born as the cumulative impact of many unremarkable moments of routine problem-solving.
We like to frame momentous innovations and discoveries as hero stories, where a single farsighted individual solves a well-defined problem in a flash of inspiration. But the process is often characterized by sociality, serendipity, delay, and unpredictability. Whether it’s the CAPTCHA software, the jet engine, or the miner’s safety lamp, multiple individuals tinkered simultaneously, building on each other’s efforts, recombining preexisting elements, often inadvertently solving problems they didn’t set out to solve. Reinforcing this messier view of innovation, a recent study showed that most scientific papers reported discoveries different from the ones anticipated in their grant applications.
A popular book among web designers and engineers, Steve Krug’s Don’t Make Me Think, influenced the early pioneers of the like button. Its central idea was that for innovations to spread virally, they should not seem novel or require any special cognitive effort from users. The like button’s success came about more through familiarity than novelty. It never needed instructions; instead, it tapped into deep human instincts. We are averse to friction: actions must be easy, social, attractive, and timely if they are to stick. It’s worth reflecting on what this means for your approach to innovation. Are you trying too hard to be original or the first and missing something glaringly obvious in the familiar or the mundane? Don’t gold-plate or make things unnecessarily futuristic. Pare it back. Simple is often the best.
It’s now apparent that the unforeseen benefits of the like button came with equally unintended side effects. How could this symbol of affirmation, rooted in our evolutionary biology, lead to consequences like digital addiction or self-esteem issues in teenagers?
The difference between online and offline liking is not one of kind but rather one of quantity. Real-world social interactions are limited by time and space. We can only interact physically with so many people per day. But digital likes can be given (or withheld) in higher quantities, at high frequencies, and without physical constraints. This can overload our dopamine circuitry, especially in young, developing brains. Ironically, by perfectly following the dictum of Don’t Make Me Think, the designers of the like button created something so frictionless that it could overwhelm the very instincts it tapped into.
The fact that the pioneers of digital liking foresaw neither its benefits nor drawbacks introduces a complex regulatory challenge. How do you regulate for unforeseeable consequences without stifling innovation? Regulation will always lag, compounded by the fact that regulators will often be the last to become fluent in a new technology, and that the science to sort out causality and test interventions takes time. The bumbling process of sorting through potential side effects and remedies may be more of a feature than a bug, as with the original innovation process.
With surprising benefits came surprising new problems, requiring new rounds of technological and regulatory innovation and retesting our capacity for social learning. The crucial difference with some digital technologies is that the spread may be faster and more pervasive than we are used to coping with, requiring innovation to speed up the regulatory learning process, through public observatories, adaptive regulations, or other means.
As the next technological wave of Gen AI takes off, the 20-year history of the humble like button has much to teach us about how innovation and regulation work: not as a clean, linear, intentional process, but as a messy, serendipitous, and deeply social one.
Martin Reeves and Bob Goodson are the authors of Like: The Button That Changed the World (Harvard Business Review Press, April 2025)
Ester Baiget, CEO of Novonesis, discusses how the Danish biosolutions powerhouse leverages modern technology and its 'library of strengths' to transform food production and washing operations....
Our world needs more scientific innovation, but founding new ventures can be challenging for deep tech scientists. Insights shared by those who have made this transition reveal the key lessons for aspiring...
Xiaomi, a company known for consumer electronics, launched its first EV last year. Just seven months later, it had delivered its 100,000th unit, one of which is driven by Ford CEO Jim...
The Swiss-based energy group is plotting a course toward a renewables-focused future, without abandoning its traditional customer base. Here we explore five key principles that have underpinned the company’s success....
Audio availableExplore first person business intelligence from top minds curated for a global executive audience