For more than two decades, digital platforms have largely been built around a powerful idea: Be first, grow fast, tip the market, and let network effects do the rest. A platform that attracts enough users can trigger a flywheel effect, leading one company to dominate — often profitably, sometimes spectacularly so. Companies such as Amazon and Google came to exemplify this model, and for a time almost all the world's most valuable firms were platform businesses — firms that facilitate interactions and exchanges between two or more distinct, interdependent groups. (See "Products to Platforms: Making the Leap," HBR, April 2016.)
That playbook is now breaking down. Many platform leaders are coming to terms with the fact that growth is rarely easy to sustain and that competition often persists even after players achieve large scale. Network effects matter, but they are rarely strong enough on their own to deliver winner-take-all or winner-take-most outcomes.
At the same time, the broader context has changed. Investors are less willing to subsidize losses indefinitely, and expectations around profitability have become more demanding. Even when platforms reach millions of users, their growth is increasingly shaped, and often constrained, by regulation, privacy rules, and antitrust scrutiny.
The result is a new era for digital platforms, one in which companies must rethink how they grow, compete, and make money. This is an exciting time for platform businesses because there's more room for competition and innovation. Markets that once seemed to have unassailable leaders are becoming more contestable, opening the door to new entrants, new business models, and new forms of value creation. One example is regional ride-hailing platforms, which continue to operate successfully alongside global players like Uber. Another is Airbnb, which has expanded the accommodations market without displacing hotels, while hotel companies have found new ways to create and capture value through retail media and other digital services.
I have studied platform growth and regulation for more than a decade, and across multiple settings I've found that the realities of platform competition differ from many of the assumptions that shaped the early playbook. The lessons that have emerged are nuanced: The strength of network effects is often context-dependent; different users value different platform experiences, not just the benefits that come with size; and growth strategies that are compelling in theory do not always produce winner-take-all outcomes in practice.
This article draws on those lessons to expand the strategic lens for platform leaders. It explains why network effects are rarely sufficient on their own, describes how regulation is reshaping the environment for growth, and presents four strategies for growth that new and established platforms alike can apply in response. More broadly, the article reflects on platform strategy's evolution from focusing on network effects and market tipping toward more richly understanding how platforms compete, grow, and create value over time.
The Limits of Winner-Take-All Thinking
For years, the belief that network effects naturally lead to winner-take-all or winner-take-most markets has shaped strategy and venture investment across tech. (See "Pipelines, Platforms, and the New Rules of Strategy," HBR, April 2016.) Network effects are important, but giants such as Google and Amazon achieved a spectacular level of dominance in their respective industries because several complementary forces aligned to create a broader set of advantages. As these platforms grew, the cost of serving each additional user fell dramatically — economies of scale. At the same time, their incumbent status helped them expand into adjacent products and services more efficiently than new entrants could — economies of scope. (See "Are You Missing Growth Opportunities for Your Platform?" HBR, May–June 2025.) Much of what the giants offered could be delivered at near-zero marginal cost, especially in digital contexts. And the data they collected from users improved the services they offered in ways that further strengthened their advantage. Together, those dynamics created a loop in which scale allowed the platforms to provide better performance at low or no price, which in turn attracted more users.
Amazon illustrates how these forces come together. As more buyers joined the marketplace, it attracted more sellers, strengthening the network effects. But the company's advantage did not stop there. Its scale allowed it to build highly efficient logistics and fulfillment networks. Those same capabilities could be leveraged across multiple product categories, spreading the fixed costs. Meanwhile, Amazon's data on consumer behavior enabled it to make better recommendations, provide more-effective pricing, and deliver more-targeted advertising, further improving the user experience. Each of those elements reinforced the others, making Amazon's position difficult to challenge.
When such forces align, markets can tip decisively. But the giants' success can mask the fact that this kind of alignment is rare — that true winner-take-all outcomes are the exception, not the rule. A case in point is the accommodations market, where multiple types of platforms exist. Traditional hotel groups like Marriott are platforms in that they don't own and manage most of their properties but do own their brands, loyalty programs, and booking services. These hotel groups compete not only with one another but also with a persistent fringe of independent hotels. In addition, when people book stays through intermediaries, hotel groups share revenue with online travel agencies such as Booking.com, Expedia, and a range of regional and local alternatives.
Over the past decade, a new type of rival has emerged in short-term rental platforms like Airbnb. These companies have created substantial value by expanding supply and increasing flexibility. While they have undoubtedly put pressure on incumbents, they have not displaced them. Many travelers still value the consistency and reliability of hotels, particularly for business travel, and remain wary of the risks associated with staying in a stranger's home.
The broader story in the accommodations industry, then, is not about a single player taking over but about multiple models coexisting. Even in the presence of strong network effects, no single platform has "won" the market, suggesting that dominance is far from inevitable.
One reason is that network effects are often local rather than global. Ride hailing provides a clear example. While Uber and Lyft compete in the United States, different players dominate in other regions — Didi in China, Grab in Southeast Asia, and Ola in India. Because ride hailing operates city by city, a platform does not need global scale to succeed; capturing a handful of local markets with sufficiently dense participation from passengers and drivers can be enough to sustain a business. The same is true in settings from waste management to labor platforms, where scale in one geography does not automatically translate into advantage in another. (See "Why Some Platforms Thrive and Others Don't," HBR, January–February 2019.)
But research that I conducted with colleagues points to another reason winner-take-all outcomes are hard to achieve: Users have different preferences, so they do not necessarily gravitate to the largest platform. In a study of a merger between the two largest pet-sitting platforms, we found clear evidence that having more users did lead to creating more matches, yet the benefits of network effects were largely offset by users' preferences for one platform over another. Indeed, when the platforms merged, users of the acquiree were forced to switch to the acquirer's interface and user experience, while the acquiree's features disappeared. Many users left after the merger, revealing that they valued design, brand, routines, or existing relationships more than the benefits of sheer size.
The implication is that even strong network effects may not be powerful enough to overcome differentiation. Platforms are not interchangeable utilities; they offer distinct products and services. And users care deeply about how those offerings fit their specific needs and preferences. That helps explain why platform growth has often come through brand acquisitions, particularly in sectors like online travel and hospitality, which have preserved the appearance of choice while concentrating control over data, pricing, and market access behind the scenes.
What's Reshaping the Game
Platforms face a changing environment in several ways. First, regulation is increasing. In Europe the Digital Markets Act imposes constraints on large platforms that would have been unthinkable a decade ago, requiring data sharing and interoperability while limiting their ability to favor their own services or otherwise leverage their ecosystems. In the United States policymakers have historically relied less on ex ante regulation to prevent specific actions or behaviors and more on ex post antitrust enforcement through the courts, but this too is evolving as antitrust scrutiny intensifies. Until recently, the last major U.S. antitrust case against a digital platform was the one against Microsoft in the late 1990s. It focused on how a dominant firm used control over a key distribution channel, the Windows operating system, to favor its own browser, Internet Explorer. That case helped preserve the openness of the web, creating the conditions for new entrants like Google to emerge and grow.
Two decades later regulators are once again focused on control over distribution, with particular scrutiny of defaults and other forms of preferential treatment. These issues are at the heart of recent cases against Google in search and advertising and Amazon in e-commerce. There is a certain irony here: The legal logic developed in the 1990s to constrain Microsoft and favor new entrants like Google is now being applied to the search giant itself.
The new volume and scope of enforcement mark a clear break from the largely hands-off approach that authorities have long taken. In the recent Google case, the U.S. Department of Justice argued successfully in district court that the tech company had maintained its dominance in search partly through exclusive default agreements. Yet the remedies have been modest, focused mainly on limiting some of these agreements and increasing competitors' access to data. (Both sides filed appeals in federal court in mid-2026.) But the broader significance of the case and others like it is that they signal a growing willingness by regulators to challenge platform strategies that reinforce and entrench dominant positions.
Privacy regulation adds another complication. Data, once treated as a proprietary growth engine, is increasingly subject to consent requirements, restrictions, and oversight. Laws such as the General Data Protection Regulation in Europe and the California Consumer Privacy Act require firms to obtain explicit permission for data collection and use. Regulators are also cracking down on "dark patterns," or design choices that nudge users to share more data than they would otherwise. The strategic tools that once fueled platform advantage are now often viewed as liabilities.
A second way the environment is changing is about profitability. For years technology companies operated under an implicit license to grow first and monetize later. Now platforms no longer have unlimited time to become profitable. Investors remain willing to tolerate delayed monetization for many AI-first ventures but not for traditional platform businesses, which face increasing pressure to deliver clearer and faster returns. Recykal, a platform for exchanging recyclable materials in India, shows how organizations are adapting to this shift. In 2024 the company chose to sacrifice some revenue in order to focus on activities that would improve its margins and overall profits.
These changes indicate that platforms' growth can no longer rely on regulatory laissez-faire, investors' patience, or the assumption that scale alone will drive continued success. Yet these same forces create an opportunity for new entrants and alternative models.
Four Strategies for Growth
If winner-take-all outcomes are rare and will increasingly be constrained by regulation when they do arise, that does not mean entrepreneurs should stop building platforms. Quite the opposite. Platforms have already created enormous value by reducing information asymmetries, expanding access to products and services, and enabling entirely new forms of economic activity. For instance, they have turned small, disconnected local markets into robust, highly active, often global ones. Craft markets, short-term rentals, and freelance labor are just a few examples of sectors that have been fundamentally reshaped.
Platform companies have also figured out how to build trust at scale. Through screening mechanisms, reputation systems, and online reviews, they have made interactions between strangers not only possible but routine. A generation ago, the idea of getting into an unknown driver's car or staying in a home that you booked online would have seemed implausible. Today it is commonplace.
Finally, platforms have introduced flexibility into markets that were historically rigid. Traditional industries, such as hotels, often struggle to adjust capacity as demand changes, and others, such as taxis, also face relatively inflexible pricing. Platforms like Uber and Airbnb enable both supply and prices to respond dynamically, reducing inefficiencies and creating new opportunities for less professional providers.
The key shift now is in understanding that platforms don't need to dominate entire markets in order to create value. In this environment, success depends on making deliberate choices along four dimensions: where to grow, how to create value, how to monetize, and how to sustain trust in the long run. Four strategies can help businesses navigate each of these dimensions.
Prioritize users that create the greatest spillovers. When platforms attract multiple user groups, growth depends on which users create the strongest spillovers — that is, whose participation generates the most benefits for other groups of users. For example, on a marketplace where adding high-quality sellers attracts many new buyers but adding more buyers has only a modest effect on seller participation, sellers are the stronger driver of growth. Investing heavily in the side of the market that's easier to attract often delivers disappointing returns if it does little to stimulate meaningful interactions. The first mistake many platforms make is failing to identify which group of users truly drives growth.
Taskrabbit, a platform for local services, illustrates the importance of figuring out these market dynamics. In a 2020 study of the platform, my colleague and I found that user participation was highly elastic on the supply side: Taskers (people available for hire on the platform) were willing to take on significantly more work as demand increased, without requiring higher prices. That meant demand — not supply — was the key constraint on growth. By focusing on growing demand, Taskrabbit unlocked more transactions, which translated into higher earnings and stronger retention among taskers. Growth thus came from activating the most valuable side of the market, not from expanding both sides equally.
This logic is familiar in platform pricing: Platform companies often charge very little or even subsidize the user group whose participation generates the greatest spillovers. But pricing incentives alone are not always enough to produce growth. What's often required instead or in addition is to build capabilities that enable higher-quality participation. Amazon, for example, could not rely on third-party sellers to provide fast, reliable shipping at scale. It had to invest in a logistics and fulfillment infrastructure to make that possible. Doing so increased the value of the platform for both buyers and sellers.
Recykal, which connects and serves several stakeholders in India's waste management and recycling ecosystem, faced a similar situation. Increasing subsidies for those supplying recyclable materials could have attracted greater quantities of waste, but most likely at the expense of quality. Lower-quality materials, in turn, could have undermined the platform's reputation among buyers — including authorized recyclers and processing plants, consumer brands, packaging producers, and large-scale manufacturers. Rather than using subsidies as a lever to attract more supply, Recykal has started investing in deposit-refund systems and supporting infrastructure to improve the quality and consistency of supply.
Both platforms illustrate the same principle: Growth comes not from attracting the largest number of users at the lowest cost but from investing in the users and capabilities that create the greatest amount of value for everyone on the platform.
At the same time, prioritizing one side does not mean neglecting the other. Persistent imbalances can erode trust and undermine the platform in the long term. Uber's early focus on riders contributed to tensions with drivers, while Amazon's prioritization of buyers has, at times, strained relationships with sellers. Platforms are ultimately trust-based systems. Growth that comes at the expense of one side of the market is difficult to sustain.
Make the market "thicker" for current users. When platforms struggle to grow, companies' instinct is often to chase more users. But growth can come from a different source: creating more value from existing participants. Market "thickness" — the likelihood that users find high-quality matches — is not fixed. It is shaped by how the platform is designed. There are two primary ways to increase thickness.
The first is expanding the scope of the platform by adding products or services that increase opportunities for interaction. Many platforms pursue this path, and it can work well. But expanding too quickly can backfire. In early-stage platforms, having more categories does not always translate into creating more matches. In fact, my research shows that platforms often achieve higher match rates and stronger user retention when activity is concentrated in a smaller number of categories, at least early on. Specialization can make the market feel thicker by increasing the likelihood that users quickly find what they need.
Uber's creation of services like Uber Eats shows the potential of scope expansion. Adding new use cases can increase engagement and utilization across the network. But this approach is especially effective when those services increase the participation of existing users and providers — for example, by giving riders another reason to use the app and drivers another way to earn income.
The second, and often more-powerful, method is improving the platform's design to create more matches among users and to increase the matches' value. Doing so involves making decisions about pricing mechanisms, how quality is signaled, and how search and matching algorithms are structured, because small changes in these parameters can have large effects. For example, e-commerce platforms recognized early that the advantage of fixed prices over auctions was the ability to reduce friction and speed up transactions, even if it meant sacrificing some price discovery. (Auctions help reveal what buyers are willing to pay.) Similarly, using verified user reviews over less informative credentials, such as someone's years of experience in a field, can improve trust and conversion without adding new users.
Uber's redesign of shared rides, where passengers traveling in similar directions share a vehicle, illustrates this second lever. By asking riders to walk short distances and wait slightly longer for pickups and shifting its algorithms from matching one rider with multiple potential drivers to simultaneously matching multiple riders with an available driver, Uber dramatically increased the pool of feasible matches. That change improved utilization, lowered costs, and made it harder for competitors to reach viable scale.
This strategy has two powerful effects. It increases value for users already on the platform, and it raises the minimum efficient scale required for competitors to enter the market or survive. In markets where winner-take-all outcomes are unrealistic, this kind of advantage matters more than raw size does.
Look beyond the obvious monetization options. The classic platform economics principle of charging the side of the market that is least sensitive to price still holds, but it is no longer sufficient. Today's platforms have access to monetization levers that go beyond simply charging users to participate or transact.
In the retail media model, for example, a platform monetizes its access to users by selling targeted advertising to third parties. Amazon pioneered this approach: Instead of charging consumers more, it charges sellers and brands to promote their products within the marketplace through sponsored search results and featured placements. In effect, sellers pay for visibility where consumer intent to buy is highest.
Other platforms are adopting these revenue models. What makes retail media powerful is that it builds on the data a platform already generates (search queries, browsing behavior, and transaction histories) to create highly targeted, high-conversion advertising opportunities. That allows platforms to monetize without raising prices on end users — and in some cases even allows lowering them — while giving suppliers a measurable return on their spending. Edenred, a payments and employee benefits platform, is exploring how transaction data and user engagement can support new revenue streams by connecting merchants with employees at relevant moments. Marriott likewise leverages customer data and its loyalty program to create targeted offers and partnerships that generate revenue beyond room bookings. The key is not simply charging more; it's charging differently by aligning monetization with where value is created.
Platforms that rely on a single revenue model often find themselves constrained. Those that expand their monetization tool kit thoughtfully and transparently can unlock new sources of growth without alienating users.
Treat customer welfare as a strategy, not as compliance. Many platforms have driven users away with short-term, extractive tactics that increase engagement or revenue in the moment but erode trust over time. This approach sits at the heart of many regulatory battles. Practices such as self-preferencing — when a platform uses its dominant position to favor its own products, services, or apps over those of competitors — may boost profits in the near term, but they often come at the cost of credibility and user satisfaction. What looks like optimization today can become a liability tomorrow, both reputationally and legally.
The platforms that endure will be those that make creating overall value and positive experiences for users a strategic principle. These businesses depend on repeat interactions, and users must feel that the platform is acting fairly. Once that trust erodes, it is difficult, and often costly, to rebuild.
I see these dynamics play out across many platforms. Airbnb has invested heavily in trust mechanisms, yet travelers still worry that the home they booked will turn a dream vacation into a nightmare, and hosts fear serious damage to their property. Uber, after years of contentious changes, faces persistent skepticism from drivers, even when introducing features intended to improve their experience. In both cases, past decisions shape how new initiatives are perceived.
Google's experience with Privacy Sandbox highlights a subtler version of this challenge. The initiative was intended to replace third-party cookies with a system that protects users' information while still allowing advertisers to target audiences. Rather than sharing detailed user-level data, Privacy Sandbox was designed to keep more information within the Chrome browser and provide aggregated or anonymized signals to advertisers. Conceptually, the approach aimed to balance privacy and monetization, and research that I conducted with others suggests it is one of the most effective ways to obtain user consent for data collection and use.
Yet despite its technical merits, the proposal faced significant resistance and was ultimately abandoned, in part because regulators and other industry participants worried that shifting data and targeting capabilities into infrastructure controlled by Google could further entrench its market power. As a result, many publishers and regulators have fallen back on existing solutions, such as ubiquitous cookie consent banners — which are widely disliked by consumers yet persist because they are seen as more transparent and less likely to concentrate control over data.
In each of these cases, perception matters as much as reality. Even well-intentioned changes can fail if users or regulators do not trust a platform's motives. Critically, customer welfare should be interpreted broadly in platform settings, referring not only to end consumers but to all sides of the market. For Amazon, that includes both buyers and sellers; for Uber, both passengers and drivers; for Google, both users and advertisers. Regulators emphasize customer welfare because they recognize that short-term profits alone may not provide firms with sufficient incentive to invest in innovation and create long-term value. Platforms should take the same approach, not out of altruism but because creating value for customers is the most durable path to creating value for the business.
. . .
In a world of persistent competition, active regulation, and skeptical capital, platforms' success depends on disciplined choices: where to grow, how to create value, how to monetize, and how to sustain trust. The platforms that thrive in this next era will not be the ones waiting for markets to tip. They will be the ones that understand why many markets rarely do and that build strategies to succeed in this new reality.
Copyright 2026 Harvard Business School Publishing Corporation. Distributed by The New York Times Syndicate.
