Free or Paid Subscription? For Online Platforms, It’s a Dance.

This phrase gained popularity as a description of the trade-off at the heart of many modern platforms—you use their service without paying, and they profit by selling your data.

But there are other business models available to social-media companies, streaming services, and apps. Some platforms have framed subscription paywalls as consumer-friendly privacy protection, promising that the fees will keep them from harvesting and selling user data. Others have pursued a hybrid approach that offers a choice: use it for free and share your data or pay for privacy.

The model that platforms choose has big implications for their bottom line, market share, and regulation. After Meta moved to a two-tiered model in Europe, the EU responded with a 200 million Euro fine, declaring the “consent or pay” system in violation of its Digital Markets Act.

In a landscape where both established industry leaders and young upstarts must balance revenue and privacy demands, how do these platforms decide which business model is right for them?

New research by Kellogg’s Sarit Markovich and Tel Aviv University’s Yaron Yehezkel sheds light on this decision by modeling how online platforms respond to factors such as market competition, data value, and network effects. They found that the features a platform offers—and how much it depends on maintaining a large userbase—has a surprisingly strong influence on whether a free, paid, or hybrid model makes the most sense.

But in the fast-moving technology world, it’s also important to be nimble when the landscape shifts.

“Online companies need to understand that this is not a binary decision—it’s more dynamic than that,” Markovich says. “If they start out with one model, it doesn’t necessarily mean they should stay there. They should always be thinking about the commercial value of their data and the strength of their network effects. And if these change, their strategy should change as well.”

More than competition

To tease out how companies choose between these approaches, Markovich and Yehezkel built a mathematical model that studies the interaction of three factors: the intensity of competition the company faces, the commercial value of the data it can collect on its users, and the network effects that drive the platform’s features.

In traditional economics, competition is usually the most important driver of decisions about price. If there are more competitors in a market, that will drive prices lower as companies chase market share. Conversely, monopolies have more leeway on how much they can charge since customers have little choice.

So why do today’s dominant online platforms such as Google and Meta offer their core services for free? The difference is data—most of these companies’ revenues aren’t driven by billing customers but by selling users’ attention to advertisers and their personal data to third parties. That changes the calculus for companies when they decide to offer customers a free, paid, or hybrid plan.

Facebook, for example, collects incredibly granular data about its users: their social connections, likes and dislikes, locations, and much more. This data enables advertisers to run hyper-targeted ads on the site—and advertisers pay a premium to do so. Netflix and Spotify, on the other hand, collect lower value information on its users, mostly limited to their streaming choices. In these cases, a subscription model may make more sense.

But companies may also be incentivized to attract more users for a technical reason: network effects. Many of the attractive features offered by online platforms are powered by the size of the user network and thus are enhanced when more people use the app.

For example, people want to join social-media sites where their friends are also users, and the more people they follow, the higher is the value they receive from the service. Streaming services like Netflix and Spotify use network-based algorithms to recommend music and movies you might enjoy, and navigation apps like Waze gather data from other people’s phones to give you the best directions.

To build the critical mass of users needed to power these features, a platform may offer their product for cheap or free, even if they don’t directly profit from selling the data.

It’s all about the network

In their analysis, the researchers found that the optimal business model depends most heavily on the strength of these network effects.

When a platform relies on strong network effects for its features and has the ability to collect granular, commercially valuable data on its users, it is most profitable for that company to choose the free, data-based model. However, when network effects or data value are low, it makes more sense for a company to choose the subscription model.

But in the middle, when network effects and data value are intermediate, competition becomes more important. In the fight for customers, platforms benefit from offering both free and paid tiers—whether they are a market leader or a company trying to challenge for market share.

Markovich notes that this is exactly the shift Netflix made in 2022. After years of being a subscription-only service, it launched a “basic with ads” option in response to intensifying competition in the streaming space, she explains, in a bid to appeal to a wider range of customers.

But Markovich says she was surprised to find that the hybrid model was not always the most profitable option.

“You might just think, hey, it’s flexible and you’re giving consumers what they want,” she says. But it can also make these platforms vulnerable for entry.

“It intensifies competition,” Markovich says. “If two competing platforms choose the hybrid model, they both compete over all users. By having one offer a subscription model and one offer a free model, they differentiate themselves. Users then choose based on their privacy preferences, and that allows for a softer competition across the platforms.”

Platforms often fail to appreciate how to adjust their business model as these factors change, Markovich says. For example, when a company first launches, it has few users and weak network effects, making a subscription model more optimal. But as it attracts more users, it may generate more revenue from its user data, making a shift to a free model more attractive.

Companies can even invest in improving those factors by gathering more-granular data or by shifting their business model to differentiate themselves in an increasingly crowded arena.

“This is where I think firms might be missing the point in terms of realizing there is a way for them to affect where they want to be, in terms of the strength of network effects or the commercial value of their data,” Markovich says.

Price versus privacy

The researchers’ model also offers guidance to policymakers who want to balance protecting consumer privacy and restricting prices.

Corporate regulation is often focused on keeping prices low through breaking up monopolies and encouraging competition in a market. For online platforms, Markovich and Yehezkel found that price competition was fiercest when multiple hybrid models coexist in a particular space.

“If we do not allow platforms to choose the hybrid model, then they’re going to split the market,” Markovich says. “They will say, ‘Okay, I’m focusing on the consumers who care about privacy; you’re focusing on the ones who don’t’—and now each group of consumers has just one option. In that way, a platform can become a monopolist over a subset of the market. That’s why it’s better to give consumers a choice via the hybrid model.”

Furthermore, if the hybrid model is banned or discouraged—as it was for Meta in the EU—it will push companies that rely on strong network effects to the free model that collects and sells more high-value personal data. Allowing the hybrid model and encouraging market competition will likely drive down the cost of the subscription tier, enabling more consumers to select the privacy-protecting option.

“Regulators should recognize that there isn’t a one-size-fits-all solution,” Markovich says. “Protecting privacy does not always require ‘big gun’ tools, like banning entire business models. In many cases, simpler measures like price caps can better protect privacy. By limiting subscription prices, you can make choosing privacy more affordable and accessible for users.”