In today’s digital economy, mobile apps are everywhere—and so are the entrepreneurs trying to build them. Over 3.8 million apps are available in Apple’s App Store, with new ones entering every day. But what determines whether an app breaks through to success—or quietly fades into the background?

In our new research, we analyze over 400,000 mobile apps to understand what drives success. We focus on two factors: (1) how similar an app is to its rivals, and (2) what happens when the platform owner—Apple or Google—launches its own competing app. These issues sit at the heart of today’s policy debates about self-preferencing, platform power, and competition in digital markets.

An important reality for those competing in the app economy is that developers live in ecosystems created by others. To publish an iPhone app, for example, a developer must use Apple’s tools, follow its design rules, and submit to reviews and pricing systems. Google’s Play Store has similar controls. But these same platforms also sell their own apps—raising the question of whether a developer will find itself at the mercy of Apple of Google, be helped by them, or engage in fair competition.

Our study groups apps into “clusters”—collections of apps that serve similar functions and compete for users. For each app, we measure how similar it is to others in its cluster, particularly the top-performing apps. We then examine how these choices of product space influence success, as measured by download volumes. Finally, we look at what happens when Apple or Google compete in a cluster.

Our findings reveal three core patterns.

First, success depends on being close—but not too close—to your competition. Apps that resemble the average competitor in their cluster tend to do better. But apps that are too similar to the top-performing app in their group do worse. This suggests a strategy of moderate differentiation: developers benefit from being in a space where other apps are doing well, but they must offer something distinct from the dominant player.

This echoes patterns seen in other industries. In groceries, for example, national brands compete with store brands by distinct branding. In telecommunications, companies that tried to resell incumbent network services in the 1990s without real differentiation generally failed.

Second, frequent updates matter. Developers that continually improve their apps tend to be more successful, reinforcing the idea that customers value innovation and responsiveness.

Third, the platform’s own behavior plays a major role. We find a striking asymmetry between Apple and Google. When Apple launches a new app, it tends to boost the fortunes of similar third-party apps. When Google launches a similar app, third-party downloads decline.

Why the difference? Part of the answer lies in the business model. Apple places apps near competing third-party apps, incentivizing Apple to grow the whole category, making the iPhone even more attractive. In contrast, Google’s apps tend to be more distinct and novel. This creates broad appeal but might divert user attention.

Our findings complicate the conventional wisdom around self-preferencing. Much of the regulatory debate focuses on whether Apple or Google should be allowed to manage their own marketplaces. Critics argue that they self-preference, distorting competition. But the reality appears more nuanced.

When Apple enters a market, it acts more like a promoter, improving the overall ecosystem. Google, by contrast, acts more like a traditional rival, creating differentiated products.

For regulators and lawmakers, the message is clear: platform competition is not a binary issue. Rules controlling app platforms may eliminate certain innovations or promotions. Apple’s approach benefits some types of consumers, and Google’s approach benefits others.

The app economy is both an opportunity and a challenge. Entry barriers are lower than ever, but so is the margin for error. Our research suggests that success in digital markets depends on the interaction of product differentiation, innovation strategy, and platform dynamics. Simply put, who you compete with—and how you compete—matters.