Skip to main content

Internet Platforms

· 4 min read
Steven Normore
Steven Normore
Author

In 2007, Marc Andreessen—a pioneer of the early internet and creator of Mosaic—shared his thoughts on internet platforms in an article titled Three Kinds of Platforms You Meet on the Internet. He categorized platforms into three distinct levels based on their functionality and the opportunities they offered developers. As we explore these categories, we’ll also examine how the landscape has evolved and consider how Andreessen might frame his ideas today in light of technological advancements and platform economics.

The Three Levels of Platforms​

Level 1: Access API​

The first level consists of platforms offering web services APIs (e.g., REST, SOAP). These APIs allow developers to retrieve or manipulate data from the platform to build external applications. For example, Flickr’s API enabled developers to display photos within their own apps.

  • Pros: Minimal investment for platform owners; developers gain access to valuable data streams.
  • Cons: External developers bear the burden of hosting and scaling their applications, often leading to increased costs and complexity.

Level 2: Plug-In API​

The second level involves platforms that allow developers to embed their apps directly into the platform’s user interface. This creates a seamless experience for users who interact with third-party apps as though they are native to the platform.

  • Example: Facebook’s early platform allowed developers to build social apps that lived within the Facebook ecosystem.
  • Pros: Direct user reach and a smoother user experience.
  • Cons: Increased platform management complexity, including performance and security concerns from third-party code.

Level 3: Runtime Environment​

At the highest level, platforms provide a complete environment for developers to build, host, and run their applications. These platforms handle infrastructure management, allowing developers to focus on application logic.

  • Example: Salesforce’s Force.com, which enabled businesses to create and deploy custom applications entirely within Salesforce’s infrastructure.
  • Pros: Reduced operational complexity for developers; platforms enforce security and performance standards.
  • Cons: Developers are constrained by the platform’s technical and economic policies.

How Platforms Have Changed Since 2007​

In the years since Andreessen’s article, the internet platform ecosystem has undergone significant transformation. Cloud computing, mobile platforms, and decentralized technologies have reshaped how platforms operate and what they offer.

1. Cloud Computing and Platform-as-a-Service (PaaS)​

Today, many platforms straddle Levels 2 and 3 by offering sophisticated developer environments that abstract nearly all infrastructure concerns. Amazon Web Services (AWS), Google Cloud Platform (GCP), and Microsoft Azure dominate as runtime environments, with developers relying on these platforms for hosting, scaling, and AI services. These PaaS offerings could be considered an extension of Level 3, demonstrating how platform capabilities have matured.

2. Mobile Ecosystems​

The rise of mobile ecosystems such as Apple’s App Store and Google Play has created hybrid platform models. Developers operate within tightly controlled environments (akin to Level 2) while also relying on the runtime capabilities of mobile operating systems (Level 3). Monetization and user access are tightly coupled with these platforms, which wield enormous gatekeeping power.

3. Decentralization and Web3​

The emergence of decentralized platforms like Ethereum represents a new paradigm. These platforms enable developers to build applications that are distributed and governed by code rather than centralized authorities. While these systems lack the integration seen in traditional Level 3 platforms, they offer unique benefits such as trustless computation and resistance to censorship.

How Andreessen Might Write It Differently Today​

Given today’s landscape, Andreessen might revise his framework to include a fourth level that reflects the influence of modern trends:

Level 4: Ecosystem Platforms​

Modern platforms are no longer isolated systems but vast ecosystems encompassing multiple layers of functionality. These platforms, such as AWS, Apple’s ecosystem, or Ethereum, integrate elements of all three original levels while adding new dimensions like marketplace economics, machine learning tools, and decentralized governance.

  • Features: Platforms provide not only APIs, runtime environments, and plug-in capabilities but also tools for monetization, AI/ML integration, and collaboration.
  • Challenges: Increased regulatory scrutiny and rising concerns over monopolistic practices. Developers face constraints that go beyond technical limitations, including high platform fees and restrictive policies.

Additionally, Andreessen might acknowledge the impact of open-source platforms, which challenge proprietary models by offering freely accessible tools that can rival commercial alternatives.

Conclusion​

Marc Andreessen’s original taxonomy of internet platforms remains a foundational lens through which we can understand developer ecosystems. However, the rapid evolution of cloud computing, mobile ecosystems, and decentralized technologies has expanded the definition of what a platform can be. Today’s platforms are more integrated, more powerful, and—for better or worse—more central to the functioning of the internet. Reflecting on this evolution underscores the need for developers, businesses, and regulators to navigate the complexities of platform dynamics with both caution and creativity.