Platform: from O'Reilly's Web 2.0 (2005) to Tirole's multi-sided market theory, and why understanding the word changes decisions

Platform is probably the most overloaded word in the contemporary marketing and technology lexicon. It's used for almost any digital product with several users, which ends up dissolving the concept. When everything is a platform, the word stops communicating.

There is a precise technical and economic sense of the term worth recovering, because it has practical consequences for business decisions: where to compete, how to build, what risks you take on by depending on other people's platforms, what dynamics affect your business when a platform decides to change the rules. This article tries to restore some of that precision and apply it to the context of communication and marketing in 2026.

The technological origin: Tim O'Reilly and "Web 2.0," September 2005

The modern use of the term "platform" on the internet was significantly popularized by an essay by Tim O'Reilly published on September 30, 2005 titled "What Is Web 2.0: Design Patterns and Business Models for the Next Generation of Software." O'Reilly, founder of O'Reilly Media (the technical publisher), had been thinking with his team about what distinguished the internet companies that survived the 2001 crash from those that failed.

The conclusion O'Reilly articulated: the successful companies were treating the web as a platform —not as a set of pages, but as infrastructure on which participatory services were built, where users were producers of value, not just consumers. His examples in 2005 were Google, Wikipedia, eBay, Amazon, BitTorrent, Napster (in its day), and the emerging players of the moment.

Some principles O'Reilly identified:

• The web as platform: building services on shared web infrastructure instead of packaged proprietary software.
• Harnessing collective intelligence: value grows with user participation.
• Data is the next Intel inside: accumulated data becomes a competitive asset.
• End of the software release cycle: continuous updating instead of discrete versions.
• Lightweight programming models: APIs and mashups instead of heavy systems.
• Software above the level of a single device: an experience that spans devices.

O'Reilly's essay is one of those texts where something already happening gets named and made visible. After its publication, the language of "platforms" as a business category quickly entered the technical, financial, and media vocabulary.

"Software Is Eating the World": Marc Andreessen, August 2011

Six years later, another influential essay extended the conversation. On August 20, 2011, Marc Andreessen —co-founder of Netscape, now co-founder of the venture capital firm Andreessen Horowitz— published in the Wall Street Journal the famous essay titled "Why Software Is Eating the World."

The central thesis: large traditional companies in any industry were being displaced by software companies, not because software was the industry but because software was increasingly how all industries were run. Amazon was eating retail. Netflix was eating TV. Spotify was eating music. Salesforce was eating CRM. The companies dominating these shifts were, in many cases, platforms in the technical-economic sense of the term.

Andreessen implicitly extended O'Reilly's observation: platforms weren't just a feature of the web; they were the new dominant form of competition across entire industries. The phrase "software is eating the world" became a recurring quote in tech investing in the 2010s and is still cited in 2026.

The economic theory: Jean Tirole and multi-sided markets

While the technologists were popularizing the concept, economists were formalizing the underlying theory. Jean Tirole, a French economist and professor at the University of Toulouse, received the Nobel Prize in Economics in 2014 partly for his work (together with Jean-Charles Rochet) on multi-sided markets or two-sided markets.

The theory: a platform is typically a market with two or more sides where the platform facilitates interactions between distinct groups of users whose decisions affect the value of the other side. Classic examples:

• Visa/Mastercard: merchants and consumers. Neither side wants to use the system if the other isn't there. The platform solves the coordination problem.
• Newspapers: readers and advertisers. More readers attract advertisers; advertising revenue funds content that attracts readers.
• Operating systems: users and app developers. More users attract developers; more apps attract users.
• Marketplaces (eBay, Amazon, Airbnb): sellers and buyers. More supply attracts demand; more demand attracts supply.
• Content platforms (YouTube): creators and audience. More creators attract viewers; more viewers attract creators.

The theoretical contribution of Tirole and collaborators was to show that multi-sided markets have economic properties that traditional markets don't: asymmetric pricing structure (sometimes one side pays and the other earns), cross network effects, particular competitive dynamics, distinct regulation. Those findings changed how platform companies are thought about in strategy, antitrust, and public policy.

The reference book: Platform Revolution (2016)

The most influential synthesis for management professionals was published in 2016: Platform Revolution: How Networked Markets Are Transforming the Economy and How to Make Them Work for You, written by Geoffrey G. Parker, Marshall W. Van Alstyne, and Sangeet Paul Choudary (W. W. Norton & Company). The book is based on academic research by Parker and Van Alstyne on platforms over prior years, plus Choudary's consulting with companies implementing platform strategies.

The central operating idea: platforms operate by different rules than traditional companies based on value chains (pipelines). Where pipelines optimize internal product creation and linear delivery to the customer, platforms optimize interactions between external producers and consumers, with the platform itself acting as the intermediary that captures value.

Some concepts from the book that have become standard vocabulary:

Inversion. Platforms differ from pipelines in that they invert the logic of control: instead of owning productive resources, they facilitate them; instead of optimizing internal processes, they optimize external interactions; instead of increasing value through control, they increase it through coordination.

Network effects. The value to each user grows with the number of other users. Types: direct (same side of the market, like WhatsApp, where your value grows with how many friends are on it), cross (different sides, like Uber, where more drivers increase value for passengers), negative (saturation that reduces value, like spam on a large social network).

Multi-homing. When users participate on several platforms at once. If switching costs are low, multi-homing is high; if specific investment is high, multi-homing is low. It determines whether a platform can sustain a monopoly or whether competition between platforms holds.

Envelopment. When a platform absorbs functions that other platforms used to handle. Microsoft Windows enveloped Netscape; Facebook has repeatedly enveloped competitors' functions; Apple has enveloped multiple categories through iOS.

The economic reality of dominant platforms

In 2026, the world economy is dominated by a small set of platforms with powerful network effects. The "Big Tech" companies (Apple, Microsoft, Alphabet/Google, Amazon, Meta, NVIDIA in its new profile) represent colossal market capitalizations sustained largely by that platform logic.

Some relevant empirical observations:

Platforms tend toward a monopolistic or few-player structure. Network effects reward the first to reach critical mass; competition concentrates quickly. In many verticals there's a dominant platform (Amazon in U.S. e-commerce, YouTube in video, Google in search) or a duopoly (iOS/Android in mobile).

The platform's power over participants is asymmetric. Sellers on Amazon can't negotiate terms; they have to accept them. Creators on YouTube don't control the algorithm. Apps in the App Store depend on Apple's policies. That asymmetry is a persistent source of tension and regulation.

Growing regulation. The EU passed the Digital Markets Act (DMA) in 2022, in force since 2024, which imposes specific obligations on "gatekeepers" (platforms that meet size and market-power criteria). Initial designations included Apple, Google, Meta, Amazon, Microsoft, and ByteDance. The U.S. and other jurisdictions have also tightened scrutiny.

AI platforms are emerging as a new layer. OpenAI, Anthropic, Google DeepMind, xAI, and others operate as platforms that serve AI inference to a growing layer of applications built on top. It's likely that the next decade will be marked by similar platform dynamics applied to the AI layer.

Types of platform in marketing and communication

For marketing and communication professionals, the relevant platforms group into categories:

Content and attention platforms. YouTube, TikTok, Instagram, X, LinkedIn, Facebook, Pinterest, Snapchat. Where the audience consumes content and where brands compete for attention.

Search engines as platforms. Google, Bing, DuckDuckGo. They work as a two-sided market between users and websites/advertisers. Google Search plus Google Ads plus YouTube + Maps + others probably constitute the most complete marketing platform in the world.

Marketplaces. Amazon, Mercado Libre, eBay, Etsy, Shopify (which is a platform of platforms), Booking, Airbnb. Where buyers and sellers transact and the platform takes a commission or margin.

App stores. Apple App Store, Google Play. The distribution layer for mobile software. The disputes between Apple and Spotify/Epic/etc. over commissions reflect that dynamic.

Productivity and work platforms. Slack, Microsoft Teams, Zoom, Notion, Figma, Salesforce, HubSpot. Where professionals operate and where companies invest significant capital.

Creator platforms. Substack, Patreon, OnlyFans, Twitch. Where creators monetize their audience directly with the platform's support.

Social commerce platforms. TikTok Shop, Instagram Shopping. The fusion of content and commerce that has emerged in recent years.

Each category has different operating dynamics, and the decision of which to invest in depends on what you're after and what you can sustain.

The fundamental risk for brands: platform dependence

The central reality brands must keep in mind: building an audience or a business on a platform means depending on the platform's decisions. That dependence is a lever of asymmetric power.

Documented cases that illustrate the risk:

Algorithm changes that destroy reach. Facebook dramatically reduced the organic reach of Pages in 2014-2018, forcing brands to pay to distribute content to their own audience. Instagram has had algorithmic changes that collapsed creators' reach.

Policy changes that shut down products. Quibi (a short-form video streaming service) shut down in 2020 after six months of operation. Platforms change rules and entire categories of players are left without a business.

Decisions that penalize competitors. The accusations against Amazon for prioritizing its own products over those of marketplace sellers, against Google for prioritizing its own results over organic ones, against Apple for asymmetric App Store rules, are recurring accusations with antitrust cases in progress.

Fee cuts or hikes. Apple and Google charging a 30% commission in their app stores has been a source of litigation (Epic v. Apple notable). Fee changes affect the unit economics of whole businesses.

Account suspension/removal. With no real recourse, a platform can cut off access. It's happened to journalists, brands, individuals. The power asymmetry is plain.

The operational conclusion is that brands should treat platforms as valuable but risky channels, not as stable infrastructure. Maintaining owned channels (website, email, CRM database) that survive any platform change is an elementary discipline of risk management.

How to choose platforms as a brand

A decision that should be made consciously:

Where your audience is active today. Not theoretically —where they actually spend time and receive content. Honest research avoids assuming your audience "is on LinkedIn" when they're really on TikTok.

What type of content you can sustain well. If you can't produce video at a decent frequency, TikTok isn't for you. If you don't have a voice for LinkedIn, it's not the platform for your brand.

What level of dependence you accept. If your business depends 80% on Instagram, you're fragile. Diversification reduces risk even though it requires more resources.

What you want to build: presence or conversion. Some platforms are better for presence/brand (TikTok, YouTube), others for direct conversion (Google Search, paid social).

How much control you retain. A newsletter on your own domain = high control. An audience on TikTok = low control. Hybridizing is healthy.

What you're willing to pay. If your presence plan depends on organic reach, you're a hostage to the algorithm. If you budget for paid from the start, you're more protected but you spend more.

Common mistakes in the relationship with platforms

Treating platforms as neutral media. They aren't. Each platform has economic incentives, algorithmic biases, and policy decisions that affect your content. Understanding the medium's dynamics is necessary to use it well.

Producing identical content for all of them. Each platform has a native culture. The same video on TikTok, Instagram Reels, and YouTube Shorts can work, but well-done adaptation improves performance.

Not investing in owned channels. The brand whose only audience is on other people's platforms is vulnerable. Email, website, and your own community are a long-term investment worth making.

Assuming today's dominant platform will be tomorrow's. MySpace, Vine, Yahoo Search, Clubhouse —multiple platforms that dominated in their day and have disappeared or been marginalized. Platform dominance isn't eternal.

Investing everything in one platform without diversifying. If something happens (algorithm change, account suspension, market shift), the damage is total. Diversifying has a cost but protects you.

Underestimating the cost of keeping up with the platform. Each platform demands specific production, knowledge of the medium, and continuous management. Multiplying platforms multiplies cost, not just income.

Confusing vanity metrics with results. Followers on Instagram, views on TikTok, subscribers on YouTube are platform metrics, not necessarily business metrics. What matters is that they translate into something commercially actionable.

Not reading the terms of service. Platforms can change the rules. Brands that are surprised when something changes often didn't read what they signed.

Trusting platform analytics as the only truth. Each platform reports in its own favor. Cross-checking with independent analytics (GA4, your own data) reveals the reality.

Platforms and creative operations

For a brand operating across multiple platforms, coordinating production across channels with different dynamics is a continuous operational problem. Producing content native to each platform, maintaining brand consistency, measuring performance by channel, reusing assets efficiently —all of this requires infrastructure, not just creativity.

That coordination is a discipline of creative operations: the communication channels explain the PESO model that helps classify platforms, the editorial calendar coordinates multi-channel production, brand management keeps the voice consistent across different platforms, and creative KPIs measure whether the multi-platform presence generates real results.

At Polimake that logic lives across three surfaces: Studio to coordinate multi-platform campaigns, Studio to produce adapted variants, Media as the repository where assets by platform are tagged and accessible for intelligent reuse.

---

If you lead marketing, product, or strategy and you got here looking for an answer about the concept of a platform, the most useful thing you can take from this article is probably the combination of three ideas: platform has a precise technical and economic meaning (multi-sided markets with network effects, not just any digital product), building a business on other people's platforms means taking on the risk of asymmetric dependence (platforms can change the rules, and participants absorb it), and diversification across platforms + owned channels is an elementary discipline of risk management, not optional for serious brands.

To round it out, the communication channels covers the PESO model that organizes platforms into categories, bots and exploits on social media covers the specific risks of manipulation on platforms, and competitive analysis covers how to evaluate the platform ecosystem in your sector.

Quick references

• Communication channels — the PESO model for classifying platforms.
• Bots and exploits on social media — the risks on dominant platforms.
• Competitive analysis — evaluating the platform ecosystem.
• Conversion funnel — how the audience moves between platforms.
• Editorial calendar — multi-platform coordination.