Chapter 3: The Zero-Transaction-Cost Society　A2A Revolution: Why Platforms Will Collapse in the Age of AI

The history of economics is filled with great theories. Yet among them, one of the most insightful explanations for the existence of the modern corporation is undoubtedly Ronald Coase’s seminal work, The Nature of the Firm.

The Secret of Capitalism Discovered by Ronald Coase

In 1937, at the age of just 26, the young economist challenged an assumption that most economists of his time took for granted. The prevailing view was that markets function efficiently through the price mechanism. Since Adam Smith, economics had argued that the “invisible hand” allocates resources optimally. However, when Coase looked at the real world, he noticed something puzzling. Alongside markets existed enormous organizations that operated according to entirely different principles.

Those organizations were corporations.

In markets, people coordinate their activities through prices. Inside corporations, however, people coordinate their activities through managerial authority and organizational hierarchy. Rather than market principles, organizational principles govern behavior.

This led Coase to ask a fundamental question:

“If markets are so efficient, why do firms exist at all? Why shouldn’t everyone operate as an independent contractor, entering into agreements only when necessary? Why do we have giant corporations employing tens of thousands of people?”

The answer Coase arrived at was remarkably simple:

Because using the market is costly. Firms are created to reduce those costs, which he called Transaction Costs.

Specifically, market transactions involve the cost of searching for trading partners, negotiating prices and terms, and monitoring contractual performance. These invisible frictions are present in every market exchange. Coase referred to their total as transaction costs.

In other words, firms exist because they reduce transaction costs.

The Reason Firms Exist: Transaction Costs

Consider the case of hiring a designer. First, you must find potential candidates. Then you need to evaluate which designer is most capable, compare prices, assess reliability, and determine whether deadlines can be met. Even after identifying a suitable candidate, negotiations must take place. Once the project begins, you must ensure that the promised quality is actually delivered.

All of these activities constitute transaction costs.

Broadly speaking, transaction costs can be divided into three categories:

1. Search Costs – the effort required to find the optimal trading partner.
2. Negotiation Costs – the time and resources required to negotiate terms and finalize agreements.
3. Enforcement and Monitoring Costs – the effort required to ensure compliance and resolve disputes when contracts are breached.

Viewed from this perspective, it often becomes more efficient to hire people as employees rather than repeatedly contracting with external parties. If salespeople, accountants, and lawyers are employed within the organization, there is no need to search for and negotiate with them every time a task arises.

This is how firms emerged and evolved.

The Twentieth Century Was a Battle Against Transaction Costs

Later, Nobel Prize-winning economist Oliver Williamson expanded upon Coase’s theory.

In works such as Markets and Hierarchies and The Economic Institutions of Capitalism, Williamson analyzed how the boundary between markets and organizations is determined.

According to Williamson, human beings possess three characteristics. First, they exhibit bounded rationality; people cannot always make perfectly rational decisions. Second, they display opportunism; under certain circumstances, individuals may break promises in pursuit of self-interest. Third, there is asset specificity; investments made for a particular transaction may have little value outside that context.

When these three factors converge, market transactions become increasingly inefficient. As a result, firms internalize activities and integrate them into organizational structures.

The great corporations of the twentieth century emerged precisely because of this logic. Ford integrated operations from iron ore extraction to automobile production. Toyota built extensive keiretsu networks. GE and IBM internalized vast functions within large organizational structures.

The principle was straightforward: in many cases, it was simply cheaper to perform activities inside the firm than through the market.

It is therefore no exaggeration to say that the history of twentieth-century management was a history of fighting transaction costs.

The True Nature of Platform Capitalism

The twenty-first century produced a new generation of winners. In my 2010 book Platform Strategy (Toyo Keizai Inc.), which was later featured on TV Tokyo’s World Business Satellite, I discussed what I termed Platform Strategy® and the companies embodying it.

These companies included Google, Amazon, Apple, Meta, and Microsoft, as well as platform giants such as Uber, Airbnb, Booking.com, and Alibaba.

What exactly were these companies doing?

On the surface, they appeared to provide search engines, e-commerce platforms, social networks, and ride-hailing services.

From an economic perspective, however, their essence was remarkably simple:

They reduced transaction costs.

Google reduced information search costs. Amazon reduced product comparison costs. Uber reduced transportation search costs. Airbnb reduced both accommodation search costs and trust-verification costs.

Platform companies are, fundamentally, transaction-cost-reduction machines.

In return, they have been able to capture substantial take rates. Apple collects up to 30 percent of app revenues. Uber often captures more than 20 percent of transaction value. Booking.com charges hotels commissions ranging from 15 to 25 percent.

These profits were not merely examples of monopoly pricing. They were accepted because the platforms removed market frictions.

AI Agents Will Eliminate Transaction Costs

Today, however, this foundation is beginning to crumble.

The reason is the rapid advancement of generative AI. Since the emergence of ChatGPT, many people have viewed AI primarily as an intelligent chatbot. Yet that is not where the true disruption lies.

The real story is the rise of AI agents.

As discussed earlier, AI agents are digital entities capable of making decisions and acting autonomously on behalf of humans. Although still in their early stages, AI agents are likely to become the primary economic actors of the 2030s.

What happens then?

Transaction costs approach zero.

Traditionally, humans required time to compare alternatives. AI can gather and analyze global information instantly. Contract negotiations that once took weeks can be completed in seconds between AI systems. Monitoring contractual performance, previously dependent on human oversight, can increasingly be automated through smart contracts and cryptographic technologies.

The three transaction costs identified by Coase—search costs, negotiation costs, and monitoring costs—move toward near-zero levels.

A2A and the Emergence of a New Economic Order

This is where A2A, or Agent-to-Agent interaction, becomes critically important.

The internet we know today was designed primarily as a network connecting humans. Social media, e-commerce platforms, and search engines were all built around human users.

That is about to change.

A world is emerging in which AI agents connect directly with one another.

A consumer’s AI searches for products. A company’s AI proposes prices. A logistics AI arranges delivery. A payment AI executes contracts.

The entire process occurs without direct human involvement.

In such a world, traditional advertising and sales activities lose much of their significance. AI agents are not persuaded by advertising. They are not swayed by brand imagery. They evaluate price, quality, delivery speed, reliability, and total value with cold precision. Superficial incentives and marketing gimmicks become far less effective.

As a result, markets become more efficient than ever before—and at the same time, more unforgiving than ever before.

The Collapse of Network Effects

One of the greatest sources of competitive advantage for platform companies has been network effects.

The more users a platform attracts, the more valuable it becomes. The more valuable it becomes, the more users it attracts. This self-reinforcing cycle enabled the rise of GAFAM.

In an A2A world, however, the dynamics change.

AI agents do not necessarily depend on massive centralized platforms. Instead, they can connect directly through open protocols. Just as email was never controlled by a single corporation, future AI communications are likely to operate through open standards.

As a result, the source of network effects shifts from platforms to protocols.

This is the essence of what I describe as the “collapse of platforms.”

A Question for Business Leaders

Business leaders must ask themselves a difficult question.

Are our profits truly derived from value creation? Or are they derived from market friction?

Are customers choosing us because we create superior value—or simply because comparison is inconvenient? Are they staying because switching is difficult? Are we protected by information asymmetry?

If the answer is the latter, that advantage may largely disappear in the age of AI.

The future will not belong to companies that rely on lock-in. It will belong to companies that possess irreplaceable assets.

Intellectual property. Brands. Communities. Physical supply chains. Proprietary data. AI-accessible APIs. MCP-compatible infrastructures.

In the AI era, competitive advantage will come not from controlling friction, but from creating value.

Ronald Coase explained why firms exist in 1937. Nearly ninety years later, we face the opposite question:

If transaction costs approach zero, how much of the firm still needs to exist?

That question may well become the defining challenge of management theory in the age of AI.

In the next chapter, we will explore how AI may dismantle large bureaucratic organizations and give rise to what I call the Hyper-Smart Firm—a fluid, highly adaptive organization built around AI, networks, and human creativity.

Platform Strategy® is a registered trademark of NetStrategy Co., Ltd.