The next major technological battle will not only take place in data centers, chips, or AI models. If McKinsey’s roadmap is accurate, a significant portion of value will end up concentrated in a much more visible layer for companies and consumers: e-commerce. In its analysis of the 18 main economic arenas with potential through 2040, the firm ranks e-commerce first by estimated revenue, with a range between $14 trillion and $20 trillion. The second category, AI software and services, falls far behind, with estimated revenues between $1.5 trillion and $4.6 trillion.
The difference is substantial. In fact, it suggests something more profound than a simple sector forecast: that AI infrastructure might ultimately find its major commercial outlet in digital purchasing. In other words, semiconductors, cloud, AI software, and payment systems would be key pieces, but the point where a huge part of the business materializes is in the final transaction. For a tech-focused audience, this interpretation matters because it requires viewing e-commerce not just as a subset of traditional retail, but as a top-tier technological platform.
The table summarizing the 18 arenas with the most potential through 2040
The following table captures the key data from McKinsey’s chart on estimated revenues in 2040 and compound annual growth rate (CAGR) from 2022 to 2040:
| Potential Arena | 2022 Revenue | Estimated 2040 Revenue | CAGR 2022-2040 |
|---|---|---|---|
| E-commerce | $4B | $14–20B | 7% – 9% |
| AI Software & Services | $85B | $1.5–4.6T | 17% – 25% |
| Cloud Services | $220B | $1.6–3.4T | 12% – 17% |
| Electric Vehicles | $450B | $2.5–3.2T | 10% – 12% |
| Digital Advertising | $520B | $2.1–2.9T | 8% – 10% |
| Semiconductors | $630B | $1.7–2.4T | 6% – 8% |
| Shared Autonomous Vehicles | n.d. | $610–2.3T | n.d. |
| Space | $300B | $960–1.6T | 7% – 10% |
| Cybersecurity | $160B | $590–1.2T | 8% – 12% |
| Batteries | $98B | $810–1.1T | 12% – 14% |
| Modular Construction | $180B | $540–1.1T | 6% – 10% |
| Streaming Video | $160B | $510–1.0T | 6% – 11% |
| Video Games | $230B | $550–910B | 5% – 8% |
| Robotics | $21B | $190–910B | 13% – 23% |
| Industrial & Consumer Biotechnology | $140B | $340–900B | 5% – 11% |
| Future Air Mobility | n.d. | $75–340B | n.d. |
| Obesity & Related Conditions Drugs | $24B | $120–280B | 9% – 15% |
| Nuclear Fission Plants | $18B | $65–150B | 7% – 13% |
| Total | $7.25T | $29–48T | 8% – 11% |
The picture is striking for two reasons. First, e-commerce holds a commanding lead by a large margin. Second, this leadership coexists with higher growth rates in other areas, such as AI software or robotics. This suggests that the value map will not be linear: some industries will grow faster, but others will capture much larger business volumes based on their position within the digital ecosystem.
E-commerce is transforming from just retail into business infrastructure
One of McKinsey’s most interesting points is how it repositions e-commerce within the technological system. It’s not just seen as an evolution of retail but as a layer connected to other sectors accelerating the digital economy: cloud, semiconductors, software, cybersecurity, payments, and automation.
This idea aligns with current practice. Modern e-commerce depends on structured catalogs, recommendation engines, anti-fraud systems, logistics platforms, APIs, payment gateways, observability, real-time analytics, and increasingly, automation based on AI models. From this perspective, the online store ceases to be merely a sales webpage and becomes a complex technological assembly, where improvements in inference, latency, integration, or personalization can directly influence real revenue.
McKinsey also provides a relevant historical perspective. Analyzing major “arenas” emerging over the past two decades, it states that these industries experienced faster growth, attracted more R&D investment, generated greater economic benefit, and showed more competitive dynamism than others. Growth alone isn’t enough: a sector is also defined by how quickly leadership shifts, the influence of new entrants, and its ability to attract investment and demand.
The next disruption: when customers become AI agents
If the first major message of the report is that e-commerce will be one of the layers capturing the most value, the second is even more disruptive: an increasing portion of that business could be mediated by AI agents. In specific research on agent-based commerce, McKinsey estimates that between $3 trillion and $5 trillion of global commerce could be mediated by agents by 2030. In the US B2C retail market, the opportunity is between $900 billion and $1 trillion.
The core idea is simple but transformative. If an agent can search, compare, select, build a cart, and execute a purchase under rules set by the user, then the traditional buyer’s journey diminishes in centrality. The website, app, and search engine still exist but are no longer the sole entry points. Conversation, context, automation, and interoperability move to the forefront.
For tech companies, this opens a new layer of competition. It’s no longer just about better design or conversion rates for human users. It’s about being interpretable by automated systems, exposing inventory and prices in structured formats, enabling delegated authentication and payments, and creating experiences that work well for both humans and agents.
Protocols, APIs, and payments: the real battleground will be in the invisible layer
The core of agent-based commerce isn’t in a flashy interface but in the infrastructure. McKinsey highlights several key elements: Model Context Protocol, Agent-to-Agent Protocol, Agent Payments Protocol, contextual personalization, dynamic planning, and agents capable of interacting with interfaces even when no direct API exists.
All of this points to a different architecture than traditional e-commerce. Instead of pages optimized for clicks, the market is shifting toward services designed for other systems to understand, negotiate, and consume. The quality of the catalog will depend less on images or product descriptions and more on semantics, traceability, and ability to integrate into automatic decision flows.
Simultaneously, the payments area becomes critical. When purchasing is performed by an agent acting on behalf of a user, the rules around authorization, identity, fraud, auditing, and responsibility change completely. An essential issue for Europe’s tech sector is trust. McKinsey emphasizes that agent-based commerce won’t scale without new mechanisms for control, transparency, and reversibility. This is not just a technical challenge but also regulatory and cultural.
For platforms, marketplaces, software providers, payment gateways, and infrastructure companies, the takeaway is clear. The biggest opportunity isn’t just adding AI features to existing systems but redesigning parts of the tech stack for a scenario where discovery, decision-making, and purchasing become increasingly automated. In this context, e-commerce isn’t just the final screen of the tech process—it’s likely where all that technology translates into revenue.
Frequently Asked Questions
Why does McKinsey rank e-commerce ahead of AI software in its forecast?
Because McKinsey projects that e-commerce could move between $14 trillion and $20 trillion by 2040, compared to the $1.5–4.6 trillion forecasted for AI software and services. The implication is that the final commercial layer may capture more business volume than some enabling technologies.
What exactly is agentic commerce?
It’s a model where AI agents act on behalf of users to search for products, compare options, build carts, and complete purchases. It doesn’t necessarily replace the human buyer but automates increasingly larger parts of the process.
What technologies does a business need to prepare for this scenario?
Structured catalogs, robust APIs, identity and authorization systems, delegated payment solutions, personalization engines, decision traceability, and tools to distinguish legitimate agents from malicious automation.
What changes for an online store if the customer becomes an agent?
Product presentation methods, traditional SEO roles, visual interface importance, fraud management, loyalty logic, and the relationship between brand and consumer all shift. In such a landscape, experiences tailored for agents can become just as important as those for actual humans.