The Arrival of the AI Era: Sequoia's 'Why, What, Now' Framework and Market Strategy

In 2025, at the AI conference 'AI Ascent' hosted by Sequoia Capital, partners Packer Radío, Sonya, and Konstantin delivered a keynote address on the latest AI market trends and investment strategies. This article explains the background of AI market growth, business opportunities, and specific strategies for companies and startups to win, following the 'Why, What, Now' framework they presented.

1. The Current State and Significance of the AI Market—Why AI Now?

1-1. The Arrival of a 'Great Wave' with Prerequisites Met

In the layer cake slide presented by Sequoia, they listed the following four prerequisites for why AI has shifted from an 'inevitable' phase to an 'imminent' one.

• Compute: Improvements in the performance of GPUs and specialized chips

• Networks: Widespread adoption of high-speed, high-capacity communication infrastructure

• Data: Accumulation of massive amounts of data from all fields

• Distribution: Establishment of information dissemination channels via social media and apps

They pointed out that these elements work synergistically, accelerating the adoption of AI compared to the past. In particular, the fact that attention was instantly focused worldwide following the emergence of ChatGPT in November 2022 is a major difference from the 'cloud' and 'mobile' eras.

1-2. Market Size Comparable to Cloud/Software

As of 2024, the cloud market has reached approximately $400 billion, surpassing the traditional software market. Sequoia predicts that the initial scale of the AI service and software market will be at least 10 times that of the cloud, with even greater growth potential in 10 to 20 years. It has been shown that profit pools are already shifting not only in the service sector but also in traditional software fields due to AI integration.

2. Trends in AI Business—What Should Be Targeted?

2-1. Focus on the Application Layer

What Sequoia prioritizes in its investments is not unicorns, but 'revenue and free cash flow.' Even in past cloud and mobile trends, the greatest value creation occurred at the application layer. Similarly, for AI, they assert that the winning strategy lies in industry- and function-specific applications, rather than just foundation models.

2-2. Startup Strategy Specialized in Verticals/Functions

Rather than a vertically integrated model, it is important to design apps by 'working backward from the customer' to address their pain points and incorporate a human-AI hybrid loop. As specific examples, 'Open Evidence,' which specializes in medicine, and 'Harvey,' which supports lawyers in the legal industry, were introduced, and it was argued that startups should leverage their deep industry expertise.

3. Specific Strategies for AI Companies to Win—How to Play?

3-1. Value Creation Thinking with the 'Leone Merchandising Cycle'

Based on the value chain (idea → product → market entry → customer support) advocated by Sequoia partner Doug Leone, AI startups will differentiate themselves from the following perspectives.

• End-to-end proposals: Not just providing tools, but accompanying the customer until the problem is solved

• Data Flywheel: Circulate product usage data to continuously enhance value

• Cross-Industry Expertise: Communicate in the language customers use

3-2. AI-Specific Requirements—The 5% Differentiation Point

According to Mr. Packer, 95% of AI company valuation is based on traditional business success factors (problem-solving ability, talent acquisition, etc.), while the remaining 5% consists of the following AI-specific metrics.

1. Revenue Vibe: Is it not just temporary revenue, but revenue accompanied by sustainable behavioral changes in customers?

2. Margin Trends: Token costs are falling rapidly. Focus on future improvement potential rather than current gross margins.

3. Data Flywheel Contribution: Clarify which business metrics it contributes to.

4. Future Outlook—The Agent Economy and Its Impact on Individuals and Organizations

4-1. From Agent Swarms to the Agent Economy

From 2025 onwards, we foresee the arrival of an 'Agent Economy' where multiple AI agents collaborate, trade, and build trust. It is suggested that a new economic sphere will be constructed where humans and AI work together to proactively conduct transactions and resource transfers.

4-2. Transformation of Individual Thinking and Management

In the AI era, a shift from deterministic to stochastic thinking is inevitable. Management skills for collaborating with agents will be required, and the ability to handle uncertainty—which traditional management methods cannot fully address—will hold the key.

Sequoia AI Ascent 2025 demonstrated that AI is at an 'imminent' technological inflection point, that business value is concentrating in the application layer, and that the Agent Economy is arriving. Now is the time for companies to focus on developing AI solutions 'back-calculated from the customer' and to exert maximum leverage while managing uncertainty. Strategic investment and implementation with an eye on the future will determine the success of the next generation of business.

Related Articles