Summary of Key Points
Recently, SaaS (Software as a Service) stocks have seen a sudden surge in value, with companies such as Snowflake and ServiceNow in the United States, as well as Kingdee in China, experiencing significant price increases. However, this does not mean that the entire software industry is safe. Instead, the market has begun to “stratify” SaaS companies: those that possess the private data required by AI and can create a “data flywheel” will see their value amplified by AI; whereas those that only focus on data recording, display, and lightweight collaboration without critical data will have their value reduced by AI. The catalyst for this trend was Snowflake’s financial report, which demonstrated unexpectedly strong growth thanks to its built-in AI tool, Cortex Code. This tool’s advantage lies in its ability to understand the “context” of customer data—such as historical issues with old projects and the specific definitions of data metrics—making it more practical than general-purpose AI.
Detailed Analysis
#### 1. What “counterintuitive” secrets did Snowflake’s financial report reveal?
Snowflake is a company that provides SaaS solutions for data management. Its revenue grew by over 30%, and its profits nearly doubled, far exceeding market expectations. But the most crucial aspect was CFO’s statement that Cortex Code was the main driver of this impressive performance.
Cortex Code is not an independent AI programming tool (like ChatGPT) but is integrated into Snowflake’s data platform. Many companies store their business data in Snowflake for analysis purposes. Therefore, Cortex Code doesn’t “intrude” on customers’ workflows from the outside; it emerges naturally as part of their data processing process, as they prefer using a tool that understands their own data better.
#### 2. Why can’t general-purpose AI compete with tools that come with built-in data?
While general-purpose AI programming tools (like Codex and Claude Code) are powerful, the biggest challenge when working with old corporate projects is understanding the context of the data:
- Why was this table designed in a certain way?
- How are data metrics (such as monthly sales figures) calculated?
- What mistakes have been made in the past (e.g., which fields are prone to errors)?
General-purpose AI lacks this contextual knowledge, but Snowflake’s Cortex Code, being integrated with the customer’s data, can access this hidden information. For example, when analyzing historical sales data, Cortex Code knows the source and potential issues of that data, allowing it to generate useful code directly; general-purpose AI might produce correct code logic but not align with the company’s actual situation, requiring manual adjustments.
#### 3. In the age of AI, has the SaaS industry’s moat become a “data flywheel”?
Not all data is valuable; only the data needed by AI matters:
- Traditional CRM records like customer names and phone numbers have limited value.
- Live data such as customer meeting recordings, sales strategy adjustments, and historical project issues are what AI needs (for example, AI can use recordings to gauge customer intentions or avoid repeating past mistakes).
More importantly, a “data flywheel” is created: the more customers use your product, the more data you generate for AI, making your AI tools more effective—and thus making it harder for them to switch to competitors. For Snowflake, the more data customers store, the better Cortex Code understands their business, leading to a mutually beneficial cycle.
#### 4. Which SaaS companies are truly at risk?
The market is starting to eliminate those SaaS companies that lack critical data:
- Tools that only perform basic functions like data recording and display (e.g., simple form software or basic task management tools) can be directly replaced by AI (e.g., ChatGPT can generate forms or help manage tasks).
- Companies without processes that are essential for AI integration (e.g., lightweight collaboration tools) can be integrated into chat applications, eliminating the need for them as standalone solutions.
The value of these companies will be reduced or even eliminated by AI.
#### 5. What does this surge in SaaS stock prices indicate?
This surge does not mean the end of the crisis for the SaaS industry but the beginning of a differentiation process.
There was concern that AI would eliminate SaaS, but the market has shown that it’s not the entire industry that is at risk; rather, there is a segmentation. In the future, the value of SaaS companies will depend on two factors:
- Do they manage data that AI needs?
- Can they create a “data flywheel”?
Companies with these capabilities will see their value amplified by AI; those without them will face increasing challenges. This surge is not a sign of a recovery in the software industry but rather the market’s way of identifying which SaaS companies will survive in the AI era.
In one sentence:
AI is not an enemy of SaaS; it acts as a filter, making companies with data advantages more valuable and those without them more marginalized. When considering SaaS stocks, focus not just on price trends but on whether the company possesses the data needed by AI and has a “data flywheel.”