Summary of Key Points
Recently, SaaS company stocks in North America have plummeted, and there has been a widespread belief that "AI is killing SaaS." However, AI is not the culprit responsible for the decline; rather, it is a catalyst for a major reshuffle in the industry. It will eliminate SaaS companies that lack competitive barriers or have fragile business models, while benefiting those with complex industry-specific advantages, platform ecosystems, compliance-related data, or robust underlying infrastructure. The core of this reshuffle is that only companies that are "indispensable to AI" will survive, while those that can be replaced by AI will be eliminated.
I. Which SaaS Companies Will Be Eliminated by AI?
The following types of SaaS solutions have little chance of surviving in the face of AI:
1. Mass-market tools charged per user seat: Companies like Adobe generate 70%-80% of their revenue from user account subscriptions. With layoffs leading to fewer accounts (where each employee used to have their own Figma account, now shares are used), and the addition of token-based costs for every AI usage (similar to paying for image editing services), their gross margins are much lower than those of traditional software suites. If users are unwilling to pay extra for the additional cost of AI, these companies will struggle to make a profit.
2. Tools whose necessity is diminishing: For example, Photoshop, which once required months of training to use layers and masks, can now be easily used with simple commands like "make the sky blue" in natural language. Non-professional users no longer need such complex tools, and these software solutions have weak barriers to entry (e.g., projects can be easily transferred to Canva).
3. Data transfer-oriented SaaS: Many companies use tools like Notion and Slack to exchange data. With AI agents that can directly retrieve data across systems, the need for these tools as mere data transfer intermediaries is eliminated.
4. Creative, single-purpose tools: In the past, creating videos required the coordination of several different software applications; now, AI-native tools (like GenSpark) can complete the entire process in one interface, effectively replacing those older tools.
5. Companies with fragile business models: While revenue from per-seat fees may be stable, shifting to a token-based or outcome-based pricing model makes their financial models more vulnerable. For example, Salesforce charges based on conversions, and if no sales are made, there is no income, leaving many companies unable to withstand the increased costs.
II. Which SaaS Companies Will Be Benefited by AI?
The following types of SaaS solutions have barriers that AI cannot overcome:
1. Professional software for complex industries: Companies like Autodesk, which design nuclear power plants and dams, use software embedded with thousands of industry standards (such as building codes and ISO regulations) and tied to companies' compliance data and engineering history. Users face high migration costs, making it difficult for AI to replace their services entirely; however, AI can enhance the design process, increasing engineers' efficiency.
2. Platform giants with large user bases: Companies like Microsoft (Office 365) and Salesforce (CRM) have hundreds of millions of users and can integrate AI capabilities directly into their existing workflows (e.g., Copilot in Office). AI reduces the development barriers, but having access to a wide customer base is even more crucial, expanding these giants' market power.
3. SaaS providers dealing with compliance and sensitive data: Companies like Palo Alto Networks in cybersecurity and Intuit in tax software handle sensitive information and are subject to legal responsibilities (e.g., signing diagnostic reports). While AI can automate some processes, they still play a vital role in ensuring compliance, becoming more efficient tools rather than substitutes.
4. Infrastructure providers: Services like Snowflake for data warehouses and AWS for cloud computing require substantial computing power, storage, and data management as AI becomes more widespread. Their demand grows exponentially with the adoption of AI.
5. Vertical platforms serving high-value users: Companies like Pixieset, which serve professional photographers, allow photographers to focus on creativity rather than spending time on post-production tasks, expanding the market for these services and attracting more freelancers.
III. Three Key Changes Caused by AI in the SaaS Industry
This transformation is not about simple substitution but about reshaping industry rules:
1. Shift in business models from selling seats to selling results: The era of stable revenue from per-seat fees is over. Companies must move towards outcome-based or token-based pricing. However, this transition is challenging: while the cost may increase (from 1 yuan to 15 yuan for the same amount of work), they need to demonstrate how AI can create additional value that customers are willing to pay for (e.g., by helping engineers comply more efficiently).
2. Shift in barriers from technical code to softer factors: While code was once a significant barrier, AI has made it less valuable. New barriers include distribution capabilities (e.g., Microsoft's user base), compliance data, and organizational habits (companies' reluctance to change their work processes). Existing giants can leverage their existing customer bases, while new players can use AI-native solutions to attract new users.
3. Accelerated consolidation: AI reduces development costs, allowing platform companies to easily incorporate features from niche tools (e.g., Microsoft competing with Salesforce for customers). In the end, a few dominant platforms will dominate the market, leaving fewer companies standing, with the surviving ones becoming even stronger.
IV. Implications for the Secondary Market
The current slump in SaaS stock prices reflects both rational shifts in industry paradigms and irrational panic:
- Real risks: Companies like Adobe, whose business models and product usefulness are under threat, see justified declines.
- Overly criticized companies: Although their stock prices have fallen, those with strong barriers (like Autodesk) still have higher valuations due to market recognition of these advantages. Investors need to distinguish between companies that are truly struggling and those that are just being affected by panic.
The decline in SaaS stocks presents both opportunities and challenges for investors: some companies with real competitive advantages will recover, while those without them may see permanent losses.
Conclusion
AI will not kill SaaS as such; rather, it will eliminate poorly designed, static software solutions based on per-seat fees and highlight the rise of truly innovative platforms with enhanced AI capabilities and irreplaceable value. For entrepreneurs and investors, this is an opportunity to reassess which companies can adapt to the new market realities and which will be eliminated.
(The entire analysis is written in plain language to make it easy for non-experts to understand the impact of AI on the SaaS industry.)