Summary of Key Points
Many SaaS companies currently falling into the misconception of superficial AI integration: forcibly adding AI buttons and features without seeing any increase in revenue. Instead, the rising cost of computing power erodes profits (resulting in a 10% to 15% decrease in gross margin). The correct direction for transformation is towards "new SaaS"—by reengineering with a headless architecture and adaptable user interfaces, the software can shift from being adapted to users to adapting to them. Using natural language as the entry point and dynamically generating interfaces on demand not only solves the issues of cost and value associated with traditional transformations but also opens up new revenue opportunities.
The Pitfalls of Traditional SaaS AI Transformations: Increasing Losses Through Superficial Efforts
Many companies treat AI transformation merely as a matter of adding labels, such as inserting AI buttons on every page or embedding features into processes. However, these changes are often superficial. For example, adding an "AI Analysis" button to a customer management system that still displays outdated data with just a fancy new interface yields no real benefits:
- No Revenue Increase: Users see these AI features as useless and are unlikely to pay extra for them.
- Costs Rise: Every time AI is used, computing power is consumed, leading to a 10% to 15% decrease in gross margin, regardless of the value of the feature.
- Wrong Approach: The core of AI transformation should be to generate additional revenue, yet in many cases, AI becomes a burden on costs.
The Essence of New SaaS: From Fixed Interfaces to Dynamic Adaptability
The essence of new SaaS is to make software more flexible. This is achieved through two key components:
1. Adaptable UI: The interface is no longer static; it can be transformed into whatever is needed for a particular task. For instance, a simple input form is generated for frequently updated customer information, and a report interface is created for one-time annual sales planning, which can then be discarded after use.
2. Headless Architecture: The interface is decoupled from the underlying business logic. The basic business processes (like customer data and order management) remain in place, but APIs are provided for AI to utilize these functions and generate the appropriate interfaces.
Here’s a practical example: In a traditional CRM, updating customer status requires logging into the system, navigating to a specific page, filling out fixed fields, and submitting. With new SaaS, you can simply request, "Update the customer status for Company XX," and AI will generate a temporary interface that allows you to complete the task with one click, eliminating the need for a complicated process.
The Architecture of New SaaS: A Robust and Flexible Five-Layer System
New SaaS does not rely on patching over an existing architecture; instead, it establishes a new layered system that balances stability and intelligence:
- Foundation Layer: Established business logic (such as order management and customer data) is retained, and these functions are broken down into reusable APIs.
- Native UI Layer: Fixed interfaces are provided for users accustomed to the traditional workflow to prevent loss of existing customers.
- Rule Layer: Ensures security and compliance by preventing AI from making unauthorized changes to data. All actions are auditable to avoid issues.
- Inference Layer: The "brain" of AI, which understands natural language commands (e.g., "Generate an annual sales plan") and determines the necessary data and functions to use to generate the required interface.
- Experience Layer: The interface can be accessed on any device—mobile phones, computers, smart assistants, or even through enterprise messaging apps—without the need for a browser.
The Advantages of New SaaS: Overcoming Traditional Transformation Challenges
Compared to traditional AI integrations, new SaaS offers several significant benefits:
1. Simplified Usage: Tasks can be completed using natural language, eliminating the need to remember complex procedures (e.g., updating customer status with just a few commands).
2. Accurate Data: All operations are based on real business data, avoiding inaccuracies caused by general-purpose AI models.
3. Maintains Business Models: Existing subscription-based pricing and permission management structures are retained, allowing companies to continue their usual revenue models.
4 Enhanced Business Value: Fixed features become more flexible, enabling the creation of custom reports as needed.
5 System Integration: New SaaS can integrate with other software (e.g., financial and office systems) to perform complex tasks.
6 Efficient Use of Computing Power: Low-cost AI models are used for simple tasks, while more expensive models handle complex ones, ensuring cost efficiency.
How New SaaS Generates Additional Revenue
The ultimate goal of AI transformation is to increase revenue. New SaaS achieves this through three approaches:
1. Expanding Value Boundaries: By recombining underlying capabilities, new features can be created (e.g., combining customer data with sales forecasts to generate intelligent business recommendations).
2. Modular Function Assembly: Users can combine existing functions to create new ones (e.g., automating follow-up emails by linking customer tracking with email sending).
3. New Revenue Models: In addition to traditional subscription fees, new revenue sources can be generated based on the results provided by the software (e.g., earning a commission for helping customers earn more money or charging for customized reports).
In Conclusion
Traditional SaaS AI transformations are often about making superficial changes, while new SaaS focuses on fundamentally reengineering the product to better meet user needs. Only by fundamentally altering the software’s logic can AI transform from a cost burden into a revenue-generating asset. This represents a paradigm shift in the SaaS industry, not just a simple technical upgrade.