Summary of Key Points
The central argument of this article is that for the past two decades, we have relied on keyword indices (such as Baidu Index and WeChat Index) to understand market trends. However, keywords only reflect what users have said and cannot capture the underlying real demands. Now, AI can identify users' intentions (for example, when a user searches for "How to choose between the Li Auto L9 and the WM Motor M9," the actual intention is to compare the models and prepare to make a purchase). The next generation of index products will evolve from simply counting keywords to inferring intentions. This will not only change market research tools but also restructure corporate functions such as marketing, product development, and sales, and may even give rise to an "intention economy"—where resources are allocated based on user intentions. However, we must also be cautious of the risks associated with the probabilistic nature of these intentions (such as misjudging market demands).
1. Keyword Indices: Only the Surface, Not the Real Demand
The biggest problem with keyword indices is that they treat language as demand. Here are two examples to illustrate this:
- Multiple expressions for the same need: Someone planning a trip to Japan might search for "Tokyo travel guide," "Japanese visa," or "Osaka family trip"—all of which serve the same purpose. Keyword indices would treat these as six separate data sets, missing the underlying common goal.
- The same keyword with multiple meanings: The word "apple" can refer to a fruit, a smartphone, or stocks; "new energy vehicle" could mean that a user is just starting to learn about the topic, comparing brands, preparing to buy one, or even looking for charging services afterward. These have vastly different commercial values, yet keyword indices combine them all in their statistics.
Therefore, keywords represent only the surface level of expression, while intentions represent the true underlying demand.
2. Intent Indices: More Than Tags—Probabilistic Judgments Based on Behavior
Intent is not a simple "buy or don't buy" label; it is a probabilistic assessment made by AI based on user behavior (what they search for, click on, how long they stay on a page, and whether they save content), as well as context (time, location, device). For instance, if a user searches for "How to choose between the Li Auto L9 and the WM Motor M9," AI might infer that 70% of their intention is to compare the models, 20% is to prepare for a purchase, and 10% is just casual browsing.
The next generation of intent indices will not merely provide you with statistics like "AI search volume increased by 40%"; instead, they will show you that there has been a decline in general interest in the AI sector, but an increase in demand for private deployment evaluations and recruitment. This indicates that the market has moved from discussing concepts to taking concrete action. These indices have evolved from mere popularity tools to systems that understand user needs and can answer questions like "What problems do users want to solve?" and "To what stage has the market progressed?"
3. Major Changes in Corporate Functions: Marketing, Product Development, and Sales All Need to Revolve Around Intent
Intent indices will radically transform how companies operate:
- Marketing: No longer focusing on keyword popularity but on addressing specific intentions. For example, if there is an increase in interest in comparing new energy SUVs, marketing efforts would focus on creating comparison tools and case studies to help users make decisions, rather than broadcasting general promotional ads.
- Product Development: No longer just adding more features; instead, products need to address unmet user intentions. If a customer hesitates to buy, it might be because the product's value is not clearly demonstrated to the company. In this case, the product should provide an ROI model or pilot solutions to help the customer make a decision internally.
- Sales: No longer pushing for contracts; instead, the focus is on facilitating the decision-making process. For example, if a customer is in the risk assessment phase, sales should provide additional safety information and customer testimonials rather than repeatedly quoting prices.
- Cross-departmental collaboration: Using intent as a common language across departments—marketing identifies intentions, product development addresses them, and sales advance them to avoid confusion.
4. The Potential of the Intent Economy: Huge, but Beware of the "Delusional Optimism"
The core of the intention economy is to allocate resources based on user intentions. Unlike the previous "attention economy" (which focused on attracting traffic and keeping users on a platform), this approach aims to make users choose your product and complete their tasks. However, there are several risks to be aware of:
- Intent is probabilistic, not absolute: A 35% increase in purchase intent might simply mean that users are researching, not necessarily planning to buy. Companies should not invest heavily without first conducting low-cost verifications (such as sending surveys or launching small-scale pilots).
- Data and privacy issues: Who has the right to infer user intentions? Could these tools induce false intentions? Platforms must use data responsibly, and users must have the right to know and control their information.
True leaders in this field do not rush to invest just because they detect an intention; instead, they identify it early, verify it cost-effectively, quickly align their efforts with the correct direction, and stop incorrect attempts promptly.
Conclusion
The shift from keyword indices to intent indices may seem like a change in the method of data collection, but it represents a fundamental reconfiguration of business logic. In the past, we focused on what users said; in the future, we need to understand what they plan to do. Corporate competition will also shift from competing for traffic to accurately understanding user intentions, effectively responding to them, and helping them complete their goals. Does your team still make decisions based on keywords, or have you already started paying attention to user intentions?