User Experience

Viewpoint: The mini-programs of major AI models merely embed the web-version features, making it convenient for users to access different services within the same app, but there is no essential difference.

Logic chain: Mini-programs integrate models such as Claude and Doubao in the form of embedded web pages for the purpose of convenience. In practice, their actual capabilities are no different from opening the pages in a browser—only slightly more convenient.

Failure condition: If a mini-program integrates additional features or offline capabilities, it may then make a difference.

Related field: AI applications

View: Chinese AI companies lack truly human-like assistants in the vein of Claude and ChatGPT. Such assistants should have a sufficiently broad set of skills and a well-designed, effective memory system. In fact, what non-technical users rely on has little to do with raw model capability.

Logic chain: The dependence of large numbers of non-technical users on AI stems from the completeness of the assistant's experience layer—such as remembering context and integrating multi-channel tools. This makes skills and memory more decisive competitive factors than model performance. Ignoring the experience layer while chasing only model parameters makes it difficult to win the mass market.

Invalidation conditions: If the target users are technical developers or professional researchers, their sensitivity to underlying model capability is far higher than to experience packaging. In that case, model capability takes priority over human-like design.

Related areas: AI consumer applications, product experience design.

Claim: SaaS tools that close off their ecosystems at the basic-feature level and refuse to interoperate with external services undermine user experience and efficiency, ultimately hurting the product itself.

Logic chain: Cross-tool data interoperability is a basic efficiency need. For example, when Feishu Calendar cannot be subscribed to by other calendar apps, it creates data silos, adds friction to user workflows, and pushes users toward more open alternatives.

Failure condition: When a product has formed extremely strong lock-in or a monopoly position, users may have no choice but to tolerate it—but the long-term risk of user churn remains.

Related domains: Product & Operations; SaaS

【Insight】In task management tools, if the visual feedback for completing a task is too subtle, it diminishes the user's positive reinforcement experience, affecting their sense of satisfaction and continued motivation to use the product. 【Logic Chain】Taking TickTick as an example, the checkmark box is too small and the completion animation is barely noticeable, leaving users lacking psychological reward when finishing tasks. This insufficient positive reinforcement makes the product feel less polished and satisfying to use. 【Failure Condition】If users treat the tool purely as a serious task recorder rather than something requiring motivation, or if they derive satisfaction from external sources, weak feedback has limited impact. 【Related Areas】UX design for tool-type products, feedback and incentive mechanisms in habit-building applications.

Viewpoint: After extended use of new energy vehicles (NEVs), their overall experiential advantages make it difficult for users to return to gasoline cars, creating an irreversible shift in consumer habits.

Logic Chain: After actually using the vehicle for one and a half months and verifying its range performance, the user fully experiences advantages in driving, riding, and charging, concluding that buying a gasoline car again would be impossible—demonstrating that the product strength is sufficient to change deeply entrenched consumer preferences.

Failure Conditions: If battery technology regresses, charging infrastructure becomes extremely scarce, or electricity costs surge, the experiential advantage could be reversed.

Related Fields: Consumption & Lifestyle, Product & Operations

Viewpoint: If a content platform does not support creator formatting needs such as Markdown, it will seriously dampen creative enthusiasm and suppress the production of high-quality content.

Logic Chain: Creators are accustomed to using Markdown for efficient typesetting. If the platform does not support it, formatting becomes messy and publishing costs rise, ultimately driving users to switch to platforms with better support.

Failure Conditions: This impact is diluted if the platform excels at short-form text or social interaction, or possesses an irreplaceable network effect.

Related Fields: Product & Operations, Content Creation

View: Complex pricing models (such as allocating quotas by feature category) increase user confusion and resistance, harming the product experience.

Logic chain: Users expect a simple, unified usage allowance. When a product provides separate credit quotas for different feature categories, users must spend extra effort understanding and managing each quota. This can make them feel manipulated, increasing decision cost and usage friction. Clear and simple billing rules are the foundation of good product design; otherwise, potential users may be driven away.

Failure condition: If separate quotas align strictly with the underlying cost structure, are highly transparent to professional users, and are accompanied by clear explanations and demonstrated benefits, some professional users may accept them. Certain complex billing models (e.g., cloud services) have become industry norms, and users have already formed mental models around them.

Related fields: Product & Operations, Business Model

Viewpoint: When membership benefit upgrades in the service industry fail to match users' actual scenarios, they expose the company's intent to hit internal metrics rather than deliver real value.

Logic Chain: A user upgraded to a high-value suite during an extremely brief stay has no time to enjoy any added benefits. Such a "perk" disconnected from real needs signals that management cares more about internal assessments than customer experience, which in turn triggers negative perceptions.

Invalidation Condition: The upgrade could have been meaningful if the benefit could be retained for the next visit or translated into practical convenience, but in this case neither was possible.

Related Fields: Customer service system design, membership loyalty, critique of formalism.

Viewpoint: On the basis of reliable quality in core powertrain components, interior quality and driving experience are what truly enhance the sense of happiness from car ownership.

Logic Chain: As the automotive industry matures, the reliability of powertrain systems in mainstream models has generally reached acceptable standards, and differences in powertrain performance are barely noticeable in daily use. In contrast, interior craftsmanship, human-machine interaction, and comfort features directly affect the mood and satisfaction of every trip, making them the decisive factors in automotive happiness.

Failure Conditions: When a vehicle's core powertrain has obvious quality defects, or when performance driving remains the primary purpose—such as in sports cars or track scenarios—the weight of interior experience diminishes.

Related Domains: Consumption and Lifestyle

Opinion: For video membership services, offline caching is a rigid requirement; restricting caching severely harms the user experience.

Logic chain: Users purchase memberships expecting to watch anytime, anywhere; offline caching is indispensable in no-network scenarios such as flights. Copyright restrictions that prevent caching directly undermine users' core usage scenarios and trigger strong dissatisfaction.

Failure condition: If the platform compensates by providing advance notice or alternative download options, the negative experience can be partially alleviated, but not completely eliminated.

Related areas: Product design, membership systems.

Opinion Progress messages pushed purely based on the passage of time create far more anxiety for users than any reminder value they provide.

Logic chain Notifications like “57% of this week has passed” only emphasize time consumed, ignoring actual output. Users may easily feel that they have accomplished nothing, leading to anxiety—this is poor product design. Good progress management should link to goals and outcomes, rather than merely selling time pressure.

Conditions for failure If combined with personalized goals and adjustment suggestions, such reminders may have a positive effect.

Related fields User experience, product design

Viewpoint: A platform’s extremely low withdrawal threshold (e.g., RMB 10) may seem trivial, but it can leave users with small earnings unable to withdraw for a long time, create the impression of a system malfunction, increase customer support costs, and harm user experience.

Logic chain: When a user’s earnings fall below the threshold, the platform neither proactively pays out nor sends reminders. After users forget about the earnings and later notice them by chance, they may suspect a system anomaly; when they ask customer support, they receive only the explanation of “insufficient accumulated amount,” which leads to a poor experience. They may then disengage or spread negative word of mouth.

Conditions that would offset this: If the platform proactively sends progress notifications, offers a visual display of accumulated earnings, or allows immediate small-amount withdrawals, the negative effects can be mitigated.

Related domain: Product and operations.

Viewpoint: When popular small independent apps are acquired by large companies, they often move toward expensive subscription models, hurting existing users' feelings and user experience.

Logic chain: To monetize quickly, the acquirer may eliminate the one-time purchase option or raise prices significantly, making longtime users feel betrayed. If the added features are not enough to justify the price increase, user churn is likely to rise.

When this may not hold: If the app keeps its original pricing after acquisition, offers a reasonable one-time purchase option, and provides loyal users with a smooth transition plan, the negative reaction will be weaker. If the subscription brings revolutionary feature improvements, users may also gradually accept it.

Related areas: Business model, product, and operations.

Viewpoint: AI search/research tools that provide overly detailed, in-depth information can cause cognitive overload for users, which in turn reduces their usefulness.

Logic chain: The deeper a tool's functionality and the greater the volume of information it returns → users must spend significant effort filtering and digesting it → fatigue and resistance develop → the tool is perceived as exhausting to use → its deep features are not effectively used, and the tool's value is discounted.

Invalidation condition: When users have a clear and complex need for deep research, detailed output is exactly what they need and does not feel exhausting.

Related fields: AI applications, product and operations.

Opinion: If AI research tools’ output format preservation and export features are not user-friendly enough, they will reduce users’ work efficiency and adoption rates.

Logic chain: For example, Grok Deep Research failed to preserve formatting, forcing users to adjust it manually. If it offered one-click Markdown copying and similar features, it could connect seamlessly to downstream workflows. Experience improvements directly affect users’ willingness to continue using the tool.

Failure conditions: This is acceptable only when users do not need to reuse the results or do not rely on formatting.

Related fields: AI application product design, user experience.

Viewpoint: The criteria for evaluating a product or service are not absolute. What adults see as a flaw may become a unique attraction from a child's perspective, so experience design needs to account for multiple perspectives.

Logic chain: Children become interested in construction sites, construction vehicles, and the like → adults view these as noise and landscape disruption → for families with children, the construction scene becomes a source of enjoyment for the child → the evaluation reverses → businesses can use this to create a differentiated experience.

Failure conditions: If the target users are all adults and are not differentiated, or if the negative factors involve fundamental problems such as safety, then the perspective difference fails to apply.

Related domains: Consumer and lifestyle, product and operations, user experience.

Viewpoint: Linear's input interaction smoothness is significantly better than traditional tools such as Jira and Confluence, improving user experience and efficiency.

Logic chain: Traditional management tools suffer from response latency during input, complex interfaces, and counterintuitive operations, while Linear creates a 'silky' interaction feel through clean design and instant feedback, reducing usage friction.

Failure conditions: If users require Jira's highly configurable workflows or Confluence's documentation ecosystem, Linear's simplicity may not satisfy complex needs; moreover, perceptions of interaction experience vary by individual.

Related fields: Project management tools, user experience design, team effectiveness.

Viewpoint: The design of functional products cannot merely satisfy static numerical specifications; it must fully account for users' dynamic behavior in actual use and their psychological safety buffer. Otherwise, it will lead to a poor experience and even safety hazards.

Chain of reasoning: Taking the standing long jump as an example, ensuring only the required distance while placing a wall immediately behind the landing point ignores the body's inertial buffer after landing and the psychological need for safety. This is a typical design mistake detached from real usage scenarios. This error reflects the designer's disregard for user behavior patterns and common safety knowledge.

When this does not apply: If the product is positioned solely for viewing or conceptual display rather than actual use, the requirement for a safety margin is substantially reduced; if the usage scenario is led by highly trained professionals who are clearly aware of the risks, a smaller safety margin is acceptable.

Related fields: Product design, user experience, safety standards.

Viewpoint: Even if the free quota is reduced, as long as it still far exceeds competing products, users will continue to perceive the product as “conscientious” and loyalty will not be damaged.

Logic chain: Ideogram’s free quota has been reduced from 25 prompts per day to 20 prompts per day. Although the absolute number has decreased, it remains extremely generous relative to the market → users perceive the relative advantage rather than the absolute reduction → comments describing it as “Buddha-level conscientiousness” indicate that users give it a high degree of tolerance and goodwill.

Invalidation conditions: If competitors significantly increase their free quotas, or if the product’s own service quality visibly declines, users may reassess the standard of “conscientiousness.”

Related fields: AI applications, business models.

View: In everyday AI usage scenarios, model response speed is an overwhelming experience factor, and users tend to reflexively choose the fastest-responding model even if there are minor differences in capability.

Logic chain: When using products like Poe, users have already entered a state of prioritizing the 4o model without a second thought because of its extremely fast speed. Speed directly shapes the sense of fluency, which in turn determines usage frequency and dependency. This reveals a key point about AI product competition: latency optimization itself is the strongest moat.

Failure conditions: When a task demands extremely high accuracy and professionalism, and slower speed only adds acceptable waiting time (such as long-form writing or complex reasoning), users may prefer a more accurate but slightly slower model; however, in the vast majority of lightweight, high-frequency daily conversations, speed still rules.

Related fields: AI applications, product and operations