It is recommended to read selectively in conjunction with the key points below.
I. Group Chats and Message Management: The User Dilemma Amidst Information Deluge and Social Burden
II. File Transfer and Search: A Paradigm Shift from Machine Logic to Human Cognition
III. Chat Histories: The Largest Personal Data Asset in the Digital Age
IV. The Blind Spots of Contacts: The Powerless Network of Relationships
V. The "Death" of Moments: From Social Sharing to an Information Quagmire
VI. The Foggy Path of Super Apps: Imbalance Between Function Integration and Complexity
Conclusion: Exploring Next-Generation Instant Messaging—From Reflection to Action
In the field of product design, the most fundamental mission is to create high-quality tools that solve practical problems and deliver lasting value. Although every product professional brings their own subjective judgments, values, and emotional appeals, this article will strive to be objective, using WeChat as a case study. Starting from the basic functions of instant messaging, it explores how to better meet humanity’s fundamental communication needs in the current era.
As the core hub of China’s internet ecosystem, WeChat has evolved from a lightweight instant messaging tool into a "super app." It is no longer merely a communication tool but also integrates social networking, payments, content consumption, and app distribution. However, as its functions continue to expand, user experience faces multiple challenges: distraction due to information overload, management troubles caused by complex features, and usage barriers created by its closed ecosystem.
Zhang Xiaolong is hailed as the "god of product design," and WeChat is an excellent product and a way of life. Precisely because of its outstanding performance, innovation in the field of instant messaging seems to have become an unapproachable topic—like an elephant in the room, it looms large but few are willing to discuss it head-on. Users have more or less grievances, yet they feel daunted and unsure where to begin improving the situation.
No matter how excellent WeChat becomes, the demand for a more ideal communication tool will always exist. WeChat’s strong network effects and ecological advantages should not be reasons to abandon the pursuit of better product forms. On the contrary:
- Precisely because of its massive scale and deep penetration does the new need to manage massive amounts of information become more prominent.
- Precisely because it affects all aspects of life and work does the exploration of better functional implementations become more urgent.
- Regulatory pressures, migration costs, and ecological impacts are also reasons for stagnation.
The path of innovation is never smooth, but the pursuit of better product experiences is always worth our effort. Facing the "mountain" of WeChat, we need not only to look up but to think about how to climb higher.
I. Session and Message Management: The User Dilemma Amidst Information Deluge and Social Burden
WeChat’s message presentation method leads to information overload, where important content is submerged in a sea of chats, forcing users to bear the high cost of information filtering.
Group chats and sessions, as social tools, should enhance the efficiency of information acquisition. In their early stages, they did indeed significantly accelerate information exchange. However, today they have the opposite effect—especially on dominant social platforms like WeChat, where users battle information overload daily.
Every morning, or even after just an hour away from their phones, users’ first task is to process a mountain of WeChat messages. Confronted with an endless stream of group chat notifications, people often feel overwhelmed; sometimes, finding an important notification from the previous day becomes a challenge. This phenomenon highlights deep-seated issues with how current social platforms present information.
WeChat’s linear message presentation exacerbates this dilemma. Conversations of varying themes, purposes, and contexts are listed in a single feed sorted by update time, constantly being scrolled past. Within sessions, there is no hierarchical discussion mechanism, mixing useful information with casual chatter. When scrolling through messages, users struggle to quickly identify and locate important information, sifting through a sea of noise akin to searching for a needle in a haystack.
Although platforms like Slack offer threading features as a solution, this approach is far from perfect. It instead increases users’ operational and viewing costs; what seems like a helpful design has become another form of cognitive burden.
On the surface, building more granular information architectures and retrieval tools—adding conversation categorization, topic tagging, sub-discussion threads, and multi-dimensional search criteria—might seem to provide clear navigation coordinates amid information overload. Yet this approach still fails to address the core issue.
The world and information are inherently prone to entropy increase; tools should help humans counteract this. Ideal tools should solve problems subtly, even making users unaware of their presence. Adding layers of filtering tags and granular options, over-reliance on complex filtering mechanisms and detailed criteria may convenience a minority but cannot meet the needs of hundreds of millions of ordinary users to efficiently manage social information.
WeChat’s shortcomings in group chat and message management expose its limitations in addressing information overload. While IM and group chats have the potential to accelerate information access, the lack of efficient management mechanisms has turned these features into user burdens. How to enhance information filtering and management efficiency while maintaining information flow has become a core challenge for WeChat and all instant messaging tools.
There are many potential product solutions to this problem. An ideal solution should embody the following traits:
1.Proactively providing fine-grained session and message filtering strategies
Instead of cluttering interfaces with excessive buttons and options, natural language interactions should enable better filtering strategies, allowing users to express their intent: “When did I last chat with a friend about this?” “Which friends have I discussed this with?” “Content related to my family/this trip” etc. Users struggle to describe specific time ranges or include/exclude keywords using regex-like language to convey filtering intent.
Beyond basic natural language interaction, it is critical to recognize user intent and context within chat scenarios to proactively automate message filtering strategies—tools that grow increasingly attuned to users. Users should no longer need to actively query or articulate needs. Chat scenarios are where user intent is most fully revealed, more so than content consumption scenarios. Richer intent exposure provides more comprehensive data for better filtering.
2.Personalized information filtering: distinct from content recommendation
In social scenarios, we need filtering, not recommendation. Humans have a voyeuristic tendency, and our social nature compels us to follow the dynamics of our immediate circles. This differs entirely from the information consumption experience of short videos or text-and-image content. In this regard, we still need to improve user experience in acquiring social information as algorithms do for content recommendation—focusing on enhancing the intensity and utility of accessing relevant social information. The core remains improving the efficiency of accessing and ingesting relevant information.
WeChat has already solved the problem of information “having or not having.” Now, under users’ limited time and attention, we must further boost information access efficiency and experience. The “most understanding” recommendation algorithms are criticized today for creating information cocoons and time furnaces, largely because their ultimate goal is to increase user engagement time. However, human information acquisition and social experiences are not better with longer durations; extended content consumption only intensifies dizziness and boredom. In social and communication scenarios, filtering information in the “most understanding” way to fulfill users’ needs in life, work, entertainment, etc., can deliver a quality social experience.
II. File Transfer and Search: A Paradigm Shift from Machine Logic to Human Cognition
Instant messaging platforms should transcend traditional storage-oriented models and adopt information interaction paradigms aligned with human cognition. After all, IM is not just a file transfer channel but a socialized information processing space for real-time collaboration.
Instant messaging platforms are critical venues for human-information interaction. However, mainstream IM platforms represented by WeChat exhibit significant experience gaps in file transfer and management, forcing power users to frequently switch between tools. This fragmentation stems not only from technical limitations but, more profoundly, from a fundamental conflict between existing file management paradigms and human cognitive mechanisms.
IM platforms have become critical file transfer scenarios due to their unique social attributes and collaborative features. In the natural flow of work and life information, file transfer often extends conversations: photos/videos need sharing, meeting notes require real-time dissemination, travel plans demand instant discussion, and project documents must be delivered promptly. More importantly, IM platforms uniquely support scenarios where multiple users interact with the same information simultaneously—viewing the same file in real time, discussing its content, and advancing work progress through dialogue. Users choose IM platforms for file transfer not because they are the most professional tools, but because they are the most natural venues for interpersonal communication and collaboration. This makes IM platforms’ file management capabilities directly impact the daily collaboration efficiency of hundreds of millions.
Yet current IM file management systems remain stuck at the basic level of transfer and storage (and even fail here), ignoring the uniqueness of file interaction in social contexts. In social file sharing, information flow is accompanied by instant discussion, feedback, and revision—dynamics traditional file systems cannot fully support. Managing files in collaborative scenarios requires not only individual user experience but also group collaboration efficiency and smoothness, presenting new challenges to file management systems.
WeChat currently faces multiple limitations in file management: files expiring; 200MB limits on regular files and 1GB video caps forcing users to turn to external cloud storage; clunky file opening experiences requiring third-party tools for different formats; and scattered attachments across chats lacking a centralized “file library” for quick access. These issues are particularly acute for power users and affect daily life as well.
More worthy of reflection is that current file management systems still adhere to traditional paradigms designed for computers, not human cognition—a conflict with natural human information processing mechanisms. The capacity limits of human working memory (Miller’s Magic Number) mean we cannot process complete file contents like computers; our cognitive systems favor extracting key details and conceptual knowledge. Yet existing systems force users to toggle between entire files, a disconnect that violates the economic principles of human cognition.
Human brain mechanisms offer critical insights: associative memory suggests building file organization based on semantic associations; episodic memory implies the importance of context-aware retrieval systems; and working memory limits demand smarter information summarization and concept extraction.
Most of the time, our brain’s context window is limited. We neither need nor can load complete file contents into our brains; what we truly need are key details and high-dimensional concepts from files.
As natural venues for human-information interaction—especially in collaborative scenarios—IM platforms hold immense innovative potential. Future file management should transcend the traditional “storage-retrieval” model and adopt a new “concept-semantic” paradigm:
- Building multimodal file representation systems supporting semantic search.
- Enabling automatic concept extraction and knowledge graph construction from file content.
- Providing information display methods aligned with human cognition.
Based on this analysis, future file management systems should:
- Achieve semantic understanding and automatic summarization of files.
- Build intelligent recommendation mechanisms based on usage contexts.
- Offer information display interfaces aligned with cognitive patterns.
- Support seamless cross-platform file flow and collaboration.
This requires rethinking the essence of file management—from meeting machine storage needs to adapting to human cognitive traits. In IM platforms, natural venues for human-information interaction, we have the opportunity to create information access paradigms more aligned with human cognition. File system design must shift from “storage for machines” to “cognition for humans,” centering on information organization and retrieval paradigms that match how human working memory operates. This is not just a technical evolution but a necessity to enhance human information processing efficiency.
III. The Fate of Chat Histories: How to Take Ownership of the Largest Personal Data Asset in the Digital Age
Personal chat histories in instant messaging platforms have become the most valuable personal data assets of the digital age, yet they face not only underutilization but also loss risks.
Today, instant messaging platforms are the most critical containers of personal digital footprints. However, platforms like WeChat exhibit significant shortcomings in chat history retention and migration: backup processes during device changes are cumbersome and uncertain, lengthy migrations often interrupt, and comprehensive cloud archiving is nonexistent. By contrast, Telegram offers robust cloud sync, and WhatsApp supports full chat exports—though their data utilization remains traditional, users can at least independently save and manage their histories. Mainstream IM platforms lag in data management philosophy overall.
This issue transcends technology. Never before in human history have so many personal interactions, thoughts, and emotions been comprehensively recorded. While a few creators leave digital footprints via publications or social media, for most people, daily chat histories are their primary and most authentic expression. These seemingly trivial conversations accumulate into contemporary individuals’ most complete “digital legacies,” carrying irreplaceable historical and emotional value. They are not only the most genuine traces of personal existence in the digital world but also vital carriers of collective memory. Beyond publicly available online information, the largest and most valuable personal data resides precisely in these daily dialogues.
In the AI era, the value of this data will grow further. Personal chat histories are foundational materials for building personal knowledge bases; processed through AI, they can extract valuable insights that may become key components of personal competitiveness. Yet under current platform dynamics, users often lack full control over their data. Platform monopolies force users to accept imposed rules, and the sunk costs of vast historical data make platform migration difficult.
This must change. Data self-determination should be a basic human right in the digital age; users deserve full control over their data, including the right to freely migrate it. We need to rethink personal data management models, treating these digital assets as prudently as financial assets.
The balance of interests between platforms and users must be rebalanced. While platforms may prioritize data monopolies for commercial reasons, users cannot entrust their most valuable digital assets solely to corporate ethics. “A state is secured not by borders but by the hearts of its people”—platforms retain users not by erecting data barriers but by delivering genuine value. Every user should proactively export and save their chat data, not only to safeguard digital rights but also to secure future value.
Platforms should transition from “data custodians” to “value enablers” for user data. This means providing more reliable storage, easier migration tools, and more open data APIs, empowering users to fully leverage their data. Platforms should retain users not through “data walls” but by offering superior communication functions and data services, achieving mutual success. Only when our digital memories can flow freely and be fully utilized can the immense value of chat histories be truly unlocked.
IV. The Blind Spots of Contacts: The Powerless Network of Relationships
Digital-era relationship maintenance faces a fundamental contradiction: the scale and complexity of social networks far exceed humanity’s innate cognitive capacity, yet existing contact management tools lack sufficient intelligent support.
Humans have a natural limit on the number of stable social relationships they can maintain—the famous “Dunbar’s number” of ~150. Yet digital-age social demands exceed this limit. WeChat users often have hundreds or thousands of contacts, far beyond the brain’s natural processing capacity. This magnitude leap creates unprecedented management pressure: users must remember each contact’s background, social context, and interaction history—information exceeding the “default settings” of human evolution. Human memory of social relationships relies on context and frequency, but digital address books sever this natural mechanism.
Compounding the issue, current digital contact designs remain stuck in simple flat management models. Users organize contacts via basic remarks and tags, incompatible with humanity’s innate hierarchical, contextual cognition. For example, a contact might be a “college classmate,” “gaming buddy,” and “badminton partner,” but maintaining such multidimensional relationships with existing tag systems is costly and inflexible. When switching devices or migrating accounts, these painstakingly maintained social details often vanish, requiring users to rebuild them.
Digital social networks exhibit a typical “ripple structure”: centered on the self, with relationship strength diminishing outward like waves. The core circle includes family and close friends requiring frequent interaction; the periphery extends to numerous low-frequency acquaintances. This layered, differential social structure is no longer confined to traditional kinship or geography but incorporates diverse connections. Users face multiple challenges in this modern landscape: carefully nurturing core-circle intimacy while preventing valuable weak ties in the periphery from fading due to neglect. Of particular note are “structural hole” contacts—those bridging different social circles. Though low-frequency, they play unique roles in information flow and resource complementarity. These relationships have natural hierarchies and are both manageable and adjustable, demanding user effort to maintain.
Unlike offline socializing, digital interactions drastically reduce the spatiotemporal costs of maintaining relationships but introduce new challenges. First, digital-era relationships are more fluid than in traditional societies; without shared contexts, important connections may fade imperceptibly, shifting outward. Second, moderns frequently switch social roles across scenarios, raising the bar for contact grouping and tagging—complexity far exceeding the past. Finally, due to the “sunk cost effect,” users often over-maintain relationships of limited actual value, diverting energy from nurturing core circles. These new issues call for smarter tools that align with innate social cognition while adapting to digital-era dynamics.
To address these challenges, contact management requires upgrades across three levels:
First, foundational improvements. Introduce batch operations accessible to average users, support multidimensional smart grouping, and enable cross-device data migration/sync. More critically, simplify operations via natural language interaction, allowing users to effortlessly execute complex tasks like “send New Year greetings to all elementary school classmates” or “tag those who helped me in the past month.” Natural language interaction holds significant potential here.
Second, enhanced intelligence. Systems could analyze chat content and interaction frequency to automatically create/update contact tags, log key relationship milestones, and remind users to nurture important connections at appropriate times. These features must be built on strict privacy protections, ensuring users retain full control over their social data.
Third, social efficiency tools. Leveraging the “80/20 rule,” help users identify and prioritize maintaining the most valuable 20% of core relationships while reducing maintenance costs for the remaining 80% via automation. Automation should extend beyond generic personalized greetings to interactive tools that spark user recollections, fostering meaningful engagement to sustain moderate social warmth.
The core value of these improvements lies in helping users better navigate digital-era social challenges: reducing cognitive load, boosting management efficiency, and optimizing social resource allocation. With advancements in large language models, future contact management will grow smarter and more humane, better understanding and serving users’ social needs.
Tool intelligence should not lead to social mechanization. Genuine relationships are built on sincere interaction; technology’s role is to assist, guide—not replace—this sincerity. Thus, while pursuing efficiency, we must preserve users’ autonomy, ensuring social management tools remain warm assistants rather than cold automation systems. Human-in-the-loop is the core principle. Using AI to augment human social interactions demands more product innovation and ingenuity.
As platforms like WeChat evolve, user social management needs continue to grow. WeChat’s recent “gifts” feature is a positive attempt to digitize the traditional practice of reciprocal etiquette. Future explorations must focus on achieving smarter, more humane social management while protecting privacy—a direction worthy of sustained effort.
V. The “Death” of Moments: From Social Sharing to an Information Quagmire
Moments is undergoing a silent crisis: it is not dying from feature scarcity but from information overload and social alienation, as users gradually lose the ability to access effective information and engage in authentic social connections amidst a chaotic information stream.
When the first Moments post was created, it embodied users’ desires to share life and nurture relationships. A decade later, this social space faces a misalignment of information attributes—vast amounts of content irrelevant to users’ actual social needs dilute genuine connections.
This misalignment manifests across multiple dimensions:
- Over proliferation of opinions and statements. Users eagerly share personal views, but these often exceed the natural needs of daily friend interactions, creating social pressure for both parties. Some users, fearing backlash, reduce sharing of everyday content.
- Covert spread of commercial information. Merchants and games exploit viral sharing mechanisms, turning users into unwitting vectors for promotion.
- Visibility range management challenges. Users often struggle to accurately determine the best audience for their posts, a concern that suppresses natural social sharing.
Solutions may lie not in directly intervening with information flow—lessons from Facebook and Twitter show forcing changes to browsing habits often backfires. Instead, we can start at the source, assisting users when posting content: for example, recommending appropriate visibility ranges based on content attributes and social relationships, or supporting flexible permission management (e.g., “Subscribe to my Moments for a month?” popular among younger users). Gentle guidance could better align information flow with social needs.
Meanwhile, users face growing cognitive loads. In the information age, “scrolling through Moments” is inherently contradictory: users want timely updates on friends’ lives and important news, yet must filter through non-essential content. This intense information processing transforms a once-relaxed social activity into a burden. Worse, facing information anxiety, users often adopt extremes—“browse all” or “give up entirely”—further weakening social connections.
These phenomena stem from product mechanism limitations. Moments’ timeline display, though intuitive, struggles to adapt to complex social needs. For instance, users cannot easily revisit content from specific periods or topics, nor adjust display priority based on relationship closeness. Traditional grouping features offer limited filtering but have high maintenance costs, failing to meet dynamic social contexts.
Social product evolution faces a fundamental paradox: the explosion of information versus inherent limits in human social cognition. Core metrics for social products should not be maximizing dwell time but optimizing information acquisition utility. We need tools that help users efficiently access truly valuable information while reducing cognitive load. A successful social product must not only solve information efficiency but also repair technologically alienated human connections.
Specific explorations could include:
- At content creation: Use AI to recommend optimal visibility ranges based on historical interactions and content features, reducing social friction. Support flexible permission mechanisms for users to adjust content display strategies across scenarios and relationships.
- At information access: Offer diverse content organization methods. For example, proactively suggest retrieving content by topic, timeframe, or interaction intensity, or provide smarter summaries to help users grasp key social updates. However, users should choose these tools, with better presentation guiding active filtering rather than brute-force replacement of timelines.
- At interaction design: Rethink how to foster authentic connections. Simple likes and comments no longer suffice for deep social bonds; designing new interaction modes that maintain lightweight yet meaningful relationships deserves ongoing exploration.
Moments’ evolution dilemma is ultimately a social one: in pursuing personal utility maximization, participants inadvertently harm the collective social ecosystem. This “broken windows effect” is pronounced in social networks—once someone uses Moments for marketing or attention-seeking, such behavior spreads virally, diluting quality content. This is not a design flaw but an inherent entropic trend of social networks.
VI. The Foggy Path of Super Apps: Imbalance Between Function Integration and Complexity
In its pursuit of omnipotence, WeChat consumes users’ cognitive resources through fragmented experiences. The future belongs to intelligent interaction systems that make complex functions simple and accessible.
WeChat’s growth trajectory is one of relentless expansion. From its origins as a lightweight instant messaging tool, it has evolved into a comprehensive platform integrating chats, Moments, payments, mini-programs, and more. Each functional expansion has tested users’ cognitive limits. With over a dozen features on the Discover page, millions of mini-programs, and constant new additions, the app has strayed from its original “ease of use” ethos.
WeChat’s current issue is not a lack of features but fragmentation between them. Mini-programs, as independent lightweight apps, lack organic interaction with the main chat interface—a disconnect that severely undermines product coherence. Users must constantly jump between interfaces, each transition draining attention and testing patience.
Mini-programs have achieved commercial success, building a thriving ecosystem. Yet from an IM product perspective, this is an suboptimal solution. Their success has hindered WeChat’s progress in deeply integrating chat scenarios and exploring innovative interactions. This siloed development model, to some extent, sacrifices the continuity and wholeness of the IM experience.
WeChat now resembles an over-adapted species. To meet diverse demands, it has evolved new functional “organs” but lost systemic coordination. Each feature operates like an independent ecological niche, lacking organic links—an evolutionary path that may ultimately collapse the system.
Breaking this impasse hinges on innovating interaction paradigms. Alipay’s “Zhixiaobao” (a voice assistant) offers an inspiring direction—using natural language to invoke services. This aligns with IM’s conversational essence and represents a potential future path.
Specific innovations include:
- Rebuilding function entry points with large language models.
- Transforming mini-programs into API-driven services.
- Constructing unified, seamless interaction experiences.
The next decade of mobile internet may witness a revolution in interaction paradigms. As large language models become intermediaries between humans and machines, users will no longer need to memorize complex operation paths; instead, they will summon services through natural dialogue. This is not merely a change in interaction style but a restructuring of humanity’s relationship with the digital world.
In this transformation, true winners will not be platforms with the most features but those that make complexity simple. WeChat must evolve from a feature aggregator to an experience reconstructor, simplifying complexity through intelligent interactions so technology truly serves humanity. This may cause short-term growing pains but could ultimately boost business efficiency and uncover latent demands in the long run.
Conclusion: Exploring Next-Generation Instant Messaging—From Reflection to Action
In an era of rapid evolution in human communication, we must not only acknowledge issues with mainstream platforms like WeChat but also ponder paths to the future. The pain points analyzed—chat management, file transfer, contact maintenance—ultimately converge on a core question: How can instant messaging tools truly adapt to human cognitive traits and social needs?
We stand at a unique historical juncture. Advances in AI offer new possibilities to redefine communication experiences. Large language models bring natural, fluid human-machine dialogue paradigms, while multimodal technologies enrich how information is expressed and understood. These innovations should not merely layer functions but serve users’ core needs:
- From chaotic information bombardment to intelligent information management.
- From fragmented functional experiences to seamless conversational interactions.
- From passive tool use to proactive scenario understanding.
- From closed data silos to open value networks.
Postscript: Notes from WeChat Open Class
Taken at Berkeley Summit House
If you’ve read this far, you likely share these concerns and yearn for better solutions.
Over a decade ago, WeChat’s mission was to create a niche chat tool for those who avoided QQ. Today, in the new era where AI reshapes human-information interaction, we hold a similar aspiration: to build a great communication tool for those who resist WeChat.
We do not seek to disrupt existing ecosystems or obsess over grand innovation narratives. We aim to pragmatically solve the issues raised here: making chats less anxiety-inducing, file management smarter, and social relationship maintenance more natural. We are turning these visions into reality, line by line of code.
This exploration needs more like-minded partners. Whether you wish to test, contribute code, provide resources/funding, or simply share ideas—we warmly welcome you. You may hail from computer science, cognitive science, psychology, human-computer interaction, or bring unique perspectives. Every background has the potential to spark innovation.
We are already on this journey. Guided by user needs, rooted in first principles, we hypothesize boldly and verify cautiously. Not to change the world, but to make daily communication a little better. If this vision resonates, join our exploration.
Chen Chunyu
Polaris-c-
[email protected]
ChatGPT and Claude also contributed to this article.
Original link: https://mp.weixin.qq.com/s/aeSVLFMpGF7cY6jgJkda-A
