If you are a botanist with spare time and energy after work, ten years ago you could volunteer for Chinese Wikipedia, editing entries on beech trees or spiderworts. Five years ago, you could identify unknown wildflowers on Sina Weibo. Three years ago, you found a new platform for science outreach, answering “Which plants are suitable for a courtyard — beautiful and easy to maintain?” on Zhihu, earning countless upvotes. This year, you started leading paid “plant exploration walks” through the paid experience-exchange platform Zaih (在行), receiving direct compensation.

If you are a talented designer with creativity to spare, a decade ago you could only get private commissions through word-of-mouth. Five years ago, you hesitantly browsed witkey websites — but were frightened off by the sight of dozens of bidders fighting for a ¥500 logo. Today, you can showcase your work on Dribbble, waiting for discerning entrepreneurs like Luo Yonghao to recognize your talent and reach out.

If you are an expert romantic matchmaker, in the past your special skills only earned you free dinners and the admiration of friends. Later you became a legend on Baidu’s PUA forum, offering free advice to countless lovelorn souls. Today, you can answer questions on the app “O’ér” (偶尔), guiding people step-by-step through relationship challenges — but now, your assistance no longer comes for free.

While driving for ride-sharing after work or waking up to breakfast in an Airbnb rental has become a lifestyle, and the idea of sharing assets is taking hold, the sharing of knowledge, experience, and skills is also becoming an “economy.” Platforms such as Zaih, 8pm (8点后), Linglu, and Kaowo have emerged this year, enabling one-on-one paid knowledge sharing.

So what possibilities lie ahead as internet-based knowledge sharing moves beyond the era of free access?

Why Has Online Knowledge Sharing Always Been Free?

Knowledge sharing has been a driving force since the early days of the internet. It is the cumulative, voluntary contributions of millions of netizens that prevent cyberspace from becoming a barren desert. The free exchange of ideas has built a richer, more diverse social fabric online than in the physical world. Without the “cognitive surplus” of countless volunteers contributing in their spare time, we could never have rebuilt the Tower of Babel of knowledge — Wikipedia, Zhihu, and others — online.

Yet online knowledge sharing never evolved into a true “economy.” The unsung heroes contributing to Wikipedia are rewarded with idealism badges. Forum moderators may become internet celebrities; top Zhihu answerers gain immediate social capital but only those at the apex receive indirect rewards. More importantly, skills and experiences that cannot be fully digitized — those that require face-to-face guidance or private advice — have until now remained offline, locked within traditional training and consulting institutions.

Then, O2O took off like a spring breeze. Home services — manicures, massages, housekeeping — are “liberating artisans” at breakneck speed. This wave is also surging into knowledge and skills services. If on-demand O2O only frees service providers from salons and agencies, then knowledge/skill C2C + O2O does more: it liberates experts from conventional consultancies while simultaneously creating a brand‑new market.

For example, how do you find a LEGO master to teach your child creative building through Transformers? Or a “compassionate counselor” when love gets confusing? These services were previously unavailable.

What made this shift possible now? Why is the knowledge economy online only becoming feasible today?

The “Cognitive Surplus” Theory Is Losing Its Grip

“On the internet, nobody knows you’re a dog” was an early mantra, a rule of the virtual playground. Unburdened by real-world identities, people could create and share freely. There were no knowledge authorities or hierarchies. If Wikipedia volunteers had to flaunt their town-clerk jobs, how seriously would readers take their meticulously compiled “Archaeopteryx” entry?

As American internet researchers often put it, the early internet was like the “Wild West,” where everyone started with a blank slate. That equal starting point allowed those who were underutilized in the real world to find an outlet.

As a previous TMTpost article noted: “The most enthusiastic, active, and time-rich users contributed and maintained the most entries — not necessarily the most knowledgeable.”

Clay Shirky’s “cognitive surplus” theory — exemplified by Wikipedia — challenged the authority of Encyclopædia Britannica, but it failed to attract subject-matter experts in large numbers. The cognitive surplus represents knowledge and skills that cannot be “monetized” in real life. And the real world follows the rule spoken by the Joker in The Dark Knight:

If you're good at something, never do it for free.

Amateur bird enthusiasts on Wikipedia can’t find anyone willing to pay for their knowledge — hence the surplus. Real ornithologists, however, have people paying for their research.

Zhihu provides social incentives — followers, upvotes, recognition — but even that has limits. As I argued in a previous TMTpost article: Why do people share knowledge, experience, and insights for free? Because they have a key but haven’t found the lock that needs opening; because they can’t find paying users and settle for performing for applause only.

Professional knowledge wants to be taken seriously, but Zhihu’s mechanisms make every “knowledge seeker” feel entitled. When all knowledge is free, the distinction between quality and mediocrity disappears. That’s why Zhihu drifts further from its original one‑on‑one “question solving” ideal: answerers care more about applause from the crowd than solving the asker’s actual problem.

Information overload is a cliché, yet personalized knowledge services are still not easily accessible. Want to find a professional doctor online for a diagnosis? Need someone to strategize when your crush seems unreachable? Have a critical job interview tomorrow? You may find pages of medical records or hundreds of “how to chase a girl” answers; but it’s very hard to find the right person who can give you a just‑for‑you, tailored solution.

Cognitive surplus solved the challenge of aggregating public knowledge, but it cannot deliver personalized solutions or transform knowledge sharing into an economy. It fails to attract those who hold the key to turning knowledge into actionable help. Whether it’s medical advice, psychological counseling, or career coaching, the vast majority of such services remain offline.

So how do we bridge the gap from offline to online? Consider first why cars and homes became the pioneers of the sharing economy. Uber vehicles aren’t yet fully IoT-connected, but their real-time location, availability, model, and other metadata are digitized. That allows Uber to match each rider with the most suitable car. Airbnb follows the same logic: only after collecting data on each property’s location, size, style, availability, and host details can it help a traveler filter for a Bauhaus-style rental in Sapporo.

The logic is clear: for personalized knowledge services to be possible, we must first build dynamic “digital profiles” for experts — their identities, specialties, availability, and personal style. Only then can we tackle supply-demand matching, pricing, rating systems, and payments.

Then there’s trust. Only transparent credentials and verified expertise will make users willing to pay, rather than treating online answers as casually as search results or forum replies. When Google launched its paid video consultation product Helpout in November 2013, The Verge questioned whether it offered anything better than Google search results or YouTube tutorials. By February 2015, a once-ambitious Google shut it down. A Quora user echoed that skepticism: Google Helpout failed because people only pay for experts they already trust. Experts who had built trust offline could not prove their value online — and after-the-fact ratings were reversing cause and effect.

Let’s look at another cautionary tale.

Why Do Designers Scoff at Witkey Websites?

Witkey, a business model native to China, once boomed alongside “crowdsourcing” and “cognitive surplus.” It was romanticized as a dream lifestyle for freelancers — designers, copywriters, programmers. Yet most witkey platforms have withered; only Zhubajie (八戒) struggles on, heading toward an IPO.

I once registered a company and posted a task on Zhubajie: ¥500 for a company name not already registered with the government. I received about a dozen submissions — all generic, probably AI-generated. I had to check each name manually, and none worked. Someone contacted me, suggesting they’d refund part of the money if I chose them. So I logged in from another IP, submitted a few fake names, awarded the “contest” to myself, and got my money back. Never returned.

That anonymous Zhihu user’s story explains why designers now say: “Stay away from Zhubajie, cherish your career.”

Zhubajie uses a bid-and-grab model similar to Didi. But unlike Didi, the platform doesn’t track designers’ credentials, portfolios, styles, or price ranges — so a logo request can attract dozens of irrelevant bids, leaving the client sifting through garbage. For designers, the absence of professional profiles forces a race to the bottom on price. Low-quality bidders drive out the good, turning the platform into a training ground for students and a sweatshop for design laborers. Any designer who can find offline work scorns it.

Luo Yonghao built his “dream team” at Smartisan by scouting Chinese UI designers active on Dribbble. He simply looked at their portfolios to see if they fit his obsessive, skeuomorphic taste.

Perhaps realizing the flaws of the witkey model, Zhubajie launched “Tianpeng” in July 2015 — a designer-led showcase and matching service. It introduced tiered designer rankings (master, senior, junior) and promised “no contests, no bidding, intelligent matching, one‑on‑one service.” This is a self‑disruption of its former witkey approach.

No LinkedIn Means Startups Must Do “Ground Promotion”

Thus, the foundation of the online knowledge economy is digitized expert profiles. In a sense, LinkedIn provides that infrastructure. Platforms like Clarity, 24 Session, and Evisors — already on track abroad — rely heavily on LinkedIn’s open social résumés. Clarity, founded in 2011, has already facilitated over 100,000 conversations. But in China, lacking such infrastructure, startups like Linglu and 8pm are still struggling to take off.

That’s why domestic knowledge/skill trading startups have to do old‑fashioned “ground promotion” — verifying providers one by one. On‑demand O2O spread quickly by poaching workers from existing offline salons, massage parlors, and cleaning agencies, completing batch onboarding and verification. Ride-hailing apps did the same by starting with taxi drivers.

But compared to standardized services like massages or housekeeping, knowledge-intensive services (medical, legal, career) don’t digitize so smoothly. Even Chunyu Doctor (春雨医生) — recently plagued by closure rumors — faces serious obstacles.

Chunyu’s “light consultation” model depended on attracting famous doctors. Yet China’s healthcare system leaves top doctors overwhelmed with offline patients — a scarcity conferred by institutions. Chunyu cannot match their consulting fees, and patients tied to hospital brands won’t pay “independent doctors” online without institutional backing. Without user payment, doctors’ online value isn’t realized, and no business loop forms. Saturated doctors have no incentive to nurture an online market, just as hotels with 90% occupancy wouldn’t list on Airbnb. That’s why Chunyu announced plans to open 300 offline clinics by year’s end, bringing doctor resources in‑house.

“To do Internet+ today, you must have traditional industry resources,” says Kelly, founder of 8pm, a 1‑on‑1 experience sharing platform. Although her earlier ventures — “Consulting Sky” (a consulting forum) and CareerFrog (career coaching) — gave her 200,000 potential users, CareerFrog’s premium services cost ¥15,000–¥40,000 per program. 8pm’s “affordable” consultations are far more accessible.

“Where to find consultants” became her first hurdle. She still sends LinkedIn InMails every day, and also uses China-specific “ground promotion”: sharing H5 pages in company QQ groups, using peer referrals and cross‑endorsements (similar to LinkedIn endorsements, but rooted in China’s trust networks).

Zaih also didn’t start from zero. Just as Zhihu used founders’ connections in the tech circle to attract early adopters, Zaih leveraged Guokr’s influence and Ji Shisan’s network to recruit industry influencers, media stars, and Weibo celebrities — many already bringing their own followers, requiring little promotion.

Zhihu started as a semi‑real‑name community, but after opening registrations, it became largely anonymous. Hence I argued that Zhihu could never become China’s LinkedIn. Zaih faces a similar challenge: after its initial network is exhausted, how does it scale expert onboarding and verification quickly?

Zaih now contacts “hub figures” in cities beyond Beijing, Shanghai, and Guangzhou to reach local experts, supplemented by peer referrals — but it struggles to validate credentials. For example, no information about Zhuge Siyuan — a top earner on Zaih — can be independently verified online, aside from her profile’s “legendary” biography.

In my earlier article “The Almighty Zhihu: Why Didn’t Zaih Originate from Zhihu?”, I explained why Zhihu — despite hosting industry experts — didn’t spawn its own paid consultation service. Similarly, why wouldn’t Weibo, where influencers have millions of followers, match them with followers for paid consultations?

In the social media age, “unrecognized talent” is no excuse. Experts across industries can showcase their knowledge on Weibo or Zhihu, escaping traditional credentialing systems. They become sought-after columnists, paid trainers, and public relations targets. But they excel at writing popular science articles optimized for click‑through rates — standardized content for mass audiences, not custom solutions for individuals. And for influencers with huge followings, personalization is inefficient. As Ji Shisan lamented, “Big names with their own fan base bring users, but they don’t have time for one‑on‑one sessions. They’re happy to address hundreds or thousands, but not individuals.”

Zaih invited Chen Zhangyu, a top Zhihu reading answerer, to offer a course on “Personalized Reading Maps.” But even Ji admits that such “interest areas” are not Zaih’s core focus.

Still, Zaih helps these influencers by providing credential verification, tagging of expertise, and rating systems — making it easier for seekers to find the right match. Weibo and Zhihu’s UGC content is unstructured; evaluating someone means scrolling endlessly through their posts.

Even “Chatting” Has Become a Marketable Skill?

“Every investor told me: if you don’t go erotic, you’ll never reach tens of millions of users. One flatly said, ‘Your product will die without erotica.’ But I know the erotic road is a dead end. Men are disloyal; as soon as a new app emerges for hookups, they’ll jump ship.” Yuan Fei, founder of the pink-hued chat app O’ér (偶尔), refuses to follow the “hidden rules.” She wants to build not another beauty‑contest chat app, but a platform for emotional solace for reserved Chinese people.

How does a chat app avoid sliding into erotic services? O’ér’s answer: find a pool of “oxygen hosts” (the service providers) who are “kind, decent, and emotionally intelligent.”

O’ér didn’t start from ground zero. Last year’s media frenzy around “virtual lovers” on Taobao eventually got the category banned. Those displaced chat providers became targets for O’ér, Peiwo (陪我), and Moca (摩擦).

Yuan Fei, an Alibaba veteran, leveraged her connections to turn former Taobao shop owners into her managers, onboarding over 3,000 “oxygen hosts” before launch. Taobao banned “virtual lovers” because they easily veered into gray‑zone erotic services. Many copycats emerged, but Yuan Fei says: “They never asked why people seek virtual lovers. Loners may pay in the short term, but once they wake up, they’ll leave forever.”

Yet the thriving online porn industry proves people will pay for a striptease — a well‑worn commercial path. So what makes people pay for emotional listening? It requires oxygen hosts to demonstrate specialty beyond looks. Virtual lovers already differ from pure erotica by offering tags like “tsundere,” “loli,” “queen” — moving beyond physique. O’ér aims to enrich tags: interests, education, personality, special traits (healing, joke‑telling). Only then can users match with the right listener, preventing a race to the bottom. And only by surfacing deeper value can O’ér offer future vertical services like palm-reading, PUA coaching, or relationship recovery.

Yet O’ér still manages hosts via administrators (similar to YY’s guild model), without directly owning detailed personal info — not even guaranteed real profile photos. And its Tinder‑like card‑swiping design encourages face‑based selection.

Knowledge Transaction or Knowledge Social Networking?

“A college student building a part‑time job app booked meetings with several investors on 8pm — and eventually received funding from one of them.” Kelly asked me what I thought of such usage. As a consultant on 8pm herself, she once had a client who consulted just to land a job at her husband’s company.

A “mentor” on Zaih privately told TMTpost: “Zaih has become a place where investors source deals. Several angel investments I made started with founders asking for advice; we connected, found each other credible, and formed investment or partnership ties.”

That helps explain why experts are willing to sell their time cheaply. As Zhihu user Pan Xin observed:

Seeing those prices on Zaih — ¥300 to ¥500 — it’s hard to believe they’re real experts. Some have incredibly deep credentials; those prices insult them. Price is the most direct signal of expertise, yet everything is lumped together, making it impossible to distinguish levels.

Still, like startup people lurking around BAT campuses in their Uber cars, experts list at “floor prices” not for the money, but to filter for truly valuable connections.

How many daily unsolicited requests do experts receive on Zhihu, Weibo, or LinkedIn? How many WeChat friend requests? Price is an effective filter. And conversion rates on Zaih are below 10% (number of actual meetings vs. wish-to-meet requests). Experts can also require prospective clients to prove their worth — promising mentees, collaboration opportunities, or peer exchange. The power remains with the expert. (One Ali veteran on 8pm keeps his price at ¥1 despite Kelly’s advice to raise it.) It’s similar to high membership fees on dating sites — keeping low‑quality users out.

As securities analyst Xie Moyan said: “The main reason is I get to choose my students. For training on internet investing, if I’m dealing with a retail investor, I really should charge ¥5,000/hour, otherwise it’s a loss. But if the other party is an internet insider, the price is just symbolic.”

This form of connection breaks through traditional circles, industries, and hierarchies, enabling cross‑boundary relationships, interdisciplinary collisions, and inter‑level dialogues full of possibilities.

However, this raises a question. Zaih may well evolve into a members‑only LinkedIn — a high‑barrier professional social network, much as Zhihu has become a “playground for insiders.”

Is Zaih focused on knowledge commerce or business social networking? It currently wavers. If knowledge commerce is the goal, the sub‑10% conversion rate is extremely user‑unfriendly. It suggests the price lever isn’t balancing supply and demand, and that most mentors aren’t planning to act as professional consultants — they can’t satisfy the surge of users. To fix long‑term supply scarcity, Zaih would need to rely on full‑time trainers or career coaches, turning it into an “online training agency” — a successor rather than a disrupter.

If business social networking is the focus, China still lacks a serious LinkedIn. How many users would a “pro version” attract? (This article first appeared on TMTpost.)