Can Paid Content Buy You Real Knowledge? Can Online Education Fix That? | FreeS Fund Free Talk
What Should the New Intellectual Youth Study?
Two Questions About Online Education
In the first article of our education series, we used K12 education companies as examples to explore the prerequisites for healthy development of education institutions:
- Founders who can effectively balance idealism with commercialization;
- High-quality content supply that builds strong brands;
- Ability to spark students' interest in learning;
- Education services that cater to good students;
- Low ratio of marketing and sales expenses to total revenue.
These same standards apply to online education institutions. According to iResearch data, China's online education market reached 156.02 billion yuan in 2016, up 27.3% year-over-year; it is expected to maintain roughly 20% growth in subsequent years, reaching 269.26 billion yuan by 2019.
In this article, we consider and attempt to answer two questions:
▍ How will the content monetization and knowledge economy trend affect online education, and in what form?
Mobile internet technology has once again achieved a revolutionary leap in distribution efficiency. Whether through Dedao, Zhihu, or ByteDance, content is being disseminated in increasingly fragmented ways, with channel costs dropping dramatically.
Following this trajectory, after content becomes abundantly available and fragmented, users' demand for efficiency will gradually increase. That is, in a state of knowledge anxiety, users will tend toward the most efficient methods of acquiring information. Correspondingly, the content that survives will be relatively higher in quality.
Going forward, the scenarios for knowledge absorption may remain fragmented, but users will no longer be satisfied with fragmented knowledge. Instead, they will pursue more systematic, structured content. This process is analogous to moving from scattered articles to collected magazines and books, and then to educational courses with accompanying instruction. We are currently at the stage where "magazines" are everywhere, and can look forward to "books" and "education" to follow.
This means that content monetization and the knowledge economy will have a major branch landing on "scaled online education."
▲ People are gradually growing accustomed to paying for content.
▍ Which domains can generate scalable demand for online education?
Considering that the two listed education companies, New Oriental and TAL Education Group, respectively specialize in English and science education for primary and secondary students, we are inclined to believe that these two domains have sufficient market space and scalable demand. Good online education institutions can emerge from them, and we have invested in Onion Math and Shanbay.
Both companies built their user bases primarily through organic growth driven by strong word of mouth (Onion Math grew from 0 to 10 million users, Shanbay from 0 to 50 million), with customer acquisition costs far below industry averages. On this foundation, in Q2 last year, they both began experimenting with charging users, and preliminarily validated the feasibility of paid online education. Shanbay has already become profitable; Onion Math's revenue is growing rapidly.
Looking back at the founding processes of these two companies, we can see from their founders some common traits in balancing educational idealism with the pace of commercialization.
Among Onion Math's three co-founders, Yang Linfeng and Zhu Ruochen graduated from Harvard University and Duke University respectively. In 2011, they co-founded Sunshine Bookhouse, a nonprofit organization dedicated to promoting rural education informatization.
There, they observed that the gap between urban and rural education lay not in teaching hardware or facilities, but in learning experience, and identified two key factors causing this difference in learning experience: teaching methods and teacher competency. In the short term, neither problem could be solved through traditional means; only digital methods via online education offered a way out.
In fact, urban education also suffers from poor learning experiences. But building standardized online education products requires enormous human and resource investment, which was beyond what an NGO could accomplish alone. Thus, at the end of 2013, they decided to commercialize. They piloted in several key high schools in Beijing as well as some rural schools, conducting practice from both ends simultaneously to validate the applicability of this standardized learning product.
The process was not without difficult moments. For example, the difficulty of producing standardized online courses exceeded their imagination, and production cycles were long. In the first year or so of commercial development, they were troubled by course supply failing to keep up with demand, production processes being difficult to standardize, and the pace of commercialization being too slow.
Shanbay founder Wang Jie completed his angel round in 2011. When he wanted to take the next step, he found that market confidence in online English learning was insufficient. He decided to pause his next fundraising plan, instead achieving self-sufficiency by charging his community users, with monthly revenue gradually accumulating from 2,000 yuan to over 100,000 yuan. This was nearly impossible in the PC era, but he pulled it off with considerable grit, also building strong relationships with users — many of whose lessons continue to this day.
We invited Tianfeng Securities education industry analyst Zhang Lufang, along with Yang Linfeng and Wang Jie, to discuss how they view the relationship between online and traditional education, profitability prospects, and the opportunities brought by the content monetization wave.
The Best Form of Education Is
You on One End, Socrates on the Other
Lufang: In the early days of online education, people focused on the online-offline debate, wondering whether online education could absorb the offline piece through a process of elimination. If offline traditional education is more of a service, then is online education essentially a service or an internet product?
Linfeng: The distinction isn't that clear-cut. Online education follows neither product logic nor service logic, but rather course logic. Online or offline, user consumption habits differ, but the consumption process contains these elements: where, what, how long, how much gained, and how much paid.
The greatest value of following course logic is making learning effective. Closing the value loop requires several steps:
- Attract students. Get them willing to invest time and attention, converting sustained attention into learning outcomes.
- Build a solid product. From content to interaction, everything centers on learning itself, justifying the time students invest.
- Perception level. Students have good learning experiences during the process, seeing results through improved exam scores; parents discover their children are willing to learn actively and continuously, with improved grades, and thus become willing to pay.
The standardized online learning experience that Onion Math has built is not entirely designed around user consumption habits, but rather incorporates consideration of user consumption habits on the foundation of course logic. For example, although we position ourselves as a purely online learning experience, considering that some parents haven't fully adapted to purely online formats, we also produce books to supplement and enhance the online learning experience.
We are trying to overlay service characteristics on top of standardized products, but not human-based services — rather using human-machine interaction, providing services through machines. The online education that Onion Math is exploring will not be a heavy "human" model.
▲ Online education is not a digital replica of traditional classrooms.
Lufang: I came across an online education company in Taiwan whose core business is selling hard drives, though they also provide accounts for an online learning platform. According to their research, only 36% of students using their products continue to attend offline training institutions. What's the situation on our side? How do you understand the relationship between online education and offline classrooms?
Wang Jie: Online education is definitely not an online version of offline classrooms.
I believe the most important mission of online education is to truly achieve personalized, tailored instruction for every individual. I don't remember where I read this: the best form of education is to place a table, with you on one end and Socrates on the other. If you had a teacher as brilliant as Socrates willing to tutor you one-on-one, you'd amount to something no matter what.
But the problem is, where in the world do you find that many good teachers?
The current offline classroom model dates back to the industrial age. It was actually designed to achieve mass education普及, making compromises on cost structure, with the compromise being that teachers cannot attend to each student's pace and individual characteristics.
Online education holds the promise of dramatically improving efficiency on both the teaching and learning ends through technological means. On one hand, the content teachers deliver can be infinitely replicated at extremely low cost; on the other hand, students' learning paths can be customized at extremely low cost. In certain relatively standardized domains, AI may even replace teachers to a considerable degree.
Combined, these factors will enable online education to create forms entirely different from offline classrooms, operating on cost structures that traditional institutions would find inconceivable.
Linfeng: We haven't deliberately defined online education as either a replacement for or supplement to offline education. Actually, whether students enroll in offline tutoring classes has nothing to do with us. Enrolling in classes is sometimes an irrational choice, influenced by the psychology of "others are doing this, so I must too." In the current exam-oriented environment, parents' anxiety and consumption habits cannot be suddenly dispelled by a few online education products from a handful of companies.
▲ Online education is born of digital devices, emphasizing human-machine interaction.
The essence of online education is that people learn through digital interactive experiences, not by facing another person. It is a learning method based on digital-native platforms and internet distribution channels, where both the content medium and interaction methods have changed. Correspondingly, product design and learning scenarios are very different from offline education.
The reason current online education products still don't seem very distinct from offline education is partly that people are relatively familiar with offline education models. Transferring offline experience directly is the fastest, lowest-cost approach. But as time goes on, the distinction between the two will certainly grow larger.
Product Pricing
Depends on Whether the Service Is Systematic or Fragmented
Lufang: How do online education startups build their brands and accumulate users in the early stages? How do you keep marketing and sales expenses as a percentage of total revenue at low levels?
Linfeng: Currently, purely organic growth accounts for over 70% of our total user growth, attributable mainly to two factors: offline word of mouth and online app stores.
First is word-of-mouth spread by students and teachers. In early 2014, our product was still in demo version. Considering that B-side schools were where student users were most concentrated, we chose several key high schools including Beijing No. 8 High School and the High School Affiliated to Beijing Normal University to validate from the B side. After teachers tried it and found it good, they recommended it to grade leaders, and gradually it spread throughout the grade.
At that time, there were very few products combining learning and practice for students' self-directed learning, so teachers were very willing to try. Students also felt that animated teaching provided a very different learning experience, and spontaneously began spreading the word.
This process continued through the second half of 2014, accumulating substantial feedback and good word of mouth for us. It was also with these key high schools as endorsements that subsequent promotion at other schools became smoother.
At this stage, we didn't deliberately pursue user growth, because a very important issue troubled us: incomplete course coverage. We felt we should at least finish all math courses for grades 7 and 8 before pushing hard on promotion. From September 2014 when the first version officially launched on web for user registration, through autumn 2015, we never did any marketing.
This period also faced skepticism. However, because our team had a nonprofit background, everyone shared relatively consistent understanding of core educational philosophy, all pursuing the polishing of products that could provide quality learning experiences, so there wasn't much wavering.
In autumn 2015, Apple Store featured our product on its China homepage for the first time, which played a significant role in building reputation and driving traffic. Over the year from then to September 2016, user numbers grew from 230,000 to 4.5 million, and now we have 10 million users. We've been featured on the Apple Store China homepage 10 times in a row, and subsequently Android app stores have basically followed suit with recommendations. For example, Xiaomi's app store even gave us its 2016 annual Jinmi Award.
Looking back, we were building reputation both offline and online, but offline speed certainly couldn't match the online app store channel. The vast majority of online education products are drill-and-practice models, while we focus on learning — the point that "learning math is fun," which is a bit "unexpectedly charming." First-time users would think, are you trying to trick me? I might as well try it myself. After using it, they'd find the experience genuinely good, and would spread it spontaneously. Early users were basically accumulated in this way.
▲ A good product won't make students feel like they're being forced to learn.
Lufang: How do online products spark students' interest in learning, thereby ensuring good learning outcomes? On this foundation, how do you pace the development of new products and expansion into new subjects?
Wang Jie: I believe the most important thing for sparking interest in learning is being able to quickly give students positive feedback. Why can't many people persist in learning? Because traditionally, learning is a process of accumulating much before achieving breakthroughs, but most people don't have the patience to accumulate, and give up before the breakthrough.
Our method is to use technological means to shorten the distance between learning and application as much as possible, to some degree achieving results from modest accumulation. For example, after users learn vocabulary on Shanbay Words, in the Shanbay Reading app they can read articles at their corresponding vocabulary level, and the app immediately highlights vocabulary they've previously learned, letting users feel their effort wasn't wasted.
Linfeng: Compared with offline tutoring classes and online live-streaming instruction, our model has a longer cycle from initial R&D to learning experience. Final exam scores are the most direct validation standard.
In this closed loop, our control logic is to first ensure learning experience, getting users willing to invest time and gain something from it. On this foundation, we package some products as goods to recommend for purchase. This is a continuously spiraling upward process. When some commercial monetization capability is validated, revenue shouldn't be used solely to expand sales, but should be invested into product R&D, while also considering more user consumption habits in the market.
On expanding subjects, we also don't want quality of learning experience to suffer from moving too fast. The essence of education is brand building, and once a brand gets negative reviews, they cannot be undone.
For example, if TAL Education Group announced today that it was entering some education domain, expectations would be very high. Even if it wasn't a first mover in this new domain, its previous reputation would still win considerable attention. That's the value of brand.
We hope to cover all stages and subjects of primary and secondary school, but aren't eager for rapid occupation. We want to gradually accelerate on the premise that learning experience quality is guaranteed. Now that we already have a large base of middle school math users who have established brand trust with us, when we expand subjects, we believe they can naturally migrate over.
The commercialization of education is essentially the high unification of a product's educational value and commercial value. Only by first shaping product value can moderate market operations and business model exploration follow. This industry cannot only emphasize speed; it must maintain strong interest in the value of the endeavor itself, and always believe that it can generate social value.
▲ Education is a slow project; there's no need to overly pursue expansion speed.
Lufang: After experiencing the野蛮生长 of burning cash and subsidies, the bottleneck of profitability and monetization for online education has gradually become apparent. From your explorations, how do you make monetization channels work? How do you balance free and paid? That same Taiwanese company — their product is priced at 100,000 Taiwan dollars, and last year they started generating 760 million Taiwan dollars in revenue. From this perspective, are users' willingness to pay underestimated?
Linfeng: I actually feel that whether online education can be profitable was perhaps still somewhat questioned in 2015 and 2016, but by this year people generally no longer dwell on it. Good online education products certainly can, and must, charge fees, with relatively high expectations for charging.
Currently, our charging logic is basic knowledge courses for free, ensuring users can smoothly use the product to learn and enjoy good learning experience, without getting stuck at some point. More targeted specialized exercises, explanations, exam focus analysis and summaries, and personalized study plan intelligent recommendations are paid items. Overall, we follow a freemium business logic.
Our users are dual: students use our product, while parents pay. Parents naturally have good payment habits for quality learning products, so for a product like ours that combines learning experience with score improvement effects, payment isn't hard for parents to understand.
Price is determined by product form, depending on whether the product provides fragmented services or systematic services. For example, early childhood education products consist of materials combined by topic, with no required learning sequence — this suits fragmented sales more, with lower course pricing. K12 requires systematic, sustained learning; courses need 3-6 months of study for obvious results, so naturally pricing can be higher.
Content and Education Are Two Ends of the Same Spectrum
Lufang: Live-streaming has been the focus in recent years. Some question-bank products and O2O platforms have transformed into live-streaming platforms, launching "internet celebrity" star teachers to attract traffic. Meanwhile, people are also emphasizing AI development and assistance. How big is the edtech pie? When can we eat it?
Linfeng: In the short term, new technologies have a certain pulling effect on online education. But overall, they've only sparked some learning interest from the margins, or created some marketing buzz in the market. They remain at the supporting role stage; core learning experience hasn't been greatly upgraded.
Live-streaming has never been the problem. The problem is that learning methods combining live-streaming can't stop at moving teachers online. AI also can't stop at using big data analysis for drill-and-practice — this only improves practice efficiency and is a shallow application.
Currently, the obstacle to technology pushing education forward doesn't lie in AI algorithms themselves, but in people's still relatively shallow understanding of learning experience, preventing these technologies from being effectively applied. We're also doing related research now. The process from a student completely not understanding a knowledge point to completely mastering it contains much valid data worth mining, which can be used to study how to improve learning experience.
Lufang: How do you understand "content monetization will have a branch landing on scaled online education"? How far are we from this step?
Wang Jie: In my eyes, content and education are two ends of the same spectrum. For example, a book is content; but if you add teaching, testing, supervision, Q&A and other links, it becomes education. Each of these links can be executed by machines or people, achieving different effects and cost structures.
▲ After information fragmentation, people will increasingly value efficiency in acquiring knowledge.
I believe people on both ends will move toward the middle: content creators will increasingly add service leverage to achieve differentiation and premium pricing; education providers will increasingly lighten certain parts of their operations, partially replacing people with programs to improve efficiency.
The most important foundation for content monetization is actually that WeChat accomplished three things in one go: providing a unified account system for mobile internet; providing very convenient distribution and publishing channels; and providing very quick payment methods.
WeChat accomplished these three things, greatly reducing costs for content products previously scattered across the internet at customer acquisition, conversion, and payment stages. Therefore, when the cost per user of content products falls below users' psychological price expectations, combined with content products' zero marginal cost, this new consumption market begins to explode. The content monetization trend will certainly have massive impact.
Linfeng: User demand for fragmented knowledge comes in waves and isn't in a sustainable state. During a certain period they want to acquire knowledge in some area, quickly "consume" it, then get "full," discover the actual收获 falls short of expectations, and enter a low period. After a while, when another form emerges, they quickly "consume" again.
What Onion Math does is K12 stage courses, which are inherently a systematic learning process. Moreover, K12 stage students aren't the mainstream consumer group in the content monetization wave, and aren't directly pulled by this trend. Student parents may be somewhat influenced — their recognition of online purchase of learning products is gradually increasing. But currently most middle school student parents are born in the 1970s, and their consumption habits haven't fully become internet-ified.
However, the content monetization trend is definitely favorable for us in the long term, though the dividend has a gradual release process, such as when the younger post-80s and post-90s parent groups gradually grow larger.
(Welcome to share to Moments. For reprinting to other official accounts, websites, or mobile apps, please reply "reprint" to understand reprint rules, and contact "FreeS Fund" for authorization. Copyright belongs to FreeS Fund.)
