Why do I feel that online education platforms are unreliable?

Youcan wrote in the front:

This article is an analysis and commentary on the online education industry that I wrote in 2014, two years ago, at the invitation of the education industry media "Duozhi.com". Although it was written a long time ago, all the views in the article are still completely applicable today and there is no difference. I believe that many of the viewpoints mentioned in the article have been verified today.

In 2014, I felt that the online education platform was a very unreliable model. Since then, I have discussed this with Chen Xiangdong, CEO of GSX, Ji Shisan, CEO of Guokr.com, and others. Unfortunately, GSX still fell into some of the pitfalls that I had predicted.

This article is being reposted today because in the Sanjieke classmates group, we once again heard the news that many online education platforms have begun to lay off employees. So I would like to repost this article, hoping to trigger some thinking among you about the industry and the "education" business itself.

This also reminds everyone: future NB products and operations must be able to "understand the industry and understand the business."

In 2012, a few friends and I plunged into the field of online education with passion and ideals, and set up a skill-based platform. At that time, it was the first of its kind in China, and we considered ourselves "pioneers."

Judging from the founders and core members of the team, we were originally a typical team that "knew a little about the Internet but knew almost nothing about education." After a year and a half, I stumbled along the way, encountered all kinds of troubles, and suffered bruises and bloodied heads. Due to ignorance, I also fell into many pits. I almost turned from a "pioneer" into a "martyr".

There are good reasons for choosing to build a platform - it has a lot of room for imagination, it looks sexy and tempting, and there are currently no similar mature platforms, so there is an opportunity.

From the perspective of the Internet, since we want to build a platform, we naturally have to align ourselves with the most successful platform in China currently - "Taobao". For this reason, we once shamelessly called ourselves one of the "Taobao for education" and went straight to the platform in terms of product form and operational ideas at the initial stage of the project.

But looking back, our understanding of the platform in the initial stage was probably the biggest pitfall.

There is a big problem and limitation in thinking here - how is a platform formed? In other words, if you look like a platform, does that mean you are really a platform?

For example, if there is a community focused on reading and learning, and through the act of "reading", it has gathered hundreds of thousands of users who like to learn and grow. At this time, it introduces teaching resources to complete the matching of teaching and learning parties. Is it considered a platform?

For example, Xunlei, Baofengyingyin, Hujiang, and Jingdong do not seem to be a platform (at least they were definitely not a platform at the beginning), but can they be considered a platform now?

In fact, looking back at the development path of successful domestic platforms, Taobao is almost the only example that went straight to the platform, and at the same time operated it vigorously for both buyers and sellers, and finally succeeded. Behind its success, there are all kinds of favorable conditions, favorable locations, favorable people and hard work. It is almost impossible to try to copy Taobao's successful path.

The essence of a platform is the recognition and dependence on the value of the platform between at least two parties that you want to connect and symmetrize.

What we generally know is that all parties connected by the platform are of value to each other on the platform. Buyers are of value to sellers, and good sellers are also of value to buyers. However, in the early stages of a platform’s growth and development, if you cannot find at least one additional value point for users outside of the framework of “mutual value among all parties on the platform,” you will likely be caught in the dilemma of “which came first, the chicken or the egg?”

This is exactly the case for us. After we started to look like a platform, the question that has been bothering us for a long time is whether to bring teachers in first, or users in first, or to do both at the same time?

Being in such a dilemma and wavering state, it is often difficult to take care of both ends and to form focus and synergy. It is like you are wielding a long sword vigorously and it looks mighty and majestic, but in fact the damage you can cause to the enemy is far less than a dagger stabbing directly into the heart.

If you ask me now, my answer would be: If you want to be a platform, looking like a platform from the beginning is probably not the best option.

Before we started building the platform, we painted a beautiful picture - first we would have a group of high-quality teachers whose courses are attractive in themselves, and then we would do some promotion, and users would quickly fill their classes in large numbers. The teachers would have a blast and make money here, so they would continue to teach classes and attract more high-quality teachers to join. The platform would form a positive feedback loop and continue to grow...

But unfortunately, the reality is very cruel, and this beautiful idea is also a pitfall. There are mainly the following problems:

First, a new platform does not have the ability to aggregate and distribute traffic in the early stages. However, for content partners and lecturers, the most rigid values ​​that they most desire to be satisfied are traffic and enrollment. It is not easy to find content providers that can truly provide stable and long-term cooperation. From our experience, at least in the first six months, out of every 100 institutions or lecturers, no more than 5 can actually stay and become important content providers for the platform.

At the same time, a point that can be extended here is that if the value that the platform provides to lecturers and institutions is merely tools , this value will be extremely non-rigid.

There are now so many online platforms and publishing tools that the cost for teachers to switch platforms is very low. Let me ask you, for a teacher, how big is the essential difference between teaching online through YY and teaching through SKYPE, Duobei, or QQ video? For offline courses, how big is the essential difference between publishing through Weibo and Douban and publishing through activity websites such as Huodongxing and Vasee?

Therefore, for a so-called platform, assuming that you have no traffic and godfathers behind you, why content providers should use this platform is a question that needs to be answered clearly.

Secondly, even if there are content providers and the platform has content and courses, the difficulty of course promotion and enrollment is beyond our imagination. Before we started building the platform, we fully believed that as long as we found some high-quality teachers, it would be natural and logical for students to sign up for their fame and high-quality content. However, the reality was far from that.

On the one hand, the field of skill-based learning is fragmented and subdivided enough, and even the best "famous teachers" can only be confined to a very small circle and range. What often happens is that you go to great lengths to find a top teacher in a niche field, only to discover that more than 95% of users don't even know who this person is, which makes everyone feel embarrassed.

On the other hand, the traditional education and training industry is accustomed to making money by relying on information asymmetry, and its marketing strategy is often to use free lectures to trigger impulse consumption. Either the course consultants follow up one-on-one and break them down one by one, or there are very few educational institutions that can deeply understand and apply marketing based on an open and interactive Internet environment. There are also almost no institutions or lecturers who can achieve their own brand image. In this situation, if institutions and lecturers simply publish their courses on the platform, the registration conversion rate they can attract will be very low. There are even some famous teachers whose course texts are so bad that people can't trust them at all, and the conversion rate is close to 0.

To this end, at the beginning of the platform’s launch, we had to help all the instructors and institutions on the platform to modify course copywriting, make course pictures, explore course selling points, repackage products, etc. In short, it's heavy.

In addition, there is another key factor - for ordinary users, skill learning is not a necessity. There are two kinds of things people need to learn. One is "what you know you don't know", and the other is "what you don't know you don't know". Unfortunately, most skills belong to the latter.

Without comparison, most people would not think that they lack communication, project management, time management and other abilities. Even if they realize that they have deficiencies in certain skills, more than 80% of users have not developed the learning habit of "improving skills by taking classes", not to mention that classes have to be paid.

Furthermore, in the domestic environment, more than 95% of users have passive rather than active learning habits. They often only have corresponding learning needs under external pressure, such as to obtain a certain certificate, a professional qualification, or mandatory requirements from companies and bosses. The learning needs that they actively generate on their own are basically very rare.

In order to make users generate skill-based learning needs, you need to educate them and provide them with some convincing scenarios, all of which are costs.

We finally made it through the first half year after going online. Although the base was not large, the platform began to generate some of its own natural traffic and had a certain user base. From the perspective of Internet people, our idea is that we have already taken this crucial step from 0 to 1, and things are finally going to fall into the rhythm we are familiar with - it seems that all we need to do next is copy and imitate the ways of playing of certain e-commerce platforms.

But, unfortunately, we were wrong again, this was also a pitfall.

For example, if you want to buy an iPhone 5s on Taobao, there are two stores, A and B, with similar ratings. Both stores sell standard products. Store A sells it for 5,000 yuan, while Store B sells it for 4,000 yuan. You will easily make the decision: buy the one from Store B.

But if you want to learn time management, and there are two lecturers, Teacher A and Teacher B, who have similar reviews, Teacher A's class costs 3,000 yuan, and Teacher B's class costs 2,000 yuan, can you judge that you should choose Teacher B's class based on this?

The answer is no.

The reason is very simple. "iphone5s" is a standard, but "time management" is not. The iphone5s sold in two different stores must be the same. Apart from the price, you only need to compare the services at most. However, "time management" taught by two different teachers may be two fundamentally different things.

Similarly, "Lenovo ideapad E430" is a standardized product, but "communication course" is not; "Wanglaoji Herbal Tea" is a standardized product, but "career planning" course is not.

The biggest difference between course products and other physical commodity products is that they are “non-standard”. If it is TOEFL, IELTS, CET-4 or CET-6, it may not be so bad, but in the field of skills, this problem will be more prominent - even the "skills" themselves may be extremely non-standard, let alone courses that teach how to learn a skill well.

Therefore, it is fine for e-commerce platforms to sell standardized goods, but there will be problems if they sell non-standardized skill courses. Non-standardization means high user selection costs, that user needs cannot be met in large quantities, and that it is probably impossible to define what a "good" product is. If the above problems cannot be solved, it means that the platform will find it difficult to serve a wider user group.

The key to the ultimate success of any platform is not what the platform looks like, but whether the platform can provide certain key evaluation criteria that are recognized by everyone, and an API that can be used by everyone to call this standard.

The key to the Qin Dynasty's success in creating a precedent and unifying the world was that the Qin people unified the writing system, currency, and weights and measures.

Caesar was able to establish the embryonic form of the Roman Empire because he established a mechanism for the transfer of power with "Roman citizenship" as the core, and opened "Roman citizenship" to Gauls, Germans and Greeks;

Intel can become the dominant player in the PC era because it has established the standard configuration of PCs and allows any brand to assemble and produce its own brand of PCs based on this standard;

The key to Microsoft 's ability to build a platform is that it created the standard Windows operating system and provided programming tools under the Windows platform, allowing developers to develop various software based on the Windows environment. The same is true for IOS and Android.

The reason why Taobao can build an e-commerce platform is because it has built a set of integrity evaluation system, allowing sellers to grow and upgrade based on this system, and buyers can make faster and better judgments based on this system.

In an ecosystem where information asymmetry is serious among all parties, if you want to smooth the process, you must unify the standards. The rule of the game regarding standards is always: the more people follow the standards, the greater the value of the standards.

In addition to a standardized integrity evaluation system similar to Taobao and other e-commerce platforms, a skill-based learning platform may also need to build a set of content (that is, courses) standards if it really wants to stimulate massive purchasing demand for certain courses. This will help content providers standardize products and ensure product quality, while helping users strengthen their understanding of products and stimulate demand.

Trying to create such a set of standards is naturally fraught with difficulties. It requires not only a deep understanding of education and curriculum, but also the ability to command the whole world to respond, so that everyone in the world is willing to cooperate with and use your standards. Therefore, in the field of education and training in the past, it was usually certain state-owned agencies and departments that played this role, and the way to achieve it was often through examinations, qualifications and certificates.

Of course, the rapid development of an industry may just start from the creation of certain standards, just as the wave of the PC era started from IBM opening up its own compatible computer standards. The popularity of studying abroad is also due to the existence of standards such as TOEFL and IELTS. If one day someone can really build a set of standards based on course evaluation, perhaps this field will develop rapidly and achieve breakthrough progress.

From this standpoint, this is both a pitfall and potentially a huge opportunity.

Since standardization is difficult to achieve, we can at least provide good services to a small number of existing users so that they can have a good experience. Since it is difficult to quickly expand in batches in the short term due to the lack of standards, we should take our time and keep every user we have. With the premise of "not being afraid of slowness", we can accumulate users little by little and drive qualitative change with quantitative change, right?

Sorry, we are naive again.

One year after the platform was launched, we found that our platform had never established a positive feedback loop in user operations . Although the traffic was slowly increasing and the learning and class experience on the platform was stabilizing, only about 20%-30% of the monthly traffic came from old users. This means that the increase in traffic is based on promotion, and the sedimentation and retention of old users have not really taken shape. This is far from our expectations and understanding. Even compared with some Taobao stores, this number seems too bad.

There is a fatal key here.

If you buy a mobile phone recharge card or a bag of snacks on Taobao, when your mobile phone bill runs out or the snacks are finished, you will definitely have the need to buy again, and if the place where you bought it last time gave you a good enough experience, it will most likely become your first choice.

But if you have taken a time management course on a certain learning platform and got a good learning experience and results, would you take this course again?

Fast-moving consumer goods and many other commodities can be consumed repeatedly. Even if commodities such as printers are difficult to be consumed repeatedly, they will provide you with consumables that can be consumed repeatedly. However, "courses" cannot. Courses and learning can often only be one-time consumer products.

This means that if you cannot give users a strong reason, no matter how good your product, service, and experience are, users will most likely only have one opportunity to consume from you. This is especially true when it comes to skills. Each skill is relatively independent. Once you have learned it, you have learned it. And once you have truly learned it, you will only become more and more proficient and adept at using it. You may even have very few opportunities to follow up with subsequent services.

Learning is such a special product. If users cannot learn it, they will leave with curses; if users learn it, they will praise you but will still leave.

The underlying logic behind this is: compared to other e-commerce companies, you may only have one chance to make money from users.

Of course, some people will say that the platform can continuously provide new high-quality courses and products to stimulate the other party's willingness to consume again. I guess this question actually brings us back to Pitfalls 2 and 3 - first, even if you can find high-quality courses and content providers, why should they come to your platform? Second, as a platform, how can you define and distinguish which courses are high-quality?

Choosing to build a platform as a startup team means that they can either secure a position early and gain an advantage, or they may unfortunately fall into a fire pit and go down a road of no return. Even if you are lucky enough to survive the previous pitfalls, you still have to face the dilemma of unclear answer when thinking about such an ultimate torture: How can you survive in the long run?

There are only three ways to survive. Either make money, or bet on the future now, or find a father.

Option one is to make money. The path of making money seems unrealistic for skill-based platforms in the short term. Due to the pitfalls mentioned above and various complex factors, there is probably no skill-based platform that can say that it has the ability to make money. Even for businesses that can make money, their sources of income often have little to do with the platform, such as the online school to Duobei.

Option two is to gain support and favor from the capital market and take a big chance. It is like what Masayoshi Son's 20 million meant to Jack Ma. It was because of the sufficient ammunition that Taobao was fearless. It used its 6-year free strategy to defeat one opponent after another, including eBay, and eventually became the dominant player. In this mode, what matters is courage, patience and boldness, and the prerequisite is sufficient food and grass and no worries in the rear.

Judging from the popularity of the online education circle and the pursuit of the capital market in the past period of time, the possibility of encountering another Masayoshi Son in the circle exists. However, even if you could meet another Masayoshi Son, you might not necessarily be Jack Ma.

Option three is to find a father. Finding a father is a feasible way, and it is also the most reliable choice in the current situation where the profit model and supporting system in the entire field are unclear. With a father, you can try and fail as much as you want. If you fail, your father can support you. If you do succeed, your father can also provide you with traffic and resources.

But the problem still exists. On the one hand, having a father is both a support and a big tree, but it may also be a constraint and self-binding; on the other hand, there are only a few fathers, and opportunities are always limited. Baidu has given a lecture, Alibaba has its own classmates, and the other giants also have potential moves. The world is booming, and there are not many opportunities and choices left. If you really want to find a father, please do it as soon as possible.

Wherever there is a pit, there will inevitably be people who will fall into it and never be able to climb out, and there will also inevitably be the beautiful hope that there is a flat road ahead of the pit. In the early stages of development of any industry, there are both challenges and opportunities. This is true for skill-based platforms and even for the currently extremely popular online education.

For the entire industry, making online teaching content more sticky and efficient, stimulating more people's learning needs, building a relatively standardized system for users, content and teaching processes, and exploring various potential business models are still the major focuses of common concern, and everyone still needs more time to come up with answers. (over)

Note: This article was written in 2014, so all opinions and facts in the article are based on 2014 and have not been changed. Please note.