Episode 61: Getting Educated on China K-12 Tutoring Edtech
Transcript:
(Y: Ying-Ying Lu; R: Rui Ma )
[00:00] R: Hey everyone, I hope you enjoyed our last Tech Buzz episode and last two Extra Buzz newsletters on how the COVID-19 virus is hitting China tech. As the country is only very slowly getting back on its feet, there isn’t a whole lot of news that’s not virus related, and so we are going to talk about it too, but today with a specific emphasis on edtech.
Y: But before we do that, let’s give us an update for those Tech Buzzers who might not have been as immersed as we have been in epidemic-related news. As of recording, there are now over 2,000 deaths attributed to the coronavirus and over 70,000 confirmed cases, all in China alone. Experts seem divided on whether or not we’ve seen the worst, and for many businesses, a full recovery seems pretty far away.
R: Now, tech and knowledge workers who can work effectively remotely seem to be doing alright, as listeners in these industries have reached out and told us, but businesses requiring physical labor, especially factories, are still in a state of chaos, as expected. In a survey of small and medium sized businesses conducted last Friday, a good 40% are still not operating and estimate that they are at least two weeks away from doing so.
[2:48] Y: Additionally, less than 10% are optimistic that they can achieve profitability this year, which is a problem indeed for those who aren’t surviving on venture capital. Some of the delay is because special permits need to be obtained from the local government in order to resume work, along with a bunch of epidemic-specific paperwork that must be filled out, some of them daily, for every employee.
R: Extremely tedious to administer, but which, as an unintended consequence, might boost the e-signature and document management business, as we noted in our Extra Buzz newsletter. Other delays are because realistically, a lot of workers went home over the holidays and each region seems to have different quarantine rules that are triggered when they return. Add to that the cost of virus prevention supplies, costs of quarantine for all staff should even one be found infected, and you can see why business owners are stuck between a rock and a hard place.
Y: Luckily, the group we’re focusing on today — K12 edtech entrepreneurs, are the lucky ones. You can probably guess that their businesses are growing, not shrinking, because of the epidemic. But what were the key trends and who were the key players before the virus hit? And how are they responding now? As always, we’ll give you all the context we can, so you can understand the conditions driving this very unique market.
[4:42] R: Hi everyone! We are TechBuzz China by Pandaily, powered by the Sinica Podcast Network by SupChina!
Y: We are a biweekly podcast focused on giving you a peek into what’s buzzing within the tech community in China.
R: We uncover and contextualize unique insights, perspectives and takeaways on headline tech news that don’t always make it into English language coverage. So you can be smarter about the world of China tech. TechBuzz China is a part of Pandaily.com, an English language site that tells you “everything about China’s innovation.” I’m one of your two co-hosts, Rui Ma.
Y: And I’m your other co-host, Ying-Ying Lu. We’d like to acknowledge our partners DealStreetAsia and SupChina, creator of the Sinica Podcast Network! In addition to Tech Buzz, you can also find Sinica which covers current affairs, NuVoices and Ta for Ta on women, the business-oriented ChinaEconTalk, and the Caixin-Sinica Business Brief from China’s leading business magazine.
R: Oh and an announcement! We just launched techbuzzchina.com, where you can sign up for our new Extra Buzz newsletter. Check out our first two issues, and subscribe for more insights and content on China tech for $2 a month! Finally, as always, if you enjoyed listening to our podcast, please leave us a review on iTunes or whatever other platform you use.
[6:11] R: So Ying-ying, we’ve only devoted about an episode and a half in all of Tech Buzz so far to edtech — there was Episode 47 on VIPKid, which was specifically about English learning for the younger crowd, and Episode 53 on gaming giant Netease, which talked a bit about Youdao, a recently listed Netease spinoff that’s got about 60% of its revenues coming from education.
Y: Well, the lack of episodes is certainly not because there’s nothing to talk about. If we disregard the technology component and look at education broadly as an industry, it has provided an immense number of investment opportunities in recent years. I mean, in the last 3 years alone, at least 25 Chinese education companies have gone public, although most of them on the Hong Kong Stock Exchange and just a small handful here in the U.S.
R: And even if we only talk about edtech, China is a force to be reckoned with. In fact, in 2018, the headline raging in edtech circles was that 50% of the venture capital that year deployed in edtech globally went to Chinese companies, more than twice as much as second-place United States. And that trend has mostly persisted, with the US pouring $1.7Bn into edtech last year, and China nearly $3Bn.
Y: That’s impressive given that China had a 60% decline in venture capital last year, but the edtech sector held steady and bucked the trend. And now with the epidemic, K12 edtech is surging, because it’s become mandatory to move all learning online. Whole cities are still under quarantine, and the risk of infection means that no schools are open until March at least, or possibly later, maybe even May.
[8:04] R: That’s a long time to be out of the classroom, right? No wonder everyone is kind of freaking out. But should they? Didn’t we just say that China had the most edtech investments in the world? And CBInsights lists 8 Chinese edtech unicorns. But wait, there is a fundamental difference between the type of services that most Chinese K12 edtech companies were offering, and actual schooling, which is what’s needed right now.
Y: Right. Most companies were in the business of offering K12 tutoring, and not instruction. That’s because education in China is controlled by the government, and the rules are numerous. I mean, it’s a public good so compulsory education is heavily regulated by governments everywhere, so China is not alone in this by any means, but some of the rules are more restrictive than here in the US, for example.
R; We aren’t education experts so we’re just picking a few of the rules and changes that have stood out to us because we’re guessing you might also not be aware of them. And we think they’re good examples of how different the system is. So here goes the first fact, which is that the government recently banned the operation of for-profit schools for the years of compulsory education, which in China, is from grades one through nine.
Y: Now, that doesn’t include international schools, so don’t worry, your expat friends in China aren’t breaking the law, and the exact definition is that the school has to be distributing profits to qualify as one, so you can make money and not distribute the profits, and be a “non-profit,” but the intention is pretty clear — the government doesn’t want to make it easy for you to just create your own school and to make loads of cash doing so.
[9:57] R: Given that public schools in China are actually pretty good, at least on the top end, and private schools, until maybe ten, fifteen years ago, used to be mostly known as places where you sent your kids that simply couldn’t cut it in public school, this doesn’t sound as weird to some Chinese people as it does to us. Add to that there were instances of shady schools faking their qualifications and charging loads of money, and you can see that there are indeed some elements of consumer protection intended.
Y: And while that hasn’t stopped private schools from operating in China — even becoming listed in Hong Kong — that doesn’t mean there isn’t disagreement, even amongst experts, about the limitations they must operate within. Basically, in China, for at least K to 9, the least risky is still working with the compulsory education system, which is exactly what the companies we talk about today do — they’re not seeking to replace instruction, but provide tutoring.
R: And just how big is the after school tutoring business in China? Well, it is $64Bn, according to Frost & Sullivan. And of that, not even 10% was conducted online, or about $4.4Bn. Not a ton of revenue, but pretty good for a market that was pretty much nonexistent a few years ago, and also a lot less competitive than adult education, which is already around a quarter online. And most importantly, yor u gotta look at the projected growth rates. The 2018-2023 CAGR for K12 online tutoring? 65%. Do you know what that means?
Y: It means that in 3 years, the market size is expected to grow by more than 10 times to $53Bn. Wow, even for China, this kind of growth is pretty rare to see. I would take those numbers with a large lump of salt though, since the whole after-school tutoring segment is only growing at 15 percent a year now… so even if it doesn’t slow down, that means we’re looking at over 40 percent of tutoring taking place online in 3 years. Not the craziest estimate, but still quite aggressive.
[12:25] R: Maybe though, it’s not that crazy, because when you guys are thinking of K12 curricula, you’re probably just thinking of math and language and history and science, in other words, the usual subjects. But in China, in yet another example we have of how much policy influences and can potentially expand this space, students are also incentivized to learn things like one of my personal favorite pastimes — Chinese calligraphy.
Y: The Ministry of Education had always said cultural education was important, but in 2017 suddenly announced that it was even more important than language, math and English, which led to some universities establishing a calligraphy major, and some provinces hinting this kind of knowledge might be tested on future college entrance exams. Needless to say, that immediately attracted investment into the space, and a bunch of such providers received venture capital funding in the last two years.
R: So that’s both a risk and opportunity for such providers. The overwhelming majority of Chinese students — nearly 10mm of them every year — still test for the gaokao, or the notorious college entrance exam, and it’s still a pretty hard thing to “pass.” about 80% make it to some sort of tertiary education, but only half of them, that’s about 40%, make it into an actual 4-year bachelor’s program, and this is already a record high in China after many years of overhauling the education system. It used to be much lower.
[14:04] Y: And because whether you go to college or not, and where you end up going, can make a big difference in a young person’s life, the 2 or 3-day, 4-session, total-9-hour exam testing you on Chinese, math, science or literature, depending on your choice of study, and foreign language, basically ends up being what the kid prepares for certainly the entire last year of high school, but also pretty much throughout all the years they’re in school as well.
R: That’s why even in these life-or-death times of the coronavirus, people are stressing out about how to keep up their children’s studies. I mean, unless you’re well off and planning to send your kids abroad, most Chinese families think the gaokao is the most important day of their children’s lives. It’s when all those years of study and paid tutoring pay off… or don’t.
Y: And if SARS is any example, then even an epidemic will not delay the test, and if anything, high school seniors might end up the first ones back to the classroom, because the gaokao is just that powerful of a motivating force. As this one student explained of his SARS gaokao experience, our teacher told us that we cannot use SARS as an excuse to do poorly — when we graduate we won’t be able to say: oh but I did the gaokao in the year of SARS. No one will care! We’ll still be judged by where we went to school.
R: Well, the teacher is incentivized to push the students, because they can get bonuses depending on how well their students perform. Really, it’s very common, and it actually goes for all standardized testing. Like one teacher received about $43,000 in cash bonuses for teaching a student who tested top 2 in the high school entrance exam. That’s on the high end, but it happens with regularity.
[15:59] Y: So what does all this have to do with edtech? Well, our point, which I hope we’ve hit home, is that Chinese people take K12 education super seriously and since most people still go through the public school system, most of the spending ends up taking place on tutoring. Quoting from previous research we did for VIPKid here, but almost all Chinese parents spend at least $1500 a year on this, and over a quarter over $15,000 a year, compared to the US, where 80% are spending less than $2000 per year.
R: So let’s take a deeper look at two of the companies, especially the ones that have been in the headlines in the last year, or have been highlighted by the press recently because of what they’re doing to help combat the virus. We’ll skip over industry giant TAL, which is about an $8Bn market cap company right now, and New Oriental’s spinoff, Koolearn, which is listed in Hong Kong and edging close to $4Bn, because TAL and New Oriental still have a large offline presence and we want to focus on those players that began with an internet DNA.
Y: We’ll pick one public company, and one private, recognizing that depending on how you segment the space, these are not the largest players necessarily, but are two of the top. So the first one is GSX, or 跟谁学, which went public last June on the NYSE. It made huge waves in Chinese media because it is one of the very few pureplay edtech companies that made it to not just the public markets, but to profitability, after just one round of funding.
R: Well, technically two rounds, a few million in seed funding from QF Capital and a $50mm Series A a few months later. So it’s not like they raised no money. But still, it is in stark contrast to some of the other edtech companies in the space, like VIPKid that’s raised over $1Bn, or the other company we want to talk about, Yuanfudao, which has raised over half a billion.
[18:16] Y: So in comparison to those, yes it’s been pretty capital efficient. It’s more than quadrupled since IPO and is now a $10Bn company. And most of those gains were actually made prior to the virus outbreak. Because the company is growing quickly, even before the epidemic. Management revealed on its earnings call just this Wednesday that revenue in 201 9 was 5-times what it was the year before, although that still only takes it to a topline of $300mm.
R: And it also reported a non-GAAP net margin of 14% for Q4 2019, which looks pretty good, regardless if you compare it to oldschool competitors like TAL or newer companies that are straight-up lossmaking. Although we do need to remember that winters tend to lead to seasonal highs in terms of profit for this industry, since summers are the most expensive in terms of sales & marketing.
Y: And boy has it been an expensive summer for the industry! It was rumored that last summer leading players were spending more than one million USD a day on advertising for customer acquisition. Why, you ask? Isn’t that the only time when kids don’t have to study? Wrong! Chinese schools have mandatory summer vacation homework.Yup, you get a stack of booklets and handouts to complete over the break.
R: Winter break too! But summer break is the longer and thus more painful one. So you have homework, but no teachers … obviously, tutoring would be really, really helpful here. So let’s briefly explain the types of online tutoring that’s commonly used these days. The first one is one-on-one, like how VIPKid started. Then there’s small class, which is often just a handful of students, from two to as many as thirty, but typically eight or less.
[20:16] Y: After small class, it’s … you guessed it, large class. For GSX, which operates primarily under this model, the large classes have grown in size from 400 a few years ago to 1700 now. But of course it’s not just one teacher lecturing 1,000 students. GSX, like Youdao and others, go for a dual-teacher model, by which they mean that there is one primary instructor who is lecturing, paired with multiple tutors who oversee a smaller group of students and make sure they’re following along.
R: And just in case there is any confusion, all of these classes are livestreamed, so they’re not prerecorded like MOOCs here in the West. That’s also a business model, but not the one we’re focused on today. As you might have guessed, GSX’s model heavily relies on star or celebrity teachers, who can command the wallets of thousands of parents. In fact, their top 10 teachers contribute 46% of their revenues.
Y: Depending on your viewpoint, that’s either a scary amount of revenue concentration or proof of their core thesis, which is that great teachers are hard to come by, and the internet is the only way to scale this very scarce resource. This is especially true of less developed areas in China, where great teachers are rare. So I guess if just for the social good this could deliver, we should wish GSX luck.
R: GSX’s prospectus doesn’t provide demographic breakdown of its users other than that over three-quarters of revenue come from K12 courses, so there’s no way to ascertain how many students are enrolling from third-tier and below cities, but at the price of about $125 per class, it’s pricey but still affordable for a good many families in China, who, if you’ll remember, are spending thousands of dollars in after school tutoring for their kids.
[22:12] Y: By the way, this wasn’t always obvious to GSX. The founder and CEO, 49-year-old Larry Chen 陈向东, came from New Oriental, having worked there for fifteen years and rising through the ranks to executive director before quitting in 2014 to start GSX. His solid industry background is what got him the extra-large $50mm Series A, no doubt, because in the beginning, GSX, whose Chinese name translates to “Who to learn from?” was a marketplace where any teacher could sign up to offer whatever course they wanted.
R: Its Series A announcement back in 2015 boasted that it had over 70,000 teachers and several million students, and hundreds of teachers onboarding every day. Contrast this with what’s disclosed in their prospectus last June — 169 instructors and 522 tutors, and you can see what happened there. They made a major pivot in 2018, going from 22% K12 education to 73%. Obviously, the marketplace model didn’t work out as they expected.
Y: Not super surprising. One of the key difficulties in a marketplace model is that it’s just so hard to quality control. Sure, everyone can teach, but most people won’t be great teachers. Having a ton of selection alone doesn’t really mean much. Especially when acquisition costs for online learning are multiples of what they are offline, this could be a really really tough business model.
R: Right, a study by QuestMobile done in the first quarter of last year showed that traditional service providers like TAL or New Oriental, with both offline and online offerings, were acquiring new customers at about $30-40, versus pure-play online edtech companies, that were seeing $100 to $700 dollars spent to acquire new customers. Of course, those courses might be higher dollar values too, but generally speaking, acquisition costs for Chinese internet companies have been soaring for a while now, so this comes as no surprise to anyone, really.
[24:25] Y: It’s definitely no surprise to Yuanfudao, another edtech unicorn that’s also kind of stumbled upon the same route after pivoting a few times. Founded in 2012 and originally named Yuantiku, 猿题库, which means Ape Test Bank, the company was exactly that, an app that focused on the gaokao and helped you prepare by showing you questions from past exams and predicting your performance based on your answers.
R: OK, it did a little more than that. It also allowed you to take photos of your homework and bring up an instructional video that will help you solve the problem you’re stuck on. It apparently even corrected your English homework with AI. It eventually covered all of the middle school and high school grades and subjects and naturally expanded into one-on-one tutoring. But sometime in 2016, four years after founding, it pivoted from a C2C marketplace to a B2C or branded service provider where they vetted and hired the teachers and created the curriculum. Pretty much what GSX did, right?
Y: Yup. Stepping back a moment in time here, a lot of the companies that got funded heavily during those years were one-on-one services, and since we did a full episode on VIPKid, that’s who we’ll keep on bringing up as the example. But a few years in, most providers found that it was really difficult to build up a reputation using that model. Again, more students generally meant higher profits, and because there’s just so much variability in each instructor’s teaching ability, it was really difficult to build up a brand reputation based on a bunch of not even necessarily professionally trained teachers. More on that later.
R: I agree, much easier to market a handful of highly experienced and pedigreed teachers instead, who might already have a following. But Yuanfudao, because it began life as a test prep tool and not a service, took more money to make the business model change work than GSX. Or at least that’s my simplified interpretation of the situation. I mean, it’s already raised over ten times the $50mm by GSX … meaning it’s raised well over half a billion dollars.
[26:46] Y: Another well-pedigreed CEO here. Li Yong 李甬 started his career as a journalist before joining Netease as an editor in their portal business and eventually being promoted to VP. When he quit to start his own company back in 2012, he was one of the few back then who could command angel funding pre-product. While the first product, which was also in education, failed, he managed to find success with Yuanfudao.
R: Not that much success. Its last confirmed revenue was about $200mm in 2018, right before it received another $300mm in funding led by Tencent, but we don’t know about its profitability or growth. Many are skeptical of its valuation, pointing out that its valuation is triple that of ZYbang 作业帮, another K12 test bank and livestreaming edtech company that claims 800mm cumulative registered users.
Y: By the way, we have no idea what to make of these user numbers, because the total number of K12 students in China is only 173mm, so we don’t know if that’s just counting multiples of the same user across different apps under the same company or what. Yuanfudao also announced with great fanfare that it’s exceeded 400mm registered users, which is also mind bogglingly large, so I think the only wise course of action is to totally ignore all these and focus on other metrics that actually matter.
R: Like paying customers. Less than 1% of Yuanfudao’s registered users were paying customers in 2018. And online tutoring, as you would expect, is much cheaper than offline. By about 50%. That means price points typically range from $6 to $20 per hour, depending on the subject and format. And renewal rates are also low, which means that as low as only 20% of students choose to renew their subscription, well below the typical 70% for offline providers.
[28:57] Y: Not to mention the always present policy risks. The government is very active in both promoting the edtech space and also in making sure things like online tutoring don’t become overheated. There’s been a fairly steady drip of circulars, or suggested implementations of the laws, and I guess most of them are meant to protect students. For example, training beyond the scope of or higher than the level of the textbook is banned. So are classes after 8:30PM. Unfortunately, these are currently not followed.
R: Other rules are also very consumer-friendly. Such as banning the collection of more than three months worth of fees in advance and requiring teachers to have the proper credentials, all very reasonable. But don’t think that means everyone is compliant. That’s far from the case. I mean, as of IPO, about half of GSX instructors did not have the necessary credentials. And Yuanfudao had been embroiled in scandal for exaggerating the pedigree of their instructors as well.
Y: All these were existing headwinds for the online K12 tutoring industry until the coronavirus struck. But then, as you now know, schools had to close, tutoring centers had to close, and everything has to be done online. And here we are now, that $64Bn market opportunity for after school tutoring left wide open for the online players to go after. The 90% of the market that isn’t online yet … is this the time for them to win them over? GSX and Yuanfudao both made some of their paid products free during these dangerous times.
R: Yeah, and GSX announced that it got 15mm new users during the epidemic, 5 times the total number of paying users for last year. I mean, these services are going to seem so much better than the regular school system. The traditional school system doesn’t know how to do this sudden shift online. Sure, you can tell your teachers, do some pre-recorded lessons, do some livestreaming, but most full-time teachers they don’t know how to use any of these tools — why would they? Sure enough, almost as soon as they started, the complaints began. We won’t enumerate them here but a lot of hilarity ensued. From forgetting to turn on mics and cameras to trying to deal with students with funny nicknames … let’s just say things did not run smoothly.
[31:38] Y: And while larger edtech companies like TAL are providing infrastructure solutions for the schools, many teachers just ended up using the available collaboration tools such as WeChat and DingTalk. In fact, DingTalk is being used a lot … and poor Alibaba actually had to beg students to stop flooding the app with one-star reviews just because they didn’t want to go back to school.
R: And that brings us to the end of this introductory lesson to the online K12 tutoring market in China. We chose to go over it today because it is most certainly getting a boost in the short term from the COVID-19 epidemic, but also because truly, edtech is a huge industry in China and definitely something we’ll cover more of in the future.
Y: Yeah, to reiterate, K12 tutoring in general is a $64Bn industry in China, and the vast majority of parents are spending a few thousand USD per year on after school lessons for their kids, and a quarter of them are spending more than 5 digits. Over 90 percent of the market is still taking place offline, but online tutoring is already an almost $5Bn business and it might grow ten times in the next few years.
R: The main reason for this, of course, is the fact that the college entrance examination or gaokao, is so important, and also so hard. Only 40% or so of the students who take it qualify to go into a 4-year bachelor’s degree program. This means that students in China literally spend all of their time preparing for it, and even the epidemic probably won’t delay it, because 17 years ago SARS didn’t. The Ministry of Education is tough when it comes to tests, unlike the College Board, which just cancelled the March SATs in China.
[33:32] Y: The Ministry of Education also has a super tight control over the education sector in China, so businesses operating in this area should always be very aware of policy changes, which could make or break your business. Luckily, they’re pretty bullish on integrating technology with education. Unluckily, some of their rules could drag on operating margins, although most of them are geared towards protecting the consumer, such as requiring the necessary professional certifications and such.
R: We picked two businesses to highlight briefly — the publicly listed GSX, and the still private Yuanfudao. GSX has gone up quite a bit since the virus hit and is now a $5Bn market cap company. Yuanfudao is last valued at $3Bn but already raised around 10 times the capital GSX had before going public, which was just one round of $50mm.
Y: While GSX started as a marketplace for teachers and Yuanfudao as a test bank for K12 students, both eventually ended up in the large-class, dual-teacher online K12 tutoring space. It’s what accounts for over three-quarters of GSX’s revenues. For GSX, the classes are getting bigger every quarter and are now at almost 2,000 students. For both, the classes are taught by one celebrity teacher and supported by a group of tutors who then monitor smaller groups of students.
R: Unlike one-on-one tutoring such as the model utilized by VIPKid, large classes are much more scalable and can build up brand recognition much more quickly. It also makes sense if you believe that it’s the teaching talent that is the most scarce resource in tutoring and not necessarily the one-on-one personalization. Regardless, these B2C large-class providers are experiencing tremendous growth right now by offering free lessons for the stuck-at-home K12 students who must continue their studies despite the epidemic. In contrast with their non-tech-savvy regular teachers, the livestreaming gurus of these platforms, we think, are going to gain lots of new fans and paying subscribers.
Y: What do you think? Have any questions? Send us all your questions and comments at ying@techbuzzchina.com or find us on Twitter or LinkedIn at techbuzzchina! And our new website, techbuzzchina.com!
[36:13] R: OK, that’s all for this week folks! Thanks for listening. We really enjoyed putting this together, and we are always open to any comments or suggestions. You can find us on twitter at thepandaily, at techbuzzchina, and my personal Twitter account is RUIMA.
Y: And my Twitter is spelled GINYGINY. Tech Buzz China by Pandaily is powered by the Sinica Podcast Network on SupChina. Pandaily.com is an English language site that tells you “everything about China’s innovation.” Our producers are Caiwei Chen and Kaiser Kuo. Thank you for listening!