EdTech Series 3: Huohua Siwei (Spark Education)

Hans Tung  

Mark is a serial entrepreneur. Before SparkSpark EdTech he was part of the founding team. China’s earliest classified website Ganji.com and served as its CTO, the company was acquired by 58.com in 2015. Mark graduated from Singapore National University with a bachelor’s degree in Computer Science. Mark, as part of founder team in ganji.com, you had a great exit in 2013 when the site was acquired by 58.com. Why did you want to start another company? Why not just retire?

Mark Luo

Yeah, actually, that happened in 2015. Actually, I was only 36 years old, an age where we were sort of starting a new venture. And again, you will have many copyrights who were eager to start a new company together once again, a company that would create some value for society. Now that’s why I start a new company.

Hans Tung 

Why education?

Mark Luo

I joke that the reason many seasoned entrepreneur has ventured into the education field is because they start having kids. I don’t know if it’s true for everyone. But at least that’s the case for me. Once by chance, I learned from my wife that my elder son was taking after school mathematics class, and decided to sit in. The class was great, and somewhat similar to the Western teaching style.

But there were clearly some disadvantages of the offline setting. First, each session was one hour and a half, too long for kindergarten kids. Second, every time my son complained to me that he didn’t get called to the board because only one child gets an opportunity to come to the front when the teacher posts their question despite many puts their hands up. Finally, the commute time from home to the school is also a natural barrier. This made me wonder how I could make the home learning experience more effective by bringing it online.

Hans Tung  

It’s not just when the founders are older and have kids, they want to do education. VCs as well, we definitely see some kind of parallel.

Rita Yang  

Yeah. So for people who are not familiar with the education system in China, why can Spark product take off at this particular moment in China and even create an entire new category? The stereotypical understanding of Chinese parents is that they’re only willing to pay when they can see a direct impact on their children’s scores.

Mark Luo

Yeah, first of all, I want to say in China, parents’ mindset has changed. Chinese parents, our primary parents born in 80s tor 90s, the expense of Chinese college enrollment means parents today are more well educated. They have been through the boring rote learning and endless exams themselves. And no, this doesn’t inspire the real interest of children and hope their own children’s learning experience would be somewhat different.

Second, our product mainly focuses on an age group between three to nine, where parents are more focused on interest formulation and the development of mathematical thinking skills, rather than academic scores. Today’s faster evolving social media such as WeChat, and Tiktok, have also changed the way products are marketed. Social media has definitely facilitated the spread of costs with good reputation. And second, the whole picture is our product is totally different between the products before our Spark. First, we solve the problem of offline costs, eliminates the commute time from home to school, and every student can solve problems or interact with teachers and each other at the same time. Moreover, because we use a lot of interactive courses well, so online classroom enables the collection and analysis of performance and behavior data, which could be applied into the course development and gives children more scientific and enjoyable learning experience. I think that’s why we can take off.

Hans Tung 

A lot of EdTech startups that invest heavily in front end sales, Spark, spend most of the investors’ money actually in R&D. Please tell us how you structure your product team. We heard that there’s an incredible mix of developmental psychologist, learning experts and even children songwriters, how do you get everyone work effectively together from these different backgrounds?

Mark Luo

Yeah, the team format in Spark is different from other teams. We invest in every cost element that sparkles children’s passion for learning, that’s our objective. We have experienced costs researchers who design the curriculum and the content suitable for children at their respective age. We have producers who packed the content into storylines that children like. And we have developmental, psychological, who make sure that the content is delivered to children in a way that they can accept. Finally, as you mentioned, we have children songwriters who embedded the classroom knowledge into beautiful and logical lyrics and songs that help children memorize the key concepts. Our children can immerse themselves into a story scene, play games with our virtual characters. We have three characters. And they play games with our virtual characters, clearing each level by solving challenging mathematical problems, and sing songs while learning important acknowledges. Yes, this is how we get everyone to work together.

Rita Yang 

We know that Spark started with small group classes where one teacher teaches six students, how did you decide on this particular way of offering?

Mark Luo

A lot of people ask me why we don’t use about one teacher to one student. Why did we start with one teacher with six students? Based on research, and my observation of my kids’ learning experience, I believe it’s the best format of learning that requires both attention and interaction. And a small group class setting offers both healthy competition, interaction with peers,  enough individual attention and the tailored instruction from the teacher.

Hans Tung  

Okay. 75% of Spark’s customers come via word of mouth. That’s an amazing statistic. How do you manage to get to that? I mean, you have a fantastic product, but what else has been done to achieve this kind of word of mouth impact?

Mark Luo

I think the most important factor is the product. You know, we all have the experience of referring some products to our friends. And the only time we become natural sales is when we truly appreciate the product. Many of our parents are highly educated people with very busy schedule. If it is not about the product, those referral bonuses, and the monetary incentive, are not something that would motivate them to recommend our product to their family and friends. Of course, as I mentioned before, social media is a good field to spread our reputation.

Hans Tung  

A lot of people don’t realize that growth hacking is indeed important. But having a superb product design will make the job much easier, at that point is not emphasized enough.

Rita Yang  

So the success of Spark actually created a new generation of EdTech startups in China that focused on products similar to Spark. Some uses AI purely, some others have a mixture of AI and human teachers. How do you look at competitors in this bucket?

Mark Luo

I think different kinds of format or companies can give a different experience to our children. I think competition is actually a positive thing. It drives efficiency and innovation. But I hope our competitors, regardless of their forms or products they offer, will remain true to our original aspiration to make education more interesting and effective for our children.

Hans Tung  

Spark, like many other EdTech companies enjoy an incredible wave of growth during COVID-19. Please tell us about that. And how you managed to keep up with the new customers and skills to satisfy their demand? What are the growing pains that happened during the process that you can share?

Mark Luo

Everyone is saying that the virus actually is a big benefit to the online education company. Right. But I think that the virus is a double-edged sword for Spark. On the one hand, more parents are exploring online classes, and they have more time to accompany their children to attend online classes. On the other hand, Spark is constrained by the teacher supply. Unlike large classes like Zuoyebang with a low ceiling of student numbers. In February and March, we have a lot of difficulties recruiting sufficient teachers and we have tons of students waiting to get into the class.

Rita Yang  

So how are you thinking about growth at this particular moment? Is that more about breaking into new consumer groups in different age? Or you’re thinking about offering new products?

Mark Luo

That’s a good question. The whole education sector is growing at a very fast pace and every participant is going all out. But as Spark will not follow the group, we pursued what we call healthy and sustainable growth.

Hans Tung  

Got it. Mark, what is the biggest adjustment that you had to make from being a CTO before to now being a CEO and founder of an increasingly large company,

Mark Luo

Being a CEO entails more responsibility. The pressure from outside is to talk to investors daily to get enough funding, from inside is where you are responsible for the cooperate, strategic organizational development, and the day to day business operations. One striking difference is the complexity of the team and the people you manage. As a CTO, you talk to technicians and programmers who speak the same language as the CTO in an EdTech company. However, you’re interacting with teachers, course researchers, sales, product managers, programmers and so forth.

Rita Yang  

If you can go back in time and give one piece of advice to yourself when you were building your first company. What would that advice be and why?

Mark Luo

Don’t rush into new ventures. Only start something when you have really thought through it. Actually in 2016, I started a new company called Wanduoduo. Yeah, I think that time, I didn’t think it through.

Rita Yang 

Okay. So now we’re going to end this episode with a few quickfire questions. So just say whatever comes to your mind first, if you could spend one day in someone else’s shoes, who would this person be and why?

Mark Luo

Everyone is talking about one person recently. Yeah, I would choose Elon Musk.

Rita Yang  

Okay. Why?

Mark Luo

He’s sharp in thinking, quick in decision making, bold in planning, and overall, such a character. It is almost unthinkable what it takes to motivate a group of intellects, entrepreneurs and engineers to gravitate toward him and his dream. I would also like to know what hardships Elon has to endure to build an empire and how he overcome these difficulties. Yeah.

Rita Yang  

What is a book that you recently read and would recommend to other entrepreneurs?

Mark Luo

Recently, I read a book called “The Five Dysfunctions of a Team” by Patrick. I recommend everyone read this book. The book is about the team different way for a team to fail. For example, is the absence of trust. the lack of commitment, something like that.

Rita Yang  

Got it. Last question. Who is the entrepreneur you admire the most? And why?

Mark Luo

Of course Elon Musk.

I also admire Bill Gates. He changed the world.

Rita Yang  

Now he is still changing the world.

Mark Luo

Yeah, actually, I was working in Microsoft before. I saw him several times.

Rita Yang 

All right. So I think that’s all for today’s episode. Thank you so much for being on the show, Mark.

_____________________

Rita Yang  

Hi, Jenny. Welcome to the show. Again, it’s so nice to have you.

Jenny Lee

Hi, guys. Good to be here.

Rita Yang  

So how do you usually explain Spark EdTech to an investor who is coming out of China given that it’s offering its quite unique in a sense that it didn’t fit into the usual suspects of EdTech product?

Jenny Lee

Well, actually, that’s why they do it in a different way. So I think first of all, they target the younger student population, 3,4 years old, up to 9,10 years old in the computational and logical thinking development area. So one, I think that’s what the team wants to do in terms of bringing better structured logical and thinking courses to them. I think when you refer to it being different is the method at which the course is being delivered. So you know, in education, we have one to one, which is what we hear a lot historically, from traditional to even the first EdTech 1.0, where it’s really just bringing teacher to student one to one. And then you have the larger class format, which is one to many auditorium, where you listen to one professor top rate, one lecture or one speaker, you have a lot of people just listening in, which is the kind of webinar format. I think what Spark education has is the in-between, they have one to six students. Basically, we call this a small class format, where you do have four to six students online, studying at the same time. But what’s different is that they utilize a parallel track. So instead of the reliance on teacher, the physical teacher delivering the content, the content is actually a whole series of gamified experience and content. So, it’s animated, they have their own content. And each of the class essentially, you’re listening to a recorded class. And they do it differently, learning experience different animation. And the teacher or the person in the loop to guide the student is like a tutor, so the tutor is there at the same time life, but the student could address to the whole class of four to six students, and be able to coach them based on their learning ability, how they’re reacting to the so called virtual teacher, right, which is you have, like a virtual teacher, and then an actual tutor. So this model is kind of halfway between, the one to one model, and to one to many. In terms of efficiency, it’s not the same level of efficiency as to have one teacher for one student who can customize hundred percent, but on the other hand, the economics, and the difficulty of finding that many teachers for the one to one causes is exponentially higher. In fact, you cannot find that, versus the efficiency of having one super teacher teaching a class of little kids, you can just imagine the chaos you’re going to have, but as well, because this is a younger student population. So, I think the in between model is really what we are looking at what Spark is doing, which is a smaller class format. But the smaller class format is further enhanced by the fact that you have this virtual teacher in the loop. And then it’s supplemented with a tutor, which is like a live tutor to help the students stay focus. And to help them along, if they meet with any roadblocks in the learning process.

Rita Yang  

It sounds like when you were first evaluating the deal, their one small class model really impressed you. Did you have any concerns back then that you think that we might need to watch out for when evaluating that deal?

Jenny Lee

I think in the early days, there were all kinds of format, the one to one format, the one to four, one to six, and one to many. And so I think the big consideration is the math of it, does it work? From a unit economics perspective, what’s the best model? Is that business model sustainable? Can you sustain this into first year, second year, third year? Can you have, can you sustain the first thousand students and then 10,000 students and more? So I think we did spend a lot of time trying to understand the economics and trying to look through the numbers, trying to ascertain that model makes sense, but it is an economic model sometimes it can look very pretty on paper. The reality, though, is very different, because at the end is the students who will be the end recipient of the service. And then ultimately, we do have to see because you also have teachers in loop. So this is very different from the format, of course, it’s all AI, which is now what a lot of the startups are also experimenting, which is you take the teacher completely out of the loop. So, it’s basically the pure AI model where you don’t have teachers. But in the first few models, where you do have teachers in the loop, then delivery of the service becomes very important. So the other aspects that we have to do on each of the companies that we look at really is we have to make those call, the calls to the parents, the calls to the students, to try and really verify that indeed, assuming the numbers work, the model work, the recurring, the renewal, all those good stuff work, then it’s really about, is this a good product? do the parents like it to the other students learning or the students engage? So those are areas we cannot just take a look at the CEO, and look at this cool number and say, Okay, well, you know, I think education really is a pretty complex model, the tech model is that it has to deliver real performance and results. And yet the whole process includes many different players or participants right from the end target of the student, then the parents, and then the teachers, and of course, the platform itself. So pretty complex for companies that can do it well, I think they have a chance to scale very well. But it’s definitely about taking the last step of the way.

Rita Yang  

Yeah, Mark actually shared with us, like during your first meeting, you already mentioned the potential of their product going into markets outside of China. So what about the product or him as an entrepreneur that makes you think that they can go beyond their home market?

Jenny Lee

I think if you really look at a few dimension. First, it is not a hardcore curriculum-based content. So if this is math, for grade six, grade seven, then typically they are tied to the school, the country, in fact, even you know the local level curriculum can shift and be very different, including the testing part of it. But when you look at computational logical thinking, it is broad-based, it’s part of holistic learning. And so globally, wherever the kids are, it is one piece of skillsets that parents want their kid to develop. So the positioning itself, the skill set that it conveys, it’s very globalized.

The model to deliver the service is also less reliant on particular quality of teachers, because the teacher is a virtual teacher. In this case, it’s that experience, it has to be localized as well. But it tends to take the form of language localization, let’s say if you go, you want to go to Southeast Asia, Indonesia. You have to localize to local language, if you want to go India, do you need to have the local Indian language, but in their case, because there’s no real teacher in the loop on the teaching part, it is also scalable, it is really a matter of technology, and refining the product. Second, the model can skill, , so I think that, of course, this need overseas as well, it’s not smaller than China, I think China is already a huge market. But when we look across, you know, to Southeast Asia to India, I think the population is three to nine. Asia tend to be younger than China. And so you can think about how the next, you know, five to 10 years, they will be the next generation that’s ready to consume in such content as well online.

Rita Yang  

And speaking of that, how do you see the competition with bigger platforms covering multiple subjects like to Zuoyebang, which is also a portfolio, or Ruangguru in Indonesia, and a vertical player like Spark that started with logical thinking, but now also expanding into language and other categories?

Jenny Lee

I think the great thing about education sector is that it’s a big sector, and they are different sets of students, whether they are the preschoolers, the the K 12, or the adult, it’s a huge market. But the challenge there is that it’s going to be really hard to have one player take all, just as in offline rule, you have all the different schools in China, like Tsinghua, Peking University, you know, all the different schools, you’re never going to find just one school. Why? Because the whole impression, the positioning, the price point, the localization can have its nuances and differences. So, parents are not tied to just one brand and be happy with the brand. I think parents continuously for this generation, I think they will continuously try to find the best for each subjects. And so if Spark can maintain their lead with this on the logical thinking side, then they have to, of course, it is not a static thing, it’s dynamic.

The company, the CEO has to continue to keep the product at the cutting edge, continue to make it engaging to the students, the students have to choose and say, Oh, no, like for logical thinking, I just want to do this. And maybe, for homework questions, they go to Zuoyebang. So again, it’s so hard, there’s so many different variances and feature set that we’ve seen across all the education tech startups very innovative. It’s almost impossible to imagine just one company doing everything and doing them at the best.

So I think our advice for our EdTech startup is that they truly need to have an anchor product. The anchor product could be the tools that Zuoyebang have as the home assistant, or it could be the content delivery that Spark has. And then of course leverage that to build a strong brand and a strong following. And then subsequently, when they have the customers, then it makes sense to expand. Now, in terms of topic of expansion as well, it has to be natural, I think it’s really tough. If you’re targeting, you know, three to nine years offset and to be targeting adults, you can, but it would require a different product, which means that there may be another player who’s only targeting adults, and then that’s where you have to contend with them. But if you’re looking at cross subjects into still the same demographic, three to nine years old, so they are crossing into, let’s say language, which is what Spark has done, then again, the demographic similar, you’re selling to the same parents. And so if you already have an external installed base of trust with the parents, then I think getting them to try the new product, or the new subject becomes a bit easier now, whether you get to keep them really depends then on the quality of the product, and the service level that’s there. So it’s going to be a fight. And, you know, we hope all players win. But I don’t think we can avoid them potentially, in some areas meeting each other. I think we just have to get them ready to say that, hey, you have to be the best that you can be in your category.

Rita Yang  

So Mark is actually another CTO turned to founder in your portfolios. There’s a pattern that you really like this type of founders. Having worked with him and Tony from Agora, what are some patterns among this particular group that you think it’s worth sharing?

Jenny Lee

Yeah, it’s true – like all the tech guys. So I feel that the next generation of new innovation will evolve. And he has evolved from just product features centric, to now also product feature centric, but it has to have unique angles. And the way to do it is to leverage technology. whether it’s leveraging natural language processing, whether it’s voice, you know, technology, whether it’s in Spark case, you know, leveraging his technology knowledge to ensure that his product, his virtual teacher. I love all startup CEOs, to be fair, but I think when CEOs come with a very strong technology background and experience, and they’re trying to disrupt traditional industries, like education, then the ability to convey the product and implement the product becomes very important. So if they are able to understand it quickly, and make the right call quickly, then that product gets developed a lot faster, and so get to market faster as well. Now, you could always be a copier and say, Oh, you know, there’s a product out there that looks like that, and then get your IT guys to copy. But you always be a step behind. And so I think in this new era, which I believe is where we are today, the CEO that has this right skill sets, the right kind of background, to now look at leveraging that skill set into a product, rich feature of rich product, that means they can run faster, innovate faster, they don’t have to wait for their customer service to tell them, the customers are telling them, they can go take the call from the customers understand whether this is a real need, and then correspondingly direct the R&D to get it done. And if the R&D don’t get it, they can work on it. So I think that’s the relevancy of those backgrounds. I have to say, though not all CTO type background, make good CEOs. So to be a good CEO, you have to rise above technology as well. We have also seen cases where the CEO is very tech-savvy, but very immersed, and sometimes enamoured with just the idea of tech. And so in this case, you may take it too far. And if you take it too far, and you think that you’re there when the market is not quite there yet, then unfortunately, there may not be a product market fit. So it’s great to have those backgrounds, but it’s important to use it in the right context for quicker turnaround, for quicker understanding of the customers, but I wouldn’t say that if you had the tech background, you’re going to be good CEOs.

Rita Yang  

Yeah, amazing. Okay. Are there anything about EdTech that you want to share that idea and I didn’t ask?

Jenny Lee

I think right now, we are in the second inning of EdTech. It’s not like in the first inning where we have a whiteboard, and we’re trying to figure out which model works, which are the format that could work and what kind of subjects I would say that all the different dimension of how to get into the tech market is pretty much laid out on the board. And so, in the second inning, it’s going to be about the sensation and execution. How you differentiate the energy could be that little tiny differentiation. Now, everybody may use a virtual teacher, then how are you different? And so maybe you have to differentiate by your customer service, by your effectiveness, things like that. So other things are all going to come into play in the second half along with that capital becomes really important as well. Because you’re gonna have some little giants and the giants right on the map. And so if you’re a little baby trying to get your start, then you’re going to be up against some very tough competition. I wouldn’t say it’s not possible, but you know, the fight in the second inning is going to be very different from the fight in the first inning. I would say it will be harder, tougher. So entrepreneurs really have to think through and not just follow the wave. Not everyone can end up doing education now, because many ways of monetization services are already being done by the folks in front of you. So to do well, you just have to leapfrog and certainly tougher, so think harder before you do another attack due to the market right now is truly in the second inning.

Rita Yang  

All right. Thank you so much, Jenny, for being on the show with us.

Jenny Lee

Well, thank you always happy to be here. Thank you.