Jay: This week’s show guest is Tim Yu, co-founder and CEO of Snapask. Snapask is a Hong Kong-based education technology company that offers tutoring services to students. With Snapask, students can instantly connect with qualified tutors to receive on-demand, academic support around the clock. Tim, welcome to the show.

Tim: Hey. Thank you. Thank you for having me.

Jay: It’s exciting to speak with you because I’ve actually heard of your company and the great things you’re doing. So, I’m excited, and as someone who has three kids myself, education is definitely at the forefront of my mind and my finances. So, it’s going to be interesting to have a chat with you. Maybe you can give some background for the audience of who you are, where you came up, and this sort of thing.

Tim: First of all, this company is actually my first thing I’ve been working on since I graduated. Even when I hadn’t graduated yet, I was working as a private tutor. A lot of students, in uni, they work as private tutors, but then it’s not a very efficient thing to do because you have to go door-to-door and spend about one-hour back-and-forth to the student’s place to do the classes. It’s the pain point of a lot of tutors, and even if they teach very well, they are bound by all these constraints. They have to commute; they have to spend time commuting, which is not very efficient.

That’s why, back then, we tried to do it online. I did it first by recording a lot of videos of myself teaching. I put it on Facebook, and quickly we caught a lot of attraction. Interestingly, a lot of students, after they watched my videos, they started asking questions by inboxing me. So, that actually gave me the idea of “Hey, why don’t we just try and monetize this, this question-asking behavior and make it into a real business instead of just answering questions online?” That’s how the idea came about.

Jay: Wow. That’s pretty interesting. Tim, you’re from Hong Kong originally? Is that right?

Tim: Yeah.

Jay: Okay. Great. Cool. So, straight out of school, then, because of your experience as a tutor, you decided to start a company. Was that from an entrepreneurial standpoint? How did your family and parents feel about that? Were they supportive of that, or were they sort of, “Maybe you should get a regular job first.” What did they think?

Tim: Yeah, there has been some struggle for both myself and my parents, what I should choose after I graduate… I was very honest to myself and to my parents as well. “Hey, this is something I really am passionate about. I think there is a future to being in this and education. As long as I can support myself, you guys should support me as well.” That part of it was pretty clear to myself and then to them, so I got their full support. But, along the way, it’s not enough to just have your parents’ support. Right?

You have to really make the business work. You have to work with your team. You have to build a team, build the products. These are all things that I hadn’t been trained to do. So, it’s a first-time manager, first time looking into products, not much work experience. So, everything was new to me, which was a pretty big struggle.

Jay: Yeah, and it’s exciting, and it’s also commendable because, most people, when they first splash out, they have some sort of work experience in the past, so there are some sort of soft skills at least that you learned. If not, not to — some people actually have technical skills that they’ve worked in tech or whatever, but you’ve done extremely well with Snapask. So, that’s congrats to you for your success thus far.

Tim: Thank you.

Jay: What subject were you tutoring that you realized that “I can start interacting with people online.”

Tim: I was teaching math back then. Math is a compulsory subject. Because I’m a math major, it’s the easiest thing for me to sustain myself. That’s why I chose math to tutor. But the interesting thing I saw was in the entire tutorial industry, back then and even right now, a lot of tutorial schools, they do classes with one too many. When I say “many” it’s actually 30, 40 or even 200 students in the same classroom.

This kind, we call it the “cram schools.” We do not prioritize interactions and the actual learning with interactions. We prioritize spoon-feeding, intimation to students, to actually train them on questions, on practice paper, and to make sure they can do the exam, which I personally think it’s not the right thing to do. That’s why even when I was doing tutoring I really did one-on-one or even small classes to make sure students were getting the most out of it.

Jay: Right. So, this is actually interesting because—first of all, if it’s multiple kids in the same class, I guess, for the parents, it would be slightly cheaper because it’s not private. Right? That’s not one-on-one. But then the methodology, you’re saying, that they use to teach also is not as effective. Is that what you’re saying?

Tim: Yeah. I mean, in a sense of affordability is something we have to think about. There are a lot more effective ways and efficient ways to make it affordable, online being one of them. But, the most important thing is not just about the price. Parents, they do want to invest as much as possible in their kids’ education. So, I think affordability is not the topmost priority.

The thing, instead, is “how well my kid can learn from this type of methodology?” and “how is this suitable for them?” “In the long run, can you guarantee they’re actually learning something instead of just being an exam machine?” These are all things that we care so much about, not just when I was teaching. I care a lot even when we are building our product right now. It’s something we’re building in terms of product features, in terms of the content and learning materials. Is it actually helping students? This is stuff that we prioritize all the time.

Jay: Yeah, that’s a good point. To something that you mentioned earlier, first of all, you’ve picked a very, very big segment or niche because education, obviously globally, but particularly here in Asia I feel is like you said, it’s not as price sensitive. People will spend money on kids’ education. It’s a good niche that you’ve picked that has long-term structural growth, so to speak. But, it’s interesting. I want to dig into the process there in a little bit.

Before that, maybe you can just tell us how you went about assembling your first team as a startup founder. Who are your team members? How did you actually—from concept to actually pulling together a company, how did that process work?

Tim: As I said, I hadn’t had much experience working in a company or even in a group. The closest thing I did as a group was a group project in school. What I spent time the most on in the very beginning was product. In retrospect, I think it’s not the best way to do it. When I think too much about the technicality, how well we can deliver it — the service, the features — instead, we should have put much more time into researching about what are the users’ needs and how would they want us to present this need to them.

We went straight in to build a team of engineers to really work on a product, to really build the first Android app and websites. Then, eventually, the product we launched, no one was using it because that was not Snapask back then. We were just building a platform for students to pose questions like a forum. Before it evolved into Snapask, we spent too much time on actually building something we assumed could be useful, and our users just didn’t think it was what they wanted.

Jay: Was that something like Reddit, or Quora, or something like there were people—

Tim: Very similar to that where people can post not just images and text, they record a video and put it there and have machines to keep asking questions and a response comes it.

The lesson we learned and eventually how we built the team was a lot of growth hackers. A lot of people who basically they do a lot of experiments. They have a very small assumption. Let’s say I want to prove if students want to interact within 10 minutes with a tutor or if they want to do a 30-minute session with someone. So we’ll do a lot of tests, a lot of experiments, before we even build products. It includes research, a lot of mockups, some prototypes with the closest possible product we can find online without any cost to approve the assumption before we build with the engineers. That’s the biggest difference from the beginning until now. We don’t just look at how to build product but why we should build this feature and what exactly it should be presented as.

Jay: Right. How about your team? How did you go about finding co-founders or developers and this sort of thing? Did you just by friends and word of mouth and that sort of thing?

Tim: Unfortunately, I don’t have a co-founder. I have just myself. Most people would say a company with co-founders would have a higher chance of success, but unfortunately, I don’t have that. So, I was working solo since the beginning. Very fortunately, I met good people who were passionate in startups, who really wanted to build education products that could change the usual way of how people learn.

So, we got our first product manager joining us. She’s from Taiwan, lots of experience with Microsoft, the days with Microsoft and in product management. We started, really, to build Snapask from the ground up. Eventually, we hired — because the thing is, Hong Kong is too small to be a market that’s attractive to any foreign investors or to any team members to really invest that amount of resources to build the product. That’s why we expanded very quickly.

We found also a lot of members from different regions and now, accept Hong Kong and Singapore and Taiwan, we also launched in Malaysia and Indonesia, Japan, and Korea lately. These are new countries that we also built our partners on. We set up an office to hire marketers, to hire business-development people. That’s about it. The entire story of building a team.

Jay: That’s incredible. This whole time you were just funding it yourself, initially, or friends and family were helping out and that sort of thing?

Tim: In the very beginning, we were self-funded because I was tutoring. There was still some cash for me to sustain the business. Eventually, we were very lucky. We got supported from some seed investors, some angel investors, who invested when the prototype was still not — we hadn’t formulated the entire strategy yet. It was still a prototype. It was still just a lot of users coming in, but we hadn’t quite figured out how to monetize yet. That’s the stage where we had very good investors, not just investing money but their knowledge and the experience in our business. That’s why we fly.

Jay: That’s incredible. Great story, though, on how you just started as a one-man band and expanded. So, Tim, can you—let’s walk us through the user experience of Snapask. Is it an app that you download, or is there a website as well that you can just log in? I’m looking at your website now. I see a login at the top. Is it linked up to the app on your phone that you can download?

Tim: The website, we do offer the same service on the website right now, but the core offering is still a mobile app. The app, it’s pretty simple to use. Once you’re logged in you just choose the subject that you want to ask a question about. Usually, it’s a student using it. High school student, every day they have tons of homework. Typically, a student will spend over 3.5 hours a day on the assignment. That’s the Asian’s life. When you have like eight subjects that you have to spend over three or four hours on, when you have questions that you cannot tackle, it’s very seldom you can find someone just from your own network to solve the problem for you. If your family’s lucky enough, you can have a private tutor, but that’s not the case. That’s not the luxury for most of the families.

They just simply use the phone to take a picture of the question and send it out on Snapask. With our matching algorithm, we actually understand the profile of the student, the subject, and we connect the best suitable tutor for them. The tutors are all from local universities. Let’s say if you’re a Hong Kong student asking a question in math, we’ll connect you with someone from Hong Kong, from the top university in Hong Kong, and who knows about your subject. That’s the same case in other countries as well.

We do vet our tutors based on their university transcripts, their public exam certificates, to make sure, if you want to teach math, for example, you have to get an A before you’re qualified to teach on Snapask. It’s kind of an insurance for both students and parents.

Jay: Sure. Okay, that makes sense. Let’s say I’m a student and I download the app and I’m having trouble. Obviously, math is one of the subjects that you…because you were a math major. What are the other subjects that you have available for this service?

Tim: Most of the questions are actually from math and science — science as in physics, chemistry, some geography, and biology. Over 80% of the questions are from these — from math and science. But, there are more and more students asking questions in subjects like accounting, economics… It’s a very wide range of different topics that they can ask about. We do cover all the subjects in any given curriculum.

Jay: What’s the age group that—what’s the earliest that someone would use Snapask?

Tim: The earliest is when they first enter high school at the age of 13, 14. We do have a lot of users from there. But, most users, they become very intensive question-askers when they turn 16, 17, when they need to face the real pressure, which is the public exam. When they’re preparing for this exam, which is the only chance for you can get into a uni, they become very serious. They have to ask questions like five, six times a day. Every month they ask hundreds of questions, so that’s a serious time.

Jay: Wow. Okay, so I’m in high school preparing for my exams. I’m 17 now, let’s say, Tim, and I’m preparing for my math exam, and I just don’t understand this question. So, I load up the app on my phone, snap a photo of it, send it through the app. You have your matching algorithm that pairs it up with a tutor. Then, what is the flow that happens after that? So, a tutor picks it up. Then, what happens?

Tim: So, we were actually called “the Uber for Tutoring” by BBC. They call us that because we actually have the instant pick up just like an Uber driver. When the question is out to the tutor, on the other side, they will get a notification, but we don’t send it to all tutors. We only send to the most suitable bunch, but they have to fight for it. It’s first-come-first-serve. If you’re the first one to pick up, you can answer the question.

We also reward the tutor per number of questions they have answered. Based on their rating, we decide the reward level. Tutors do have the incentive to pick up as fast as possible. Now, the average pick up time in the peak hours is usually five, six seconds, so we do guarantee this kind of status level for most countries during the peak hours.

Jay: Wow, that’s incredible. When you quantify with metrics like that it’s pretty interesting. It’s interesting to hear there are actually peak hours that people are asking questions. I assume that’s after school, after dinner, that time when they’re actually digging in. Right?

Tim: Yeah.

Jay: Tell us a little bit about the revenue model and how that works. How much does it cost for a student and how much can a tutor potentially make?

Tim: For tutors, they get paid per number of questions they answer. A typical tutor they get paid about $1 U.S. per question which lasts for about 5 to 10 minutes, but they can answer multiple students at the same time which is very different from private tutoring where you can only cater to one student at a time.

For the student’s side, we charge them per month. Every month we charge a monthly rate of about $80 U.S. It’s all-you-can-eat. You can ask as many questions as you want on any subjects, which is a pretty affordable cost as compared to most of the other tutorial services.

The peak hours are usually after 8:00 p.m., and most students ask them until 2:00 a.m. It’s a peak of the peak season, peak hours during the day.

Jay: Wow. I switched gears. So, a tutor picks up my question. What do they do? Do they just literally answer the question or is there more interaction? Do they explain the process? Do you know what I mean? How does it work from there?

Tim: The entire teaching experience is within Snapask, where after they were connected, they answer into a one-on-one session just like when you’re using Websmap, or WeChat, or any chatting app. Usually, students and tutors use text, they use audio, and image. They write down their solution, take a picture and send it back, and you elaborate with audio. That’s usually how the entire teaching flow goes about.

Jay: Ah, I see. The other question I had was the tutors, are they all college students or they’re college students and above?

Tim: Most of them are university students. They just finished the public exam so they have a lot of experience in how to take an exam and how to actually solve a problem with the simplest way. That’s why we recruit them as our tutor. But of course, we do quality assurance as well to make sure even if you have the top grades that are qualified to be a Snapask tutor, we still need to do a continuous process to make sure you’re not doing inappropriate things, you’re acting according to our terms and conditions, and you’re providing good experience for our students. It’s both students’ rating to a tutor’s answers and also our very regular checkup with the individual accounts.

Jay: Right. So, after the question is answered, does the student rate the teacher thereafter, similar to Uber?

Tim: Yeah, exactly. They rate according to the experience they have given.

Jay: Has there ever been a time when say the tutor actually couldn’t answer the question? [Laughter]

Tim: That doesn’t happen a lot, mainly because the student provided the question—

Jay: Ahead of time.

Tim: Yes, ahead of time. Tutors get plenty of time to actually digest the question and figure out the solution. The most important thing we want to convey here is our tutors don’t provide just the answers. They stress a lot on how to elaborate the concept — how to apply the skills, how to help them to do it themselves. We’re not just helping students to copy homework. It’s not our intent.

Jay: Right. That’s the key because that would not be very helpful, actually, for the student either. Just by simply having the answer, they’re still not going to be able to pass their exam if they’re just given the answer.

Tim: Exactly.

Jay: I think it’s a great, great company. A great concept, Tim. I’m just thinking of the numbers that you told me earlier. $88 a month, U.S. Is that what you said?

Tim: $80 U.S.

Jay: Eighty. Eight, zero U.S. Okay. That’s very, very affordable. I just think about, we have like a Mandarin tutor. We have some tutors for my oldest daughter. She’s still very young, but we hire them. They’re already around $20, $30 U.S. an hour.

Tim: Yeah, exactly.

Jay: So, I can just imagine how affordable that would be especially if it’s, like you said, all-you-can-eat. I think that’s great. Okay. Cool. Talk to us a little bit about—now that you have a bunch of students. How many students do you have now on the platform and how many tutors?

Tim: We have a little bit over 600,000 users on the platform. For tutors, we have 120,000 qualified tutors.

Jay: Very nice.

Tim: Yeah. In the eight cities.

Jay: Is there any way that you guys are able to track success metrics of any sort. I know it’s difficult because it’s literally… But there must be a way that you can quantify, okay, a student is better because of after having used the app or not. I don’t know if you guys have figured out how to do that yet, but I would think that there has to be a way. Right?

Tim: That was a struggle in the very beginning where we are trying to measure how students are progressing and improving in school. We try to get results from schools. We try to ask them, “Hey, why don’t you provide the students’ learning progress report to us?” It’s not that teachers don’t want to provide but they simply don’t have these reports for us to analyze.

It’s more of a teacher who knows,this student is improving well, but I could tell maybe within the next couple of tests and exams, which is not real-time support for. If you know they’re not performing well on an exam, it’s already done, so you cannot do much to fix it. That’s why we understand this is a big gap between actual effective learning and what we have right now in the school system. So we do also actively measure how students are progressing within the app. Simply by…let’s say you ask the question in math that’s about linear equations. You must have challenges in understanding the problem and applying certain skills and concepts in this particular question. That’s why you ask the question. Right?

Because of this information, we would pinpoint that you must have some weaknesses or some concept that you don’t have too much understanding in it. That’s why we’ll push quizzes and relevant questions to you to test how you’re performing after the tutor helped you and illustrated the steps to you.

In this way, because, they gave us feedback by answering that question, so we know if they’re right or wrong which proves they have certain progress after a certain session. These are very tiny things that we’re doing in the product to measure success. Eventually, we realized we have built up a very big, gigantic database on students’ performance.

Now, we even package this data and help teachers — their own class teachers — to understand more about the student. “See, okay, I don’t just have a class of 40 students in Class 5B, but actually 40 individual students with individual profiles with their very specific weaknesses, and it’s all demonstrated on the report.” That’s the important thing and I think it can change education in a very wide range of things.

Jay: Absolutely. That sounds exciting because data is very important. Like you said, you could help the actual education system as well with that data.

Tim, just a couple of final questions for you. What are some of your goals — near-term and long-term goals for Snapask — maybe for the rest of the year and maybe for a couple of years out? You said you were in eight countries now around the region. Is that right?

Tim: Yeah. Eight countries right now. The long-term goal is pretty simple. We hope this way… I mean, our assumption and our way of solving the grand challenge in education by providing personalized learning method should be global. It should be solving a global challenge. Long term… Long-term is just three years. Right? We’re trying to grow us to as many countries as possible.

We have eight countries right now. We’re trying to grow to 25 countries by 2020. Within the next two years, the shorter-term goals will be to make sure we can work with schools, to make sure the frontline — the teachers — are actually spending time to understand their students instead of just marking homework and doing a lot of administration work, which is not the best use of their time and talent.

So, we want to make sure we do the work for them, make their life easier, by analyzing students’ weaknesses, visualize it, present it to teachers so they can focus on the pinpoint weaknesses for students and help them. These are the two things. Making that work, bigger reach, and a better way of helping students and educators.

Jay: What do you think, as someone that’s in the space of education, how do you think education is going to change in the longer term, say a decade out from now or even further? I mean, it’s obvious that traditional forms of education are losing, are changing because of the internet, because of accessibility of information. How do you see education? Is it going to still be in the same state as it’s in now, in 20 years from now? What are your thoughts?

Tim: I think the mode of delivering education will change, will definitely change, because, as you said, the accessibility, the variety of available tools out there… The way of conveying message and information will change, definitely. But, unfortunately, one thing that will never change and will be very difficult to change in the next decade will be how we measure success, how we measure success in education.

It was the same a couple of decades back. We’re still assessing students’ performance with the same rule: the exams. We’re still doing math, English, physics, all these subjects to actually balance students’ performance. Say, if you’re not good in physics you cannot be a physician or you cannot be a good university student. These are very disappointing things that I still have to accept. But, still, based on this, we can still do a lot to make sure students learn and get the best out of this current education system.

The thing is, the biggest change that is quite aspiring as an educator and in this field is that we could understand students much better because of their availability and the abundance of data. The data and all the notes that we can actually collect from students, make sure we understand and can provide the most personalized way of learning.

It’s just like if you are a patient. You want to know, what is the best way to treat me as a patient? And I want my doctor to know 100%, to know what exactly is the medicine that is suitable for me and what are my allergies. These are things exactly the same in education. You want to make sure teachers are not teaching the same thing to you and wasting your time, but instead, you should use your time in things that you have weaknesses, you have misunderstanding in. Right? I think that data can help educate us to do a much better job in the future, which is very inspiring. We intend to build the right tools for them.

Jay: Absolutely. Tim, two final questions for you. The second to the last one is, as an entrepreneur who has basically started a company on your own without any help and having gone through your startup/entrepreneurship journey, if you had one piece of advice to give to aspiring startup founders or aspiring entrepreneurs maybe trying to build a company here in Hong Kong, what would that piece of advice be?

Tim: In short, it’s just don’t quit. Don’t die. It’s very simple. Most people, like management books, tell you how to be a better CEO, how to be better in management and all that. But the truth is I think a lot of the cases are different. The same methodology could or could not apply in your situation. A lot of different factors and risks appear, but the only common thing is, if you quit, you cannot do it again. If you quit, the entire company is gone. You’re gone. The team is gone. The investors are gone. There is nothing left. I guess the single piece of most important advice — I do keep to my chest very close as well — is: don’t quit. Don’t stop.

Jay: That’s good advice. That’s great advice. I think every entrepreneur has felt the urge to quit multiple, multiple times, and it’s being able to embrace that and just plow forward is how you succeed. I think a lot of people, they actually don’t truly understand how difficult it is to be an entrepreneur and to build a 100-million-dollar company or a billion-dollar company or whatever. I think when you come out on the other side, what I hear, I haven’t done it, but when you come out the other side, you always hear about people saying how difficult it actually is. So, it’s almost worth it. If your company’s valued at X, Y, Z, then I’m pretty sure that you’ve worked pretty hard for that.

Tim, it’s been a pleasure. Thank you so much for coming on the show and sharing with us your exciting company that you’re working on. What’s the best place that people can find you, follow you, or connect with you, or if they want to learn a little bit more about what you’re working on?

Tim: If you’re a student or even if you’re not a student, you want to teach, you want to be involved in Snapask, just go and download Snapask on Play Store and Apple Store. If you want to know more about the company, one of the open positions, you can go to our website, Snapask.com. You can always find me on LinkedIn or Facebook. Always happy to talk.

Jay: Okay, awesome. Facebook, Linkedin, and we’ll get your links all linked up in the show notes. Thanks again, Tim. It’s been a pleasure and best of luck. I’ll be monitoring your progress very closely because, like I said, I have three kids of my own and I’ll probably be using your service very soon.