Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
Cracking the Engagement Code to Build an EdTech Behemoth | Anshuman Singh (Scaler)
345 Plays3 years ago
"Education is about addiction, not content."
This provocative insight from Anshuman Singh challenges the very foundation of modern education. He argues that the secret to effective learning isn't just delivering the best lectures or materials, but creating a highly engaging, motivating, and supportive ecosystem that gets students "addicted" to the process of self-improvement, leading to incredible real-world outcomes.
Anshuman Singh is the Co-founder of Scaler, an ed-tech platform valued at over $700 million. A former engineer and tech lead at Facebook , he helped build and scale foundational products like Facebook Chat and Messenger. Frustrated by the massive gap between academic knowledge and real-world industry skills he witnessed while hiring, Anshuman is now on a mission to forge over 1 million world-class engineers. To date, Scaler has raised over $76.5 million in funding to fuel its global ambitions.
Key Insights from the Conversation:
- From Content to Community: The journey from InterviewBit to Scaler was driven by the realization that content is a commodity. The real, defensible moat is a structured, high-touch ecosystem that provides live interaction, peer learning, and constant support.
- Mentorship as the Core Differentiator: The pivot to a paid, intensive program was based on the data-backed insight that personalized guidance from experienced industry mentors was the single biggest accelerator for student success.
- Engineering 'Addiction' for Education: Scaler’s product is not just a learning platform; it’s an engagement engine. It uses gamification, daily streaks, and community dynamics to keep students spending an average of 186 minutes per day learning.
- The Power of the Prepaid Model: Anshuman explains why they deliberately moved away from an Income-Sharing Agreement (ISA) model to a fully prepaid one. This attracted more motivated and committed professionals, leading to significantly better outcomes and a 2.7x average salary increase for graduates.
- Building a Brand on Tangible Outcomes: Scaler's growth and ability to raise capital are built on trust generated by delivering tangible, life-changing results for its students, creating a powerful referral engine that now accounts for over 30% of new enrollments.
YouTube Chapters
[00:00] - Introduction
[01:51] - How a Global Programming Competition Got Me Into Facebook
[04:32] - Why I Chose High-Risk, High-Growth Facebook Over Google
[14:22] - The Inside Story of Building Facebook Chat & Messenger From Scratch
[20:26] - The "Why": Leaving Facebook to Fix India's Broken Tech Education
[25:06] - Phase 1: Building a Free Tool (InterviewBit) & Our First Business Model
[33:25] - The "Aha!" Moment: Discovering the Real Secret to Effective Learning
[40:15] - The "Online Stanford" Playbook: How We Re-engineered Education
[55:18] - Why We Killed Our 'Pay After Placement' Model & Pivoted to Prepaid
[1:08:03] - Scaler Today: The Path to $100M ARR & Replacing Universities
Hashtags for YouTube
#FounderThesis #StartupPodcast #AnshumanSingh #Scaler #EdTech #StartupIndia #Entrepreneurship #BuildingaBusiness #FounderStory #ProductMarketFit #Pivoting #Education #FutureOfWork #TechSkills #VentureCapital
Recommended
Transcript
Introduction to Founder Thesis Podcast
00:00:03
Speaker
Hi, I'm Akshay. Hi, this is Aurob. And you are listening to the Founder Thesis Podcast. We meet some of the most celebrated sort of founders in the country. And we want to learn how to build a unicorn.
Facebook's Impact on Scaler Academy's Approach
00:00:26
Speaker
Hi, I'm Anshu Bansing. I'm one of the co-founders at Scaler Academy.
00:00:30
Speaker
Everybody knows the reason why a company like Facebook slash Meta is worth hundreds of billions of dollars and the reason for that is that it has cracked engagement. It knows how to keep users engaged and increase the amount of time spent on their various platforms. Even though that engagement has a dark side to it, things like hate speech are outcomes of that engagement.
00:00:51
Speaker
Now imagine if you took this superpower of engagement and used it for something like education. Imagine building a platform which keeps students engaged and therefore delivers better learning outcomes.
Evolution and Mission of Scaler Academy
00:01:05
Speaker
And this is exactly what Scaler Academy has done. Founded by an ex-Facebook product leader Anshuman Singh.
00:01:12
Speaker
Scalar Academy started life as an interview coaching platform and then through a number of insights which Anshuman had while trying to solve the basic problem of improving talent quality in India, they ended up with what today is like an online version of Stanford. Something which gives the highest quality of education at affordable prices to thousands of students.
00:01:36
Speaker
Scaler Academy has proved outcomes with their students getting significantly better job opportunities after completing courses in software engineering and data science with Scaler. Here's Anshuman talking about his fascinating journey, starting with how he got into
Anshuman Singh's Journey to Facebook
00:01:51
Speaker
Facebook. How did you get into Facebook? I mean, you're not from like an IIT, like those typical hunting grounds for Facebook and Google kind of companies. How did you get into it?
00:02:04
Speaker
So back in those days, Facebook or Google mostly wouldn't hire from India for the US office. So it was very unconventional. This is back in 2009, 2010. In fact, like recession had had just gotten over.
00:02:22
Speaker
And this is, by the way, just by chance. I used to, as I said, I used to like solving really hard problems. And one of the places where I got to solve very hard problems was these programming competitions that would happen all over the world. So-called hackathons.
00:02:39
Speaker
Yeah, it wasn't called Hackathon back then. Today, the more popular term is Hackathon. But these contests, and one of them was called ACM ICPC, which is almost a 40-year-old contest framework. It's termed as the Olympics of competitive programming. Every single college in the world sort of participates in that.
00:03:04
Speaker
So there I was again very fortunate to win on India level and hence got to represent India in the world finals where they invite top 100 teams from the world.
Choosing Facebook Over Google: A Personal Decision
00:03:17
Speaker
So I did that twice. I did that in my third year and then once in my fourth year. So then I went, the fourth year thing happened in Harbin, China, which is
00:03:28
Speaker
which is a freezing place. In fact, we built out an ice city there. So there I met a few recruiters from Google and Facebook, both from their headquarters. They were not supposed to be there, but they were just gate-crashing the party.
00:03:49
Speaker
There they invited me to interview with Facebook and Google. Obviously, I, being the kid I was, I was very, very excited. I took up those opportunities and ended up doing well in the interviews and hence got the offer. You got from both, is it? Yeah, Google Mountain View and Facebook and Palo Alto, both.
00:04:11
Speaker
If you like to hear stories of founders then we have tons of great stories from entrepreneurs who have built billion dollar businesses. Just search for the founder thesis podcast on any audio streaming app like Spotify, Ghana, Apple Podcasts and subscribe to the show.
00:04:32
Speaker
So why did you choose Facebook or Google? So my parents did not want me to go to
Facebook's International Expansion
00:04:37
Speaker
Facebook. They wanted me to go to Google, given it was a more established company by then. And everyone had heard of Google. It's a verb, basically. Exactly.
00:04:50
Speaker
Facebook at that point of time was a very high risk company because in India Orkut had launched and Orkut had not done well. So there was this apprehension as well like you'd go to Facebook but then Facebook would crash and then you'll be left stranded in the US.
00:05:09
Speaker
But when I talked to people, people who were far more senior to me and I considered them to be my mentors of sorts, few things became clearer. One was Facebook was not such a large company where it would be difficult for me to make impact.
00:05:27
Speaker
given I like problem solving, I like to make impact. It was the right scale where I could go and join teams and actually build things that would go out to the audience for them to use at scale. The second thing that, again, came out was that, look, Google is already a very large company. The probability that Google becomes 10x of what it is is relatively very low. It has become 10x, but it is relatively low.
00:05:55
Speaker
However, Facebook is a very small company right now for the kind of company it is. And while there is a high risk that it might go down, there is also the other side where it can become massively, massively huge. And people in Silicon Valley keep looking for companies like these. If you're in your life at one company that has made it big, then you're sorted for life from a financial standpoint.
00:06:22
Speaker
And then that's an opportunity not everybody gets. So when I had those conversations again, I mean, I was still out of college. I did not know much about life. But I think some of those arguments made a lot of sense to me. So I ended up saying no to Google and ended up joining Facebook.
00:06:41
Speaker
To be honest, it's a decision that I still like to this one. If I were to go back and take this decision again, in a heartbeat, I would still choose Facebook. Initially, they had that only an .edu ID would let you sign up. Had it grown beyond that .edu stage, was it open to everyone? What was Facebook like
Competitive Edge: Newsfeed and Growth Teams
00:07:06
Speaker
when you joined it?
00:07:06
Speaker
Facebook had grown beyond that stage. Facebook was in that stage where they were expanding massively internationally. In fact, they were about to launch in India. And I remember the first video that I saw, before they would even launch in India, there was this video that had gotten viral, which was just like how Facebook has become so addictive, etc.
00:07:30
Speaker
And I looked at the video and I was like, this is Orkut. Why is there another company called Facebook doing exactly what Orkut is doing? But then later on, once I went into details of it, then I could make sense why Facebook has succeeded where Orkut has failed. Tell me that. What is the difference? I think the primary.
00:07:50
Speaker
The primary reason was to, one, the newsfeed. The newsfeed was to be honest a major winning point for Facebook. It made consumption and engagement really really high. The second thing that actually
00:08:09
Speaker
that stood out for Facebook was their growth team. So growth team as such does not make sense but Facebook had this growth team which would take up these problem statements and come up with very innovative solutions and that would be state of heart. For example, Facebook had this very small widget on the side where it would say people you may know. This would encourage people to keep adding more and more friends. And the wider your social network, the more posts you would see.
00:08:37
Speaker
The one thing that Orkut actually did not do really well was helping people expand their network. And that team was again a game changer. Like people you may know was a game changer because now you did not have people with just 10 friends or 20 friends. Everybody had like hundreds of friends and hence you had a lot of content to consume. So those two were major differentiators back then and a bunch of things happened after that. Was it a culture shock for you to come to US?
Cultural Challenges in the US
00:09:02
Speaker
from India, small town India to US. Yes, I mean, multiple things. One is you were adjusting to this new place, the accent was very different.
00:09:20
Speaker
It is one thing to listen to that accent in movies and TV shows, and the other is there's a person right in front of you talking really fast in that accent, and then you have to make sense of it. I remember this one incident. So essentially when you go there, you can't actually commute using trains, et cetera. You actually have to get a car. So I remember we had rented a car.
00:09:47
Speaker
And we were going somewhere. Our headlights were not on. So we were pulled to the side by this cop. Now, when you pull to the side by the cop, you're already very, very scared. And this person comes to the window and starts saying something really fast. I couldn't make sense of it. And I said, I'm sorry. Can you please repeat? He again said the same thing really, really fast. I again.
00:10:17
Speaker
I again asked him please can you repeat and then the third time he got really irritated and he said the same thing in a very very angry tone. Again I couldn't make sense of it but because I did not want to say that please repeat I just randomly said yes.
00:10:39
Speaker
He went to his car and he wrote a ticket and then he gave me a ticket. Later on I figured out like he was asking me whether I was aware, I was speeding and whether I was aware. I had my headlights on. And since I did not understand him, I ended up saying yes. I mean, those kinds of incidents that happened, things that you look back to now and then you can laugh at it.
00:11:00
Speaker
But the other thing is US is also one place where, unlike India, you don't have household help. You have to do your cleaning. You have to cook food, et cetera, et cetera. It's a bad thing about me. I've never cooked for myself in my life. I've never been a person who has been great at maintaining homes either. So those things also come into the picture.
00:11:30
Speaker
You start missing Indian food. That comes into the picture. I ate a lot of bland food. I mean, to be honest, the food in US and Europe both is very, very bland as compared to Indian food. And it takes a while to get used to it.
00:11:49
Speaker
So while Facebook had free food inside the campus, I remember we would eat once in Facebook and then we would go figure out an Indian restaurant. We would go to that restaurant, eat and then come back home.
00:12:03
Speaker
So you grew up in like very, very small towns. I'm sure you wouldn't have been very fluent in English, you know, in terms of your communication skills and all. So how did you navigate that? I mean, today when I speak to you, you're obviously very articulate and fluent, but I'm sure growing up that would not have been the case.
00:12:22
Speaker
Right, right. So that came with the initial anxiety of fitting in. So I remember when I was in college, whenever I would try to attempt speaking English, I would speak really fast. And that happens. When you have a bunch of things in your mind, you just want to blurt it out. So instead of focusing on whether the other person is able to understand me, you find the right words and you just mix it and you blurt it out.
00:12:53
Speaker
That's the first thing that happened to me in the US. The moment I would say things really fast, people would give me a very puzzled look ahead.
00:13:01
Speaker
And then it would become really clear that I'm making no sense to them. In India, people would still understand because my accent and that combined would still make sense to them. There, people would not understand me at all. So hence, the first exercise that I did was I would consciously think about things, talk about them in a tone which people would be able to understand.
00:13:25
Speaker
not the accent, but just the pace of talking. It took some time, but then eventually I got used to it. Plus the thing is Silicon Valley, where I was, it's a place where you can find all kinds of people. It's just a melting pot of you have people from Asian descent, you have people from Europe, you have people from Africa, you have people from US. When you talk to different people with very different accent, then you start to empathize.
00:13:52
Speaker
I mean, the same way I am finding it hard to understand their accent, they find my accent also hard to understand. So I have to make that extra effort to make sure I break it down, I slow it down so that they can understand those words. Even if maybe one or two words doesn't make sense, they start to understand it. So that happened for about six months or so. But after conscious effort,
00:14:17
Speaker
It became normal. And so which team did you join in Facebook? What were you doing in those years that you spent at Facebook? I joined
Developing Facebook's Chat and Messenger
00:14:28
Speaker
the Facebook messages team. So back then, what had happened was that Facebook did not have chat to begin with. Facebook was just this website which would have newsfeed. And you could go and post on people's wall exactly like Orkut, right? You could write set up books.
00:14:48
Speaker
One of the things that Facebook used to do was they would do these hackathons, which were real hackathons, which were like these two-day events on the weekends where you could stay back in the office. There would be a lot of pizzas and a lot of food, etc. Stay back in the office, build anything you want, anything you want.
00:15:06
Speaker
Some people would just go and make paintings, some people would just go and create these sculptures inside the office. I remember there was this team that went and painted the streets and then cops had come to the office. But you could also build software. One of the things that came out from that hackathon was Facebook chat. Somebody had just put together a very rough chat app.
00:15:29
Speaker
and and demoed it internally and the team said okay let's let's launch it let's see what happens so this this chat which would never store messages it would be very basic let's say you are online if you are online then you'll get my message if you're not online you will lose that message
00:15:47
Speaker
it was so broken but even after that this chat started seeing almost a billion messages every day which was mind-boggling and it was very anti what the thesis that Facebook had that people would not chat a lot on Facebook
00:16:04
Speaker
It proved that wrong. Hence, the next effort was how do we build a chat plus messages thing, which is actually really solid, has a large infrastructure in the back, does store messages. Even if you're offline, you still get those messages so that people can then use that as a communication channel as well. That project when it was starting, it was that project that had joined when it was just getting started.
00:16:34
Speaker
Okay. And you spent the entire four years on the chat product, or like, you know, what was your path at Facebook? So the first three years were on this chat product. And this was the time. The early coding or did you also get into like product?
00:16:52
Speaker
Yeah, so I mean, initially, I was purely coding. Then later on, I was, again, very fortunate that I moved up the ranks fast enough. So then I started leading certain efforts. So one of the efforts that I led was the launch of the Facebook Messenger app. The second effort that I led was, so we were competing against WhatsApp back then. WhatsApp and Facebook were different companies.
00:17:20
Speaker
One of the things that we realized was that in certain markets, Messenger was losing because it was not as fast and reliable as WhatsApp was. The other effort that I led was just the reliability and the latency for the app, just creating things around that.
00:17:41
Speaker
And then the acquisition of WhatsApp happened, so a lot of the war-like thing that was going on that came to rest and we were obviously the winners given the consolidation that happened.
00:17:57
Speaker
My last year, which is like the fourth year at Facebook, was primarily spent on setting up this new office for Facebook, new engineering office, which was the first engineering office outside of the US, which was the Facebook London office.
00:18:12
Speaker
It still continues to hire massively from India. So hiring the first 100 engineers there, putting over a bunch of projects. I was working as part of this Facebook pages team that would make sure there are no duplicate pages. And every, let's say, entity, let's say a movie or a music, et cetera, has a page that people can go and like and then recommend to other people.
00:18:39
Speaker
So I was doing a lot of that. This was like quite a leadership role. Were you like talking to Mark also during this time? Not so much with Mark. It was Mike Shrep. He was the CTO of Facebook. So a lot of that communication would happen to maybe Mike Shrep and that was actually not very frequent.
00:19:06
Speaker
But mostly to business leaders in various domains just to figure out what project can be ported over to the London office. We were very fortunate that we already had a few leaders who had come and joined the Facebook London office when it was being set up. So we had enough projects and hence we could go and do hiring for that.
00:19:30
Speaker
without actually depending on putting more and more projects in the beginning. Why did you choose this role? It's like a very different role from your previous role of more on tech and product.
00:19:43
Speaker
Right, right, right. One of the reasons was that it,
Leadership and Expansion at Facebook
00:19:50
Speaker
a couple of reasons. One is this was a very different kind of role. So I wanted to maybe do this just to figure out what kind of challenges would come in this kind of a role. The second is it was a different location as well. So I felt that being in the US at least
00:20:05
Speaker
for the first couple of years had a large learning curve to it because of the cultural differences. So I felt like maybe being at this new place would again have a new learning curve which might just lead to higher personal development. So those two combined prompted me to take that role. Okay. So what made you finally decide to leave Facebook?
00:20:31
Speaker
Yeah, so I keep saying that story over and over again. But one of the things that prompted me to do that was, in general, realizing that this one stream of education which has had a very large role to play in my life, which was just tech education, that in most parts is really broken for the Indian ecosystem.
00:20:58
Speaker
I had my own brothers studying there, and I could see the kind of education happening in colleges is really, really, really bad. Now, maybe why is that really bad? I can go into one some other day. There are so many problems there, but... There is bookish, basically.
Inspiration Behind Scaler Academy's Mission
00:21:20
Speaker
Bookish plus there is no incentive to do a lot more. A lot of these colleges are run by politicians. They care more about other aspects of college rather than the educational outcome of people. Most colleges are actually just making sure that AICTE has set certain regulations. We are following those regulations. Nobody can come in tomorrow, take away our certificate. Some of the core things or core innovations that would
00:21:47
Speaker
that is expected from a college to make sure there is higher educational outcome, that is never on their radar. So those are some inherent reasons, but that being said, what is happening as an outcome was
00:22:02
Speaker
In that engineering, even though it was a very, very large vertical in a country like India, it did not have quality colleges slash universities. Hence, the entire IIT versus non-IIT piece. That's something that I felt very strongly about.
00:22:24
Speaker
my other co-founder Abhimanyu, he comes from a family which had very educational roots. So he also felt very strongly about that. So hence we both thought that is one field where we believe we have native expertise. We did not have the answer of how, but we felt we should figure out a solution to that. And the thesis was, if we would be at this problem for long enough, we will eventually figure out some solution.
00:22:54
Speaker
One more belief was we are not very dumb. Both of us are simply smart. What is Abhimanyu's background and how did the two of you meet?
00:23:05
Speaker
Avimanyu was my batchmate in college, same batch. In fact, even in college, apart from competing in programming competitions, we would keep building these random products. In second year, I remember we had built this product which would help you control your electricity
00:23:25
Speaker
electrical appliances from your laptop and it would monitor the user's pattern and raise a flag if let's say certain electrical appliances were working outside of the regular pattern.
00:23:38
Speaker
So we had done some bit of that. We had also built out this Facebook group kind of a thing in college, which was anonymous. And we very soon figured out why should you not build an anonymous network? People keep posting very offensive things there. So we shut it out.
00:23:57
Speaker
But because we had worked together on building things, there was a lot of comfort. He was working with this company called fab.com. Fab was an e-commerce company from New York, had grown massively. They were, I think, the quickest company to become a unicorn, if I'm not wrong. Now, I think that is with some other company. Mensa is probably the quickest company to become a unicorn now.
00:24:25
Speaker
But in that time, they were the fastest company to become a unicorn. So he was with them. He wasn't thinking of leaving, but when I had pitched this that I'm thinking of looking at education, then he was also very excited by that. And then he said, let's figure out what we can do.
00:24:47
Speaker
So I mean, while at Facebook, when you were employing people, you figured out that there is an employability issue in graduates from India. And that's when you started chatting. And OK. OK. So then tell me that like that journey from idea to launch.
Scaler Academy's Gamified Learning Approach
00:25:06
Speaker
Right. So in 2014 and our thesis was that maybe the right way to solve for education is through content and product.
00:25:15
Speaker
make free content and have as many people consume it.
00:25:21
Speaker
One of the... And technical content like on... Yeah, how to code, exactly. So one of the products that we were really inspired by was Duolingo. So Duolingo is this app you use to learn languages, but it is very gamified and hence has very high engagement. And we felt like if we could build a Duolingo for tech, I mean, that should solve for education.
00:25:46
Speaker
So, hence we went ahead and did that. So, mid of 2015, we launched, we picked one new space. You had quit your jobs and like gone into this full time. Yeah, one thing which we were very clear was that there is no such thing as committing part-time. Either you do it or you don't do it. So, if we want to solve this problem, we have to commit 100%. So, hence we had left our jobs. And you had that comfort of the savings from those probably some ESOPs and stuff like that also too.
00:26:15
Speaker
Yeah, all ESOPs, to be honest, all of my savings came from stocks. Hence, I keep telling people, please, please value your ESOPs if you're at a company that makes it big. I mean, ESOPs are probably going to be 10x your earnings anyway.
00:26:33
Speaker
So anyway, I had a lot of financial cushion because of Facebook. Abhimanyu also had a bunch of financial cushion because of fab. So hence it was an easy decision for us. Also, it was very conscious because since we had financial cushion, we knew that at least for the next five, six years, we can survive without taking salaries and we wouldn't have to take a decision which is good for short term, but bad for the long term.
00:27:02
Speaker
So hence we had said, let's leave our jobs, let's focus on this full 100%. So we started with the dwelling approach. We couldn't obviously take it so wide, you can't cover all of tech in one go. So we picked one specific niche, which was, which probably people have the highest amount of motivation for is jobs in tech. So we said, like, if you're preparing for a job in tech, I'll take you from this,
00:27:29
Speaker
zero journey to the final 100 journey. I have prepared this Duolingo for that, and you come to me, and then I'll take you through that. This was to crack the selection process, basically. Yeah, yeah, yeah, yeah. So I mean, cracking the selection process, I mean, there are a certain number of concepts that you need to learn and understand, which is also important, to be honest, to be a good software developer. Interviews are a reflection of what you need to be successful at jobs.
00:27:59
Speaker
So we have built something around that. Over the next three, four years, we needed a business model as well. So the natural thing was that you have this website where a lot of people are coming and practicing. Why don't we go and talk to companies and we start charging them money for finding people here. So this becomes the ground for hiring tech talent.
00:28:27
Speaker
And which by the way worked out really well. We signed up more than 500 companies. We were doing fairly decent revenues as well. You were charging like the standard agency fees like one month salary. That was the bottle.
00:28:44
Speaker
We initially started to build a SaaS where we wouldn't charge them for hire, but rather for getting access to the pool. What we realized was that India is a market, especially in the hiring space, where it doesn't value SaaS. It values outcomes a lot more. In fact, it's very service-driven. That was a learning over a period of time. Hence, we pivoted and we said, let's do what the market wants.
00:29:12
Speaker
So we built out this separate vertical internally that would work with companies and help them use the tool, push them to use the tool to actually finally find the hires. And when they would find a hire, then they would pay us exactly like an agency.
00:29:28
Speaker
Okay. And you would provide like that replacement guarantee and that whole standard agency? Yes. Yes. Yes. So we would do all of that. So which is very low tech kind of a business. This is very high touch business. Exactly. It was very operationally heavy. While we had this added advantage of access to this pool that is that was hanging out on interview bit while we had data saying that look, I mean, this is what
00:29:58
Speaker
somebody has done and hence has a higher probability of being in the right fit, not just the resume, which would help us with the selection ratios. And hence we could get more done with less number of people as compared to a regular agency. But still it was very operationally heavy. It was not really the kind of thing that we had done in the past. It was a North heavy thing which we also learned as we were executing.
00:30:29
Speaker
And probably you would have also needed to use like Nokri and others to really fill up client mandates because not all mandates would get filled from your pool. Right. So we intentionally try to stay away from that because I mean,
00:30:46
Speaker
If we ever like would fall back to not create etc, then the thing is then you just then actually actually actually competing with agencies with the same same tools and armory, right? So then you know in no way better. And in that case, then you start playing on on margins and a bunch of other things. So.
00:31:05
Speaker
We said our superpower is that we have access to people that we don't need to pay for. We have access to their capabilities that we have seen through by way of coding. Let's use those superpowers because then a single person can probably get done the same amount of work which let's say five or ten people at a regular agency can and then that becomes your margin strength more or less.
00:31:31
Speaker
And because this was like, this must have been a kind of business that doesn't need funding as such, right? Because you were like, I assume cash flows would have been decent enough too. Right, right, right. So this kind of business actually by default is profitable right from the word go. So hence did not need to raise funds. But in 2018 actually, we were thinking that look, I mean, this is not why we started.
Refining Scaler's Mission: Quality Over Placements
00:32:00
Speaker
If we were supposed to run this, we were better at our jobs. And it's not really solving the problem that we wanted to solve. The problem isn't that there are a lot of jobs and people can't find those jobs. The problem is that there are a lot of jobs, but people are not capable to fit in those jobs. So the real problem to solve is the talent. How do we go and create more talent?
00:32:26
Speaker
And there the question arose that if the thesis was that a Duolingo kind of a thing, free content can solve it, why are we still struggling with finding the right talent? That problem should have been solved by now. And we realized by looking at data that
00:32:45
Speaker
Product and content alone are not enough. We were actually, I mean, interview been evolved to become a supplement to a lot of people in tier one colleges. Most tier two, tier three colleges were still scared. They would come, they would maybe try out a few problems and then they would leave because they didn't have the right support in general. It was too high level for them.
00:33:08
Speaker
I mean, they would start slow. But then as things started to get complex, in a real college, you still had your peers, your seniors who would help you out.
Mentorship Model and Hiring Insights
00:33:18
Speaker
There was nobody for people who did not come from that environment and hence they would drop off. In fact, there's something very interesting that happened in 2018. So in 2018, there were a bunch of companies that were driving their entire campus hiring also from interview bit.
00:33:34
Speaker
and hence we got to see the funnel, you know, like who was applying, who was getting to round one, round two and so forth.
00:33:43
Speaker
Most people, a lot of high paying companies hired. In fact, one of those companies was Uber. Uber is, by the way, the company which is in the news for that two CR salary in the US from the IITs. In India, they also pay really well. We looked at the funnel from Uber. Everybody, most hires came from IITs and IITs and NITs.
00:34:09
Speaker
But one thing, one aberration that we noticed was there were few people who came from colleges that I had not heard of. And I am, by the way, I don't know about every single college in India, but like a bunch of colleges that I had not heard of. And then there were people from those colleges who had made it to Uber. So what we did was that we felt maybe if we talk to these people, they will give us insights on what is the real way to solve for education. Because these are the people who have been in this non-supportive environment and still made it big.
00:34:40
Speaker
So we picked a sample of 100 people, 100 people who did not come from the top fancy colleges, but rather what you would call as three colleges and still made it to moon shot companies. So we talked to them, all of them very hard working, very self motivated.
00:35:00
Speaker
One very surprising fact that we came to know with 98 out of those 100 people, so except for to every single person, had either an elder cousin or elder bhiyadili or a very close senior who was already working at a good company. Okay. So, I mean, there was and it became very evident when we were talking to them, they they one they knew it is possible
00:35:28
Speaker
I mean, they had hope, they had aspiration and they knew it is possible because my brother has done it or my elder cousin has done it. Second thing is they had guidance and they had support. I mean, every time they would get stuck, they would just call up this elder cousin and say that, hey, look, I'm stuck. Can you please help me get unblocked?
00:35:49
Speaker
So that was such a high correlation that that became more or less an eye opener. It was then that we realized that education is, to be honest, not about content. That's a very common misconception all of us have. We feel education is about content and hence you'll see that everybody's advertising, look, my content is best. My content is best. My content is best. It's not about content. Content is hygiene. Content has to be good. It is hygiene.
00:36:15
Speaker
In fact, what we also realized that day was that what we had built till now was a very good book. I mean, you have a lot of books, some books are great, some books are not so great. We had built a very good book. But book alone, again, does not make sure that people learn and then succeed. I mean, I have hundreds of books at my home. I don't read all of the books.
00:36:38
Speaker
And hence, like all of your Coursera, Udemy, all of these places, they are all books. They are recorded courses, they are books. Hence, they don't have very high completion rate because books is content. Content, one very common mistake people make is they believe content is what wins you the education game.
Engagement in Education: Lessons from Stanford
00:36:59
Speaker
Content does not win you the education game.
00:37:01
Speaker
Content is hygiene. That day it became clear to us, education is about cracking engagement and motivation and aspiration. If you can get people hooked onto the learning thing, they are coming and spending five hours every day or three hours every day on your platform learning things. They are motivated because they believe they can achieve something. Then you have cracked education. Education is not about content, it is about engagement and addiction.
00:37:27
Speaker
That was a big realization we had then. And engagement and addiction comes from various ways. One is, I keep saying support ecosystem. Support ecosystem is the ecosystem that pushes you to come back. That makes sure that when you're falling down, they get you back up on your feet. Which in a college, for example, if you look at IIT Bombay, this is a further reflection that we did. Why is IIT Bombay?
00:37:54
Speaker
Is it because it has the best content? No, actually, ID Bombay curriculum has not been updated for a while. It's probably, yeah, curriculum is not too different from tier one, tier two colleges. Maybe tier three might be different, but tier one and two have similar curriculum. Actually, tier three also have very similar curriculum. So the curriculum isn't the winner. In fact, like if you start comparing ID Bombay's curriculum with a lot of world-class universities, a lot of them have much, much better curriculum.
00:38:24
Speaker
Is it ID Bombay? ID Bombay, because it has great teachers. I mean, just like quotas, if you look at quota, they would advertise the teacher. Yeah, superstar teachers. Jithubaya is teaching, this bhaya is teaching, that bhaya is teaching. That's why you should come and study at us. But one of the best institutes in the country,
00:38:44
Speaker
never advertises about his professors. And to be honest, those professors also know this respect meant to them, but they're probably not that great teachers anyway. If you go and talk to graduates from IT Bombay, and I had a few of them with me when I was in the US, a lot of them would crib about a lot of professors.
00:39:06
Speaker
Their way of teaching wasn't very engaging. So then what makes IT Bombay an IT Bombay? Why is IT Bombay an IT Bombay? One is obviously it has a lot of smart people. The peer group. But those smart people, the peer group. But just because you have 10 smart people, are you going to become an IT Bombay? Not really. It's actually the ecosystem.
00:39:26
Speaker
ID Bombay has cracked the engagement game, engagement and addiction game. You're constantly being pushed by your peers, your seniors. You know when you get stuck, you can just knock at the door right opposite yours and there's a senior living there and you can ask them, Sir, can you please help me out with this? I'm getting stuck here. You know you're in the vicinity of these people. The ecosystem makes it work.
00:39:50
Speaker
That was the realization in the end of 2018 that helped us create scale up, that education is not about content. Content has to be good. It is hygiene. But the real source or the secret source of education is engagement, addiction, and the ecosystem that you create.
00:40:10
Speaker
And hence, if you have to really solve for education, you have to solve for these pieces. I want to know how you executed this thing. I understand the insight behind it. So essentially, this insight would mean that you need to have a cohort-based faculty-led, something which mimics the campus experience.
00:40:38
Speaker
Exactly. Exactly. So we said, let's actually break that down. What exactly is IT Bombay doing? In fact, IT Bombay was not the metric for us. We said, let's pick the best university out there, which is Stanford. And we said, what does it take to create an online Stanford? By online Stanford, we don't mean that we want to create a university and that should give degree, et cetera, et cetera.
00:41:06
Speaker
degrees don't mean jack shit. It's essentially what people who are graduating, what they can do, that's what is the core sense of that. So we started breaking it down. In fact, we talked to a bunch of graduates from Stanford to figure out what was so different about them. Maybe I can talk about what content we borrowed from them. But the breakdown essentially resulted in the following. One is when people are studying content, that is
00:41:36
Speaker
that is challenging, then it is important to give them the right support so that they don't drop off. The very common human tendencies as things become tough, let's just give up or let's just delay that. If you have the right support ecosystem, then you carry on. The second thing is having a bunch of push items to make sure, so one of the reasons why people don't read books is because there is no time limit to that.
00:42:03
Speaker
You can always keep postponing. You can say that I'll read the book maybe two weeks down the line, three weeks down the line, or maybe I'll read the book tomorrow, but the tomorrow never comes. So what is this synchronous thing that keeps pushing you to do things, which if you don't do, you feel ashamed about? We feel ashamed about not doing things in college because I see other people doing it. I know that.
00:42:28
Speaker
There is a deadline. I know my professor is going to ask me, there is going to be an exam and hence I have to prepare. I can't push it. It is the same for everybody. Hence the cohort came into the picture. For support, we actually introduced the role of somebody called a teaching assistant, who was supposed to be your senior, who you could ever go at any point of time and ask doubts. This person could clear your doubts. Because that entire Bhaya Didi and elder cousin insight came from that Uber experience.
00:42:57
Speaker
We said given that has such a high correlation, we should make sure that every single person gets a personal mentor, who is exactly like your elder brother, probably talking to you once every two weeks, one-on-one, and is actually a successful software engineer or a successful decade. It comes from the same domain. It's not just some random person that made your mentor.
Real-World Skills Through Practitioner Faculty
00:43:21
Speaker
And most importantly, you have still have the faculties, faculties also like that we felt like things that are broken is that in our colleges, when it comes to tech, there are usually professors teaching and professors mostly haven't written code themselves for the last 10, 20 years. They're not in touch with what is happening in the outside world.
00:43:39
Speaker
But if you look at medical science, in medical science, doctors are practitioners. They practice and then they teach. Hence, they are able to give you very practical examples that look pure as what goes wrong if you do what you're suggesting to do. One other conscious choice was that faculty should be people who are currently working at companies.
00:44:04
Speaker
maybe building things, just like I was building things. In fact, I am still one of the teachers. I take a class almost every day, a three-hour class almost every day. Even today, yesterday night, I had a class. So that became one of the core designs that people teaching should be practitioners.
00:44:29
Speaker
And eventually, the most important piece, the most important pillar that we figured should happen is we should get people to talk to each other a lot. IITs are IITs because of the PR ecosystem. PR learning is the strongest portion. And the networking helps you even after you've graduated. And to be honest, that's a very hard problem statement in a virtual sector.
00:45:00
Speaker
So we have a few dedicated product teams, product and community teams, where their entire key result, KR, that they go after is. How many times are people talking to each other in a day? So if I have 1,000 people, then how many peer-to-peer interactions have happened? And when people graduate, we do a qualitative study to figure out if I ask a person, how many people are you really close to from your badge?
00:45:29
Speaker
Then do I get at least three people that they relate very, very closely to? Which is like, do they graduate with three best friends almost from this cohort or not? So those kind of things we started prioritizing. That became one very core pillar of scalar. So we brought all of this together.
00:45:49
Speaker
In fact, there is another thing that we do. This is not a core pillar, but just something we blatantly copied from Stanford. Stanford does this view from the top where they invite industry leaders to come and talk about what is happening in the outside world. So every month, every month or every two months, we get one person who is an industry leader, let's say a CTO of a large company or a CEO of a large company, to come and talk to the cohort. Just tell them about what is important in the outside world versus what is not.
00:46:17
Speaker
how to and in fact prioritize, right? And then just open up, make these people ask questions, get their questions answered. There are a bunch of these small, small choices inside that I have not mentioned. I've only mentioned the larger design choices and there's been working terrifically well till this point.
00:46:39
Speaker
No, tell me like the evolution, you know, when you launched the first cohort, what was that experience and what was the product like in that stage? And how did you get in place this one on one mentorship? And, you know, how did you get your faculty in place? Like tell me about that cohort one and you know, from there, what it is today like that evolution.
00:47:03
Speaker
So when we started, the one good thing that we had as a result of the work we had done in the past is the InterviewBit platform. So InterviewBit platform by then had grown to almost 2 million software developers. And since it was operational from 2015, there were a lot of people who had benefited from the platform. They held it in very high regard.
00:47:25
Speaker
So that community helped us get a lot of mentors, DAs, et cetera, because when we reached out to them saying that, hey, now we need your help to give back to the ecosystem, everybody was super excited about giving back to the ecosystem.
00:47:41
Speaker
I mean, we've filtered away some credentials, that's not the best kind of filtering, but at least like the credentials help to give us some confidence that, okay, find the mentors and the TAs have some amount of capability. No, credentials matter to the cohort, no, I'm assuming that they would respect. Yeah, even the cohort actually respects the credentials a lot more.
00:48:03
Speaker
But we filtered based on that. The teachers, to begin with, we started with just me and there was a friend who was running his own educational company. So we had requested that person to actually come and help us out for six months. So there were only two teachers in the beginning. With the understanding that if we ever had a problem, we have the entire team here, somebody or the other would step up and then also try to teach.
00:48:30
Speaker
On the side of product, I mean on the product, because for Interview Pit, we had built a lot of these gamification features. So we just translated that to scalar in the beginning. And then we started trading on scalar to make some of the, let's say the mentor interactions, TA interactions, et cetera, as a product. And the chat, did you have that in place? Like chat between cohort?
00:48:56
Speaker
For that, there are these solutions. Instead of building, we bought these solutions like Slack, Flock. We initially started with Flock.
00:49:09
Speaker
And now today we use Slack. So we just set up Flock for everybody. We call it scale at chat. And we say everybody has to be here chatting to everybody. We would do a bunch of events or competitions there so that people get accustomed to using Flock. But then later on now, I mean, now that I look at the evolution,
00:49:33
Speaker
A lot of interactions happen on Slack, but a lot of interactions also happen on WhatsApp. A bunch of interactions happen on the website. So we've split the community piece into a collection of a suite of products, which is across multiple platforms. So Slack, WhatsApp, and our core product. Which makes real life, because in real life also you...
00:49:59
Speaker
Exactly. We're also looking at some physical element, but I'll maybe talk about that in six months from now. What was the pricing for cohort 1 and what technology were you focusing on? Was it like a full stack development program or what was it like?
00:50:18
Speaker
On the technology piece, even when we were starting, the one core thesis was that tech is an ecosystem that keeps
Curriculum Focus: Problem-Solving Skills
00:50:29
Speaker
changing. And hence, focusing on a particular technology is a bad decision. It could be one specialization that you do. But in tech, technologies are tools. So if you look at a mechanic, a mechanic never says that I'm a screwdriver mechanic.
00:50:47
Speaker
or I am this electricity tested mechanic. I mean, you're a mechanic, you're supposed to go and solve a problem, right? And you have a bunch of tools that you can take help from. So, hence, it was very clear to us that if we have to build this cohort, it should be long enough, where we don't focus on one particular tech stack or one particular skill, but we'd rather make you really good at the art of learning and problem solving.
00:51:14
Speaker
So even today, the cohort is nine months. But the first five months of the cohort are essentially saying that you go figure out the tool. But we will give you a lot of problem statements. So in those five months, you get approximately 400 problem statements, which are real-world problem statements. And you have to solve those problem statements. Come back to me with a solution. Give me an example.
00:51:39
Speaker
For example, I mean, these are very technical examples. I'm just trying to think if there is a very non-tech. For example, one of the technical examples is that in tech you have something called a cache. Caches, imagine I have to store some information. Now if that information is a lot, for example, it's a YouTube videos. So if the videos are stored in a database somewhere, which is a machine farther off, it takes me some time to get that video.
00:52:08
Speaker
But let's say if a video is very popular, let's say there is Gangnam Style which video which is very popular, everybody is fetching that, then why not keep it closer to me, where let's say that video is residing somewhere in India and everybody requesting for Gangnam Style is fetching it from that place.
00:52:27
Speaker
Now then the question is, how do I choose which videos to keep in the machines in India versus what gets first from the US? So there is a concept of LRU cache. LRU cache is least recently used cache, which is keep putting every video in the servers in India and keep removing the videos which have been least recently used. So Gangnam Style, if everybody is requesting for it, it will never be least recently used. It will have a recency of use.
00:52:57
Speaker
So for example, one of the problem statements is like, go build a LRU cache, right? Like you have this use case of storing videos. It can only store, let's say five videos at a time. Now go build this, write the algorithm for this, what five videos to choose. And I'll keep telling you like what people are requesting for. And then based on that, you figure out like what five videos to retain here versus what to keep in the US DB.
00:53:24
Speaker
That is one slightly complicated example. One simpler example would be that imagine I have WhatsApp. WhatsApp has all of your messages stored locally inside your mobile. Now, if you search for anything in WhatsApp, let's say you search for the word vaccine, it will show you where all places where vaccine has occurred in your conversation.
00:53:46
Speaker
So one problem statement could be just build that, but I'll give you, yeah, just build search, searching for words in a very, very large corpus. Do people need to have some prerequisite skills before joining or do you teach them from basics? So right now we require them to know some bit of coding.
00:54:10
Speaker
But then, I mean, so there is first one month which is focused on just strengthening their coding skills, making sure they get comfortable with coding. And then the next five months I said, like completely focused on strengthening your problem solving muscle, right? Which is you get accustomed to seeing any kind of problem and not getting overwhelmed, but rather thinking, okay, fine, let me break it down. This might be step one, step two, step three, et cetera.
00:54:35
Speaker
Then we go into the specialization. And that's when we talk about, look, here is what a full stack engineer can do. Here is what a backend engineer can do. Here is what a mobile engineer can do. And you can choose your specialization and you can do that. But our core philosophy is that all of that is useless. I mean, that's another mistake which I feel most tech players are doing, that you keep focusing on the tools.
00:54:58
Speaker
The first step to any process is you help people learn the art of learning. And they apply the art of learning then to, let's say, a specialization. That's when you're preparing them for life. Otherwise, you're designing them to fail at some point later on.
00:55:16
Speaker
Okay, amazing. So like we were talking about the evolution from cohort one to today, like tell me about that. So cohort one was also nine months? Yeah, so actually, there has been some amount of evolution there, in fact, a lot of evolution there, because we were emulating Stanford. So the first thing was that, look, I mean, and this is a learning that we've had. So at that point of time, we believed Stanford is Stanford because also because of its intake.
00:55:45
Speaker
The intake is very high IQ, very high capability people, very selective in nature. Maybe like single digit percentage gets through. Yeah, same for IITs as well. So every single large institute, a successful institute, we looked at everybody was very selective in nature. So we said maybe that's something that is required to be a great institute. So let's focus on that. Now, how do we build that?
00:56:10
Speaker
So to build that one is you need a lot of pull, which is a lot of people want to come and study at your place, and then you are very selective. So to build a pull, we said, let's create a model which nobody can say no to. And that was a post-paid model, where we said, look, we're not even going to take money from you in the beginning. You come here, you study here, and if you find a great job, then you're liable to pay. Otherwise, don't pay. That's it.
00:56:39
Speaker
So this is what we started with because we wanted to be very, very selective. Now we realize a bunch of problems with that model later
Prepaid Model for Student Commitment
00:56:47
Speaker
on. It was just attracting the wrong kind of people in certain cases.
00:56:56
Speaker
I mean, it helped us be very, very selective. I mean, our selection rate was less than 1%. It helped us attract really, really good people. But six months down the line, and the program was also shorter. It was six to seven months. And what did they pay once they got a job?
00:57:14
Speaker
They would pay 17% of their salary for 2 years or 3 lakhs, whichever was smaller. So the maximum amount you would pay would be 3 lakhs or 17% of your salary for 2 years in case that amount was smaller.
00:57:31
Speaker
One of the learnings in six months, by the end of 2019, one of the learnings we had is that when we started looking at people who were graduating and where they were ending up, so we started doing this correlation analysis. The thesis was that if we are very selective, we are very selective because the smarter people have higher chances of succeeding.
00:57:54
Speaker
Is that correlation still holding or not? One of the things we realized was there was absolutely zero correlation in the entrance score that they scored versus the outcomes they had. For example, Google is considered to be one of the companies that is very, very hard to get into.
00:58:19
Speaker
Extremely selective. And some of the people who got into Google had almost the lowest score in entrance tests. We were about to reject them. Had we been slightly more selective, we would have rejected them. But they were very high on motivation. They were very high on grit. And that's why they were able to learn a lot and hence make it
00:58:48
Speaker
make it like one of the best outcomes of the batch. So then we started questioning our assumption that is the assumption that if you're very, very selective only then people succeed. Is that correct? Did you have an entrance exam? Yeah, we did. That's how we were very selective. So we would do this massive scale coding exam. We would take people the best of the best in the cohort. We would then polish them even more and then they would get into the companies.
00:59:19
Speaker
But that correlation showed us that our thesis was flawed. You have to filter for motivation and grit, not for their current level of skills. We were filtering for the wrong things. Now, how do you filter for motivation and grit? That became the challenge. It's not very easy to judge how motivated a certain person is.
00:59:47
Speaker
Plus, we had also started noticing that ours was very heavy on community play. We had started noticing that when you say that, look, I mean, you only have to pay when you get a great job, then certain people that you attract, they're not interested in learning at all. They're interested in only the final outcome. So they come into the cohort and they're like, now it's your responsibility. I actually love it. There's no skin in the game for them. There is no skin in the game. Yeah, yeah.
01:00:15
Speaker
So we said this is a model that's also not sustainable. Let's switch to the conventional prepaid models. Like you went completely prepaid, like 100% prepaid.
01:00:26
Speaker
Yeah, we went 100% prepaid in 2020. No postpaid element whatsoever. But by then, we had also created a bunch of these outcomes, which also got people to trust us. So hence, that decision was easier. One of the other data points that we started seeing was that in the cohorts in 2020, the first few three cohorts that we did, we gave people an option. You can either opt for completely prepaid, and in that case, the amount you pay is half.
01:00:54
Speaker
1.5 lakhs versus you can offer post-feed, which is fine, but then when you do find a great outcome, then you pay double the amount. We had started seeing half of the batch coming in complete prepaid without any job angle whatsoever to their learning. Those people had higher adherence. They had higher motivation, higher grit.
01:01:17
Speaker
So it became very clear that like, look, I mean, if motivation and grit is what you're optimizing for, then maybe like one of the ways to do that is by just altering your entrance criteria, like your entrance criteria shouldn't be your current level of skills.
01:01:34
Speaker
So we went completely prepared, we made a bunch of design decisions again on how to make sure these people still have the same end outcome. So we broke that, I mean, so we started imagining this, if you look at the entire learning as a hundred meter race, then your current levels of skill means somebody might be at 10 meters, somebody might be at 20 meters, somebody might be at 40 meters. So there is one, I mean, some amount of pocketization required, because if you are early on in the journey, then you probably need more time to reach the end point.
01:02:03
Speaker
So the program duration cannot be the same for everybody. So we said like broadly, let's split it into three parts. There are people who are advanced and maybe they only need seven, eight months, but then there are people who are intermediate. They probably need nine to 10 months. And then there are people who might be beginner, who might need maybe a year or maybe more than a year to actually reach the same point.
01:02:25
Speaker
The second thing that we figured was that if this is 100 meter race, if we are somehow able to filter for motivation and grit, then even then like how fast do I run is also varies from person like people in the outside world term that as IQ, I don't believe in IQ. I believe like, again, if you are determined, if you want to do certain things, I mean, it might take you more time, but you'll eventually get there. And that's what matters.
01:02:51
Speaker
But I mean, the pace of learning, the pace of absorbing something also varies from a person to person. So hence, again, like saying that everybody will learn at the same pace is also the wrong assumption. What you need to do is that even if in intermediate, you have a bad starting, you have to identify people who learn really fast, you have to identify people who learn really slow, and you have to keep have them in different cohorts so that like you're optimizing for their pace, not your base.
01:03:19
Speaker
So we made some of those core design curriculum choices. We created cohort space on that so that eventually everybody succeeds. This happened in June of 2020. So it's been a half years. It seems like our thesis is correct. So every single person from every single batch till this point, every single person.
01:03:48
Speaker
in this prepaid mode has had massive outcomes. Our average, this is, please don't put that in the official video, but our average salary pumps is 2.7x.
Impact: Salary Increase for Graduates
01:04:04
Speaker
So people who come, I mean, they almost end up tripling their salary in a year.
01:04:10
Speaker
Is it like what percentage of your cohort is already working? I thought you would have a lot of college kids like who have not yet started working.
01:04:19
Speaker
So when we did the pivot from post-paid to prepaid, one of the things that we said was, look, I mean, let's start with audience that is slightly more mature. One of the things that actually solves the motivation and grit is that if you've had, there's a phrase in Hindi, then you get very serious about your learning.
01:04:41
Speaker
And you're more focused, you know, this is what I want to do. Correct. Yeah. And then you know that this is what really helped me, right? Otherwise, in college, you're optimizing for, I mean, this is a very unfortunate education system actually makes us optimize for getting that degree, passing that exam.
01:04:58
Speaker
Why am I learning that? That is not my priority at all. So hence we said like maybe in the beginning, let's not go after college students. Some of them might be great, but let's just filter for people who have had a lot of Thappads in their life, because then they know why they're learning. And from June onwards, June 2020 onwards, every single person that we've enrolled, they've all been working professionals.
01:05:25
Speaker
So you like just say no to freshers or it's just a coincidence? Right now we say no to freshers. So right now we say no to freshers. There is a vision to build a parallel college as well, which is taking people right after class 12, maybe create a 15, 18 month program for them, not have them go to college at all. So we'll solve for those segments later on. You need a different product for freshers.
01:05:56
Speaker
Yeah. But right now, very clear focus on working professionals. They are harder to teach, to be honest. Because they already have a job in parallel, they're focusing on that. Plus, even on the problem solving bit, every single thing that you teach, you have to explain them the why.
01:06:20
Speaker
You can't teach them anything in the air. It has to go in a fair amount of depth. Whereas if you're teaching students, they get happy with even small lollipops. So they harder cohort to teach, harder cohort to please.
01:06:39
Speaker
they put in their effort if you tell show them the why and then you'll see like a lot of them then finally making it big in life which is at the end of the day what keeps us going. So like from June 2020 now we've had almost 1500 odd graduates and the data point that I was telling you 2.7x salary increase that is across those 1500 data points
01:07:10
Speaker
So, they would be people with like a 10x increase also.
01:07:14
Speaker
Yes, there would be people with tennis increase. In fact, I don't talk about these stories because then people get a little... You don't want to advertise the outliers, basically. Yeah. For example, we had this person who had like five years of experience hired by Ulan for 1.1 CR in salary.
01:07:40
Speaker
which is a very outlier case. But the thing is, if you advertise those, then everybody starts to expect that. So you don't advertise. But all of this is a good sign that at least what you're doing is in the right direction. You're creating people who can go in the real world and make very large impact. What was your intake in cohort 1? The very original cohort 1 had 198 people.
01:08:08
Speaker
Okay. And now you're doing about 1500 people a year, something like that. No, no. So this was when we 1500 people graduated. So June 2020 onwards 1500 people graduated. So this would be maybe intake from June 2020 to December 2020.
01:08:27
Speaker
Now we're in taking at much, much faster pace. For example, this last one month, we onboarded about 1300 people. We do about 1300 to 1500 people intake every month. This is across two different cohorts now. There are people who want to build a career in the direction of tech, versus there are people who want to build a career in the direction of data science. Those are two different parts altogether. Okay. You launched a different program for data science.
01:08:55
Speaker
Yeah, same philosophy, same online Stanford, every single decision very similar, driven by content which is relevant in the real world. In fact, even in data science,
01:09:10
Speaker
But we went one step further and we said all of the same first five months throwing problems at you, all of these problems will be some business case study from a company, which is a real business case study. So we've tied up now with more than 70 companies.
01:09:28
Speaker
and some of them are from the Indian ecosystem, the larger ones who have recently IPO'd. So taking business cases from them and then throwing that at them to, again, develop that muscle of problem solving. And that is important because our education system plus a lot of our IT services companies, unfortunately, train people to think in terms of tools and not the solution.
01:09:57
Speaker
So that thought process of don't worry about the tools. Tools are only tools. But worry about your approach to the solution. That's what makes you a better data scientist or a better software developer. That's a mindset shift.
01:10:13
Speaker
that has to happen in the masses and hence focusing on that. So you do a lot of personalization. So what does a cohort look like today? Is it like batches of like 20, 30 people because you're doing so much segmentation like beginner intermediate advance and then within intermediate also you have two separate levels. So how does that happen? How does your intake, how's the whole flow managed?
01:10:44
Speaker
Right. So, I mean, if you look at it like there are about 1300, 1500 odd people being taken broadly five, six segmentation. Actually, what we figured out is like in advance, we don't need the fast versus slow segmentation usually. Yeah, you don't. Yeah. But again, intermediate and beginner you do. So, typically a cohort has about 200 odd people.
01:11:08
Speaker
who are very similar in their pace of learning and what they know today. Like say intermediate lower would be one cohort, intermediate higher would be a separate cohort, okay. Yeah, yeah, yeah. I again wouldn't say lower or higher while the pace of learning is different, but yeah, I mean, that is one cohort, correct. It is that it takes them more time to learn the same thing. So that means that has some implication on the duration, et cetera, which is fine.
01:11:38
Speaker
Do you still have an entrance exam?
Entrance Exams and Cohort Placement
01:11:41
Speaker
We do, but the selection rates are now 30% instead of a single digit percent.
01:11:47
Speaker
And that entrance exam is only to just make sure that at least you know that much so that the course is effective for you. For example, we don't spend time teaching you class six, seventh maths. We expect that you're already good at algebra, you're already good at trigonometry, you're already good at, let's say, some elements of number theory. Very similarly, like very basic level of coding. For example, if I ask you to write a code to add up two numbers, at least you're comfortable with those elements, right?
01:12:16
Speaker
That is barely minimum required to be effective in the course, so hence we filter for those. Plus, would that entrance exam also tell you where, like which cohort to slot people into? Yeah, yeah, yeah. And how do you judge who's a fast learner and who's not?
01:12:31
Speaker
So initially on day zero, we don't really know. What we do is that every single class, we get feedback from people on how, one is like how the class went, how was the instructor, but also did you feel the class was slow or fast? And that starts to, once we've done one or two weeks of classes, then that starts to feed in as input, where we then have conversations with people.
01:12:57
Speaker
So it's still very operationally heavy. But we have conversations with people. We tell them that, look, I mean, you found these classes to be very slow. We recommend that you should be in the faster moving batch. And then based on that, we do the segmentation. On day zero, it's like combined for all the matches together. The first two weeks is where you're sort of discovering
01:13:23
Speaker
The segmentation, however, in advanced, intermediate, and beginner, that has happened on day zero because the test gives you that. But the pace-wise, we don't do. In fact, in the early days when we had smaller cohorts, it was possible we would not end up doing the segmentation at all in case everybody was finding the pace of lectures, etc., to be the right pace. Now the cohorts are so big that we need to split anyhow.
01:13:53
Speaker
And how does that one-on-one mentorship happen? Like you found people, so as you told me through interview bit, you found people who were willing to, were they like volunteering or was it like a paid gig, like these mentors?
01:14:06
Speaker
So we give them options. Some of them actually volunteer because their employer contract does not allow them to take payments outside of their employment. Some of them choose to get paid. So we let mentors choose that. But mostly, I mean, the payment is also so small that nobody is doing, to be honest, for the payment. It's only the spirit of giving back.
01:14:33
Speaker
So typically what happens is that when let's say you as a learner are on the platform, you see a list of mentors and along with that some text from our side, which is our systems recommending why we are recommending it to you. So for example, let's say you currently in DCS.
01:14:49
Speaker
And let's say this person was in TCS at some point in time, but today is let's say working at a hot star or let's say Google, et cetera. So we're recommending because this person has been in a situation like yours and hence might be a great fit. So you see a bunch of these options and you choose one that you like.
01:15:08
Speaker
Tell me about the product evolution. You are doing live classes, and the chat is also through Slack. What is the product that you have developed? Is it for assignments and quizzes and stuff like that, which happens on the scalar website or the scalar app? Do you have an app also? What is it?
01:15:34
Speaker
So we don't have an app because most of the coursework involves you to write code. It's very hard to write code on mobile. So we intentionally state desktop only. We in fact instruct everybody to get a laptop if they don't have one already.
01:15:50
Speaker
That being said, the product, which is the core product, the primary metric that that product is driving, apart from assignment quizzes and a bunch of other things where you can write code, is the engagement. How much time are people spending on this product every day? And hence, what kind of things to build so that people have a reason to come back every single day and spend a lot of time here.
01:16:15
Speaker
So the product is primarily around gamifying various elements. For example, let's say you're doing a problem, you get stuck, you raise a help request. If there is a fellow person who's coming and helping you, then that person earns some number of coins.
01:16:31
Speaker
You have certain house dynamics, just like Harry Potter has houses. You have houses here as well. Every single thing that you do, either makes your house gain points or your house lose points. So your house leader is after you to keep doing certain amount of things. Yeah, so you have a leaderboard where you can see the entire rank list of where you stand in the class. And there's a concept of a streak. Every single day, there is a target of points that you need to achieve. Even if you miss on a single day, then your streak reduces to zero.
01:17:01
Speaker
There are people right now in the cohort who have like 360 days streak or 400 days. So, last 400 days they have been on the website every single day achieving certain number of points. And as you hit certain streak milestones, for example, 25 days streak or 50 days streak, you earn a large number of coins.
01:17:20
Speaker
which lets you buy certain things in the real world with those coins. So those kind of things that are there in the product, which just solves for engagement. People come at the point of entry, we solve for motivation and grit, but then making sure they stayed motivated.
01:17:38
Speaker
They don't have to continuously push themselves. That is the responsibility of the product to a certain extent. What are the elements in the product? One is like an assignment and quizzing platform and then you have these forums where people can ask like a Quora kind of a
01:17:59
Speaker
Right. So essentially, the product has this place where you can chat with your mentor, this place where you can see all of your classes. So recording of classes, the corresponding assignments. So if you missed a class, then you can watch the video. You can watch it, right? Yeah. Even when you're watching the video, there is a way to, if you get stuck somewhere, you can raise what we internally call as concept help request.
01:18:26
Speaker
which is a TA will then come on a one-on-one call with you and explain that concept to you. That didn't make sense to you. As you're solving the assignment, every assignment has, I mean, code that you write, it feels that you're getting stuck. There are these hints as nudges that the assignment, the platform gives. Have you looked at sorting the data? Maybe that might help.
01:18:50
Speaker
So that kind of a thing. If let's say you're still not able to progress, you can raise, I mean, you can look at either the video solution yourself, or I mean, if something isn't making sense, you can raise a request and a TA will again come and help you. Maybe do a quote review and say that here is maybe where you're making a mistake.
01:19:09
Speaker
There is a rank list to compare with your peers. Just see like what my peers have done today versus what I have I done today. Where are my peers with respect to levels versus where I am? There is a news feed of seniors with seniors in this context that people have enrolled in earlier batches. What all jobs are they landing? And this is like a constant news feed. So just like Facebook had this ticker on the side. Yeah, so this is motivation.
01:19:37
Speaker
So that keeps happening on the side. There's a bunch of things. I mean, I might be missing out on a lot of things. There is a job tracker as well, where you keep seeing these companies that have just released a new job. We don't let you apply until you've covered some certain part of the course. But people who have, they see which jobs are relevant for them. And they can apply. And they say that I want to apply. There is an entire funnel there that we show, like, which stage are you at, et cetera.
01:20:07
Speaker
And does, like, does this data help you create more like, like a smarter platform, like if someone raised the concept request, and then you see that a lot of people are raising this concept request, then does that get productized, like, faculty, maybe, or like, you know, how do you make this smarter using tech?
01:20:28
Speaker
So every single thing has a feedback loop in certain sense. For example, by the way, the platform where we hold our live classes, that is also in-house because it has a bunch of things inbuilt. Even the things that you're doing inside the classroom contributes to your rank list. So within the classroom, there are also quizzes and that point actually gets added up there. There are coins that are rewarded, et cetera.
01:20:55
Speaker
the feedback that you give once the class ends, that goes to the instructor to tell them what have they not done well versus what have they done well. Let's say if on a problem
01:21:06
Speaker
because the assignments don't change. I mean, the first five months where I throw a lot of problem statements at you. So if I have seen certain kind of people raising a request to the TA, and I know from the past that there have been similar times where other people had raised requests, then instead of actually going to the TA that might take me 15 minutes, I immediately show them. By the way, there was a student just like you who had also raised a request at this point of time.
01:21:34
Speaker
with a very similar looking code. And here is what the TA had suggested. See if this is helpful. If not, I will connect you with the real TA.
01:21:44
Speaker
we call that internally as automated TA. Concept help requests as well, we try to create group concept request sessions. So for example, let's say I say, look, I mean, this particular concept recursion did not make sense to me. Can somebody explain to me? So the moment, let's say I raise that request, it gets broadcasted to all students in the same batch that there is going to be this concept help request session by this TA in an hour from now. If you want to join, you can also join.
01:22:12
Speaker
So basically there's a lot of buzz happening all the time, like in an army, it's a lot of real-time engagement happening. Hence, you want to be on the platform all the time because you miss out otherwise. Every single concept help request then creates a recording. So in case, let's say there is, you can't wait for the TV event for 15-30 minutes, you can look at a recording and then you can move forward and so forth.
01:22:37
Speaker
There's so much happening inside that is very hard for me to articulate what exactly the product is. The whole and sole goal of the product is to solve for engagement. Make sure that people are just hooked, addicted, and are spending as much time on the platform as possible. They never go out of this ecosystem. Because if you've solved for that, I mean, right now, for example, the average amount of time spent per user per day is 186 minutes. 186 minutes is...
01:23:03
Speaker
Even if I'm spending time watching Netflix all day, I still don't spend money. True, true, true, true. Essentially, I assume it would be a lot like, say, working at Facebook, in the sense that constant buzz, camaraderie, learning from each other, doing exciting stuff. Right, right, right. So have you raised funds? Once you started your cohort one, did you raise funds at that time or subsequent?
01:23:33
Speaker
Yes, we did. Our first fundraise was when we were pivoting to this scalar model, online Stanford model. This was in February of 2019. We were in the first cohort of SQL Server. We raised about 1.5 million.
01:23:54
Speaker
Then, once we had started running this cohort, then sometime around August, September, the same year, which is in six, seven months, we did our series A, which was about 20 million from Sequoia and Tiger. Sequoia led the round.
01:24:08
Speaker
And we very recently, because we are now also expanding internationally, we very recently raised another 50 million or so in the recent fundraise that we'll be announcing very shortly. So why do you need funds now? Because it's prepaid, so it's like a cash machine, no?
01:24:32
Speaker
Right now, to be honest, we haven't spent all of the money we raised even in the past. We don't need funds to sustain the core business.
01:24:43
Speaker
All of the fundraise we're doing are either for acquiring maybe new companies that might help us become stronger or helping us strengthen the brand because brand is an expense that we do today to help people tell our story but might not give us immediate returns. It's a long-term investment.
01:25:03
Speaker
So brand building, M&As, and international expansion. So whenever we go to a new region, for example, right now we're launching in the US, there is a lot of upfront cost to set up operations in a new country. In terms of hiring the TAs, the faculty.
01:25:22
Speaker
Yeah, hiring the team, hiring the faculty, hiring the TAs, mentors, they might not, I mean, it might not start making money immediately. But I mean, the assets, etc., that you build in the region, they'll eventually start to become cash positive.
01:25:36
Speaker
So we raise funds preemptively for those reasons just to make sure that our growth trajectory isn't stalled and we're doing everything we can to become a large company. I mean I believe that scalar for the problem it is solving and the approach we have can replace universities someday. Yes absolutely. Actually like six years on the line I mean I'm
01:25:59
Speaker
What I believe is that people would not go to colleges for regular reasons. There might be these off-cases because of which people go to colleges. In that lens, it's a company that can become a $100 billion company, so raising funds to accelerate that journey.
01:26:17
Speaker
That's the only reason why we raise funds.
Global Engagement and Local Contexts
01:26:20
Speaker
Okay. So when like say your US cohort starts, will it become like a global engagement and interaction? Like the peer group, will it become like a global peer group or the India cohort will have a separate Slack channel and like, what do you see that as happening?
01:26:36
Speaker
No, there is merit in actually having it global. There is merit because both parties learn from each other. If I am somebody in the US, I have a bunch to learn from the Indian ecosystem. And if I'm somebody in the Indian ecosystem, I have a bunch to learn from the other ecosystem.
01:26:52
Speaker
Some of the things that we do here, I mean, for example, we get people to meet locally in a meetup, etc, or we play crickets on weekends, we get the cohort to come together, play cricket, those we won't be able to do. But the other aspects we would still be able to execute on.
01:27:08
Speaker
So the idea there is like, get people to mingle with each other. The one just difference might be that the faculty might be different for let's say the person in India versus person in the US because and the mentors group, etc, would also be different. Because there is local context. Yes, yes, yes, yes, yes, yes. Yeah, to build engagement, you need that kind of.
Skill-Based Hiring Over Degrees
01:27:28
Speaker
And do you still run your hiring business?
01:27:32
Speaker
Yes, so we don't charge companies anymore. In fact, it has gotten deeper. Now we work with more than 750 companies in India and Europe. There are some certain IPs we are creating there. For example, we've created something called Fairplay, where because we've had such long relationships with companies, we have now been able to convince them that
01:28:01
Speaker
degrees don't matter. So why don't we create skill-based assessments with you? And we'll get people from Scaler, even outside of Scaler, to take those skill assessments. But as long as people are clearing that assessment, you don't ask for resume. And we would not give you resume.
01:28:22
Speaker
which gives you the skill assessment profile and then you take the column whether you want to interview this person or not, which is truly unbiasing. Now not saying that I only want to hire, let's say, a 20-year-old versus I won't hire a 40-year-old versus I only want to consider people from a certain college, etc. You don't get access to any of that.
01:28:43
Speaker
And to be honest, companies are more than happy to. The only reason why they had been using degree as a proxy is because that's a filtering mechanism. There was no better way to do that filtration. Exactly, exactly. If you can give them a better way of filtering, which is like you're saying that, look at me, here are the things that are important for being successful at your job. And this person has done it in a proctored setting, which gives me confidence that this person is capable.
01:29:10
Speaker
And you have no reason to say that I will only look at certain x, y, z degree. So you don't monetize from companies at all, like even if they want to, and you're still doing that interview bit pool, like helping those people get connected to companies or is it only for the scalar pools? Right now it is only scalar pool.
01:29:32
Speaker
and that has actually happened organically. Given the the higher focus has been to make sure that people and skillers succeed, their entire team focus would anyhow go there. So even without saying that that has started happening and so we ended up formalizing it. We don't charge companies right now because we believe that the model that we were operating in even at interview bit
01:29:55
Speaker
unless it has a large value add as compared to the agency ecosystem. Where the large value adders look, I mean, I'm so accurate with my recommendations that you probably end up hiring every single person that I recommend to you.
Lead Generation and Alumni Trust
01:30:08
Speaker
If I reach that stage, then it makes sense to monetize because you're just adding immense amount of value. Before that, if I'm just another agency, then there is no differentiator that I have. In that case, I'm in monetizing, it's only going to harm me in the short run. Right now, we're in this phase where we are just collecting a lot of information and data and figuring out what are the hiring patterns at different companies.
01:30:32
Speaker
So, working with companies helps you make a better course also because then you see what... Absolutely, absolutely. In fact, that's the entire idea of Fairplay. If you have these assessments, you see where your students, even before they go to the company's unit, you see where your students are doing well versus not and hence what needs to improve in the curriculum. So, Fairplay is a standalone web app or is it within scaler or what like?
01:30:55
Speaker
Right now, it is within scalar. At some point of time, we will make it stand alone, open to all. But right now, it is within scalar because it ties into... Any company can sign up and use it to assess anybody, even if they are not a scalar student. Any student can be screened. Like any kind of tech hiring companies can use that as a free tool for screening.
01:31:18
Speaker
At this moment, any company can come and create assessments with us and roll it out to all students in Scaler as of this moment. At some point of time, we will make it so that it is accessible for anybody out there.
01:31:33
Speaker
But at this moment, which would become a lead gen engine for you then? Like if say Google starts hiring through this product, then it'd be like maybe tens of thousands of people who would go through that assessment and that would be like a lead gen engine. Currently, how do you do your lead gen? How do you do your customer acquisition?
01:31:54
Speaker
So currently it happens either through interview bit. I mean, a lot of people practicing on interview bit, maybe some of them struggling and that becomes the lead Gen Engine. We also do these master classes on the weekend where we pick one very interesting topic and we do a class on that, which is just to show like how good we teach.
01:32:18
Speaker
That actually gives people confidence to enroll. And then we have usual ad channels that people go through. Actually for us, Google Facebook. For us actually, reference is a big channel. Today, 30-35% of our intake actually comes from reference. People in the cohort referring their friends that please enroll.
Financing and Expanding Market Reach
01:32:42
Speaker
So referrals, interview bit, master classes, those are the primary leads and channels. And then the example of it is happening today through Google, Facebook. Some bit of branding effort that we're doing also actually contributes to that.
01:32:56
Speaker
And then, I mean, in education, to be honest, just add a loan or lead gen loan is not enough. You have to generate trust. So for trust generation, then we do a bunch of things. For example, to anybody, we let them connect with alumni and have a very free flowing conversation with alumni just to figure out with them unfiltered what did they like versus what they did not like. The other thing, for example, that we do is I
01:33:26
Speaker
or me or Abhimanyu, one of us does a free live class every day, which is just to talk about, look, I mean, here is what is important to succeed in tech, which is more of showing people the path. And here are free resources where you can go and learn. But if you want a structured ecosystem, then come together. That's a class we do almost on a daily basis.
01:33:52
Speaker
That is again another trust builder like when you actually show them the breakdown. And then there are a few more efforts like that that we do which helps build trust. So there's a lead gen part, there's a trust generation part and then you have the final enrollments happening. Does the conversion happen? Is it like an online checkout or does it happen over a call? Do you have like an outbound sales team also?
01:34:17
Speaker
Yeah, we do have an Auburn sales team, primarily because the ticket size is higher. So a lot of people require help with financing. So the Auburn team actually helps them figure out what are the right financing channels for them. We have a bunch of tie-ups, so they connect them with those tie-ups and then the financing happens.
01:34:42
Speaker
Also, the outbound sales team actually also helps them answer certain questions that people have, like people in general are at different stages of their lives. So they usually want to understand like, like, you know, in my context, it's a big amount. So I think like, at least in India, we want to talk before spending such a big amount.
01:35:04
Speaker
Yes, yes, yes, yes. So when are you unlocking the bigger market of pressures? I think that is like a much, much like that will probably be 10x the size of what you're currently doing, right? Like being able to take someone who doesn't know coding at all and make him get a job at Google. Like, you know, that achieving that would be like 10x value, right?
01:35:29
Speaker
So actually, there's some pre-work happening there. We still want to make sure that everything that is good about the university is still retained. For example, we have done a great amount of work on community and peer-to-peer learning.
Metaverse and Future of Education
01:35:45
Speaker
Colleges are usually the first time when you when you're staying outside of your home in this in this hostile setup, right and that's that's a That's a core part of your life where you create these really really good friends So right now we're in the phase of figuring out answers to some of those questions while we believe that degrees are very relevant we still believe those relations are very relevant those net that networking is very relevant, right and
01:36:12
Speaker
And then while we have like sign-offs from employers already that look, I mean, if you create this 18 month program, you have people coming after class 12, we are giving you a sign-off, we don't require degrees. So we already have a lot of employers who have signed off on that. But I mean, figuring out these pieces is important. So right now we're working on that. Tentatively, maybe sometime around August or September next year is where we're targeting. But that is not, given this is such a long play, I mean, it's not written in stone yet.
01:36:44
Speaker
So it's going to happen very soon. Do you think the whole move to metaverse is something? I think that would also be a game changer for you, right? Because it actually then elevates you to a level playing field with the regular college experience. Right, right.
01:37:06
Speaker
Metaverse is going in a great direction. To be honest, I still feel for Metaverse to become really adopted by majority. It still may be a year or two out. It will eventually get there. I'm thinking like five, ten years out.
Opportunities in Global Tech Education
01:37:22
Speaker
Yeah, I've been following a lot of the news, etc., that is leaking about Metaverse. So I'm just going by that. So maybe two years out.
01:37:31
Speaker
So it is headed in the right direction. It is doing all the right things. It's just like maybe it's just a matter of time. But I mean, if we can't wait for two years, we will not rely on Metaverse to be honest. We'll figure out certain alternative solutions. But you would build a Metaverse version of Scaler, right? I mean, that would be like, again, another game changer, right?
01:37:54
Speaker
Yeah, I mean, so I would say it is completely dependent on the objective and whether it is the best solution to achieve that objective. The objective is get people to talk to each other a lot, have a lot of peer-to-peer learning happen, and have people graduate with real friends in the ecosystem. Essentially do what, let's say, an ISB does in those 11 months. So if Metaverse is the best solution for that, we'll build Metaverse. But if there is something else that is a better solution for that, we'll build that.
01:38:25
Speaker
Are the tools in place to help you easily build like a metaverse university or do you have to do a lot of things from scratch right now? I mean, you know, like for building an e-commerce company, you have API-based tools which can just help you stitch together stuff and not really worry about infrastructure. Is that in place yet?
01:38:47
Speaker
I mean Mirabus is still a very relatively new ecosystem so the tools etc if there are to be built we would have to build ourselves.
01:38:57
Speaker
Good thing is, I used to live in Facebook with two other flatmates. One of them is currently leading a large division within Oculus. If we need support for that, we have support for that. That being said, as I said initially, Metaverse is a tool.
01:39:20
Speaker
Not the solution. So whether that is the solution or not, that's a problem to be solved. I'm not saying that metaverse is the solution. Maybe there's a different solution. We will arrive at that. And maybe in two, three months from now, I'm at the liberty of talking about that. But I mean, we would solve it before launching a university. So are you also looking to expand into other areas beyond tech or do you want to remain focused on tech education?
01:39:49
Speaker
The shorter answer is that we want to remain focused on tech education because we believe that the macro is going towards tech. What is essentially happening is, and this has been happening for the last 10 years,
01:40:03
Speaker
is that everything is moving to become tech-focused. Even the car industry, if you see, Tesla is now a car plus tech, and hence the largest car company in the world. Car is secondary, tech is first there, yeah. Tech is first, right. What has COVID, in fact, done in the last two years is that every single industry in the world is now going or realizing that there is no option but to go digital, every single industry.
01:40:32
Speaker
What that means is that what we're going to see in the next five years is that the demand for people who can do tech, people who can be, let's say, either a software engineer or a data scientist or some variation of tech is going to explode. There's a report by GitHub that says that there are going to be 50 million new tech jobs created in the next four or five years.
01:40:55
Speaker
If that is where the world is moving, every single job is going to have a functional element of tech, then that's a large vertical to solve for. That's where the ecosystem can fail because we don't have enough people in tech right now. Hence, the focus is going to be completely on tech.
Diverse Tech Roles and Hiring Strategies
01:41:14
Speaker
I have two life examples to actually quote.
01:41:18
Speaker
I was in London for setting up the London thing, and there I would meet a bunch of people. Yeah, Facebook London. So in that duration, I saw two major changes. One was this incident where I would go and take up these metros, which were called tubes. In six months, I saw the ticket counters where there would be people sitting replaced by these vending machines.
01:41:42
Speaker
And I had a friend I met somebody there who was working in this oil company whose only job was to review these oil pipelines in the Middle East and And identify if there is a leak happening because the oil leaks were very expensive
01:41:58
Speaker
What he told me was that when he joined there were about 100-150 odd people doing the same thing. In his office, maybe there might be a lot more. In just again a duration of a year, number of people working on the same thing was reduced to almost 10.
01:42:15
Speaker
The remaining was taken up by a tech person or more than one tech person who would be building these systems that would identify the leaks and then flag it to the human person. That is where the world is moving. Every single thing that is repeatable is being automated in certain sense. And governments are recognizing that too. Singapore has made coding education mandatory from class six onwards.
01:42:45
Speaker
In fact, there is most probably Indian government will also do so in some time if it has not already done so. So every single government, in fact, like Middle East has made it. I'm not sure what is the clause there, but Abu Dhabi government has said that if you come and set up a company here, then the first few years, I think the first two-year salary of your employees, we will pay.
01:43:10
Speaker
If it's a tech company, it's a tech innovation company, then that's what would happen. Because every single country, every single ecosystem is now realizing that this is where the future is going to get built. It's not a fad. It is actually happening. At scale, it is happening.
01:43:29
Speaker
If that is the problem statement to be solved, that's a large enough area to be focused on. The good thing is it's not dependent on a region. It is truly global. If you create really good software engineers, that software engineer could very well be working on a problem statement in the US sitting here.
01:43:45
Speaker
If you can solve the problem of talent in tech, that is a fairly big opportunity to crack. And currently, I think the talent in tech problem is massive. Companies typically have to maybe offer 10 people to eventually have two people joining them. So what type of roles can companies look to scalar for in terms of filling their hiring needs?
01:44:13
Speaker
Mostly all kinds of tech roles. We have people who join the cohort who have experiences from one year to 21 years. We had people graduate who had gone and joined as VP or a CTO in certain companies. We had people graduate who are now architects in certain companies. And then there are people who go and join as SD1s, SD2s, SD3s.
01:44:37
Speaker
Now, obviously on the more senior roles, there is more targeted approach. I mean the fair play kind of thing does not work on senior roles. The fair play kind of thing works more on the junior slash mid-level roles.
01:44:49
Speaker
So, the numbers would be higher on, let's say, an SD1 or SD2 kind of roles. The volume would be lower on, let's say, a little more senior roles. And it's a self-signup, like any employer who wants to hire can sign up and list a job.
01:45:08
Speaker
Yeah, yeah, yeah. Get applications. All right. OK. Yes,
Scaler's $100M Revenue Goal and Global Vision
01:45:12
Speaker
yes. OK, cool. Amazing. OK, so my last question to you. What kind of revenue are you currently doing? Like, what do you expect to end this year at?
01:45:23
Speaker
Yeah, so actually, we track the revenue run rate. And this is actually, maybe the revenue run rate is a long terminology. It's a booking run rate. Essentially, our bookings sort of increase month over month. I mean, we look at what is the current monthly booking run rate. And then we multiply that by 12 to get an analyzed booking run rate. Oh, OK. Got it. So currently, we are at about $45 million booking run rate analyzed.
01:45:53
Speaker
And we expect to cross 100 before June next year. Amazing. And this is before US launch, or you are incorporating some assumptions of US launch in this? No, so this is without the US launch. This is India-India launch. OK. US will give us further boost. Right. So you could probably be at 200 million, say, maybe 18 months or so from now. Considering that US numbers will be much bigger,
01:46:23
Speaker
Yeah, hopefully. Fingers crossed. Amazing. How does that compare to other edtech companies? How much do these more famous edtech names do in terms of rendering? Do you have any idea? I only know from what has been released in PR. Baidu is obviously really huge. They do about almost a billion dollars in Japanese every year.
01:46:52
Speaker
An Academy, again, just from going by what has been released in the PR probably does about $150 to $200 million in revenue every year.
01:47:06
Speaker
Upgrad does about 100 million or so in revenues every year. Those are the largest players that I know of. And then there are a lot of smaller players that probably do closer to 1 million, 2 million, 3 million kind of a run rate in a year. So you would probably be like in the top two in say 18 to 24 months time, I guess, like top two, three at least.
01:47:31
Speaker
We hope so. We definitely hope so. For us as well, to be honest, this is a play and both me and Abhimanyu are very clear on that. This is not really a play for the next two years, three years. We believe that we are solving a very large problem and for that, it is okay if we solve that problem completely in a 10-year timeframe as well. For us,
01:47:57
Speaker
The question is, can this become a hundred billion dollar company in 10 years from now? Now, the path through that might be that we become, let's say, a two hundred million dollar revenue company in two years and maybe then a billion dollar revenue company in four years, etc, etc. Like that could be the path.
01:48:15
Speaker
So, that we are less concerned about. In all decisions that we take today might not lead to immediate revenues at this moment. In fact, some decisions might be counterproductive to revenues.
01:48:28
Speaker
because they are the right thing for the audience. Like not monetizing companies. Yeah, like not monetizing companies, correct. But they will eventually get us 10 years down the line to where we want to be. So that's the lens with which we look at things. And you have investors who have that patience to see you through that journey as well. Yeah, we're very lucky to have investors like that.
01:48:54
Speaker
Except for one. We were lucky to have very long term investors with a very long term lens on things. In fact, all of our investors are very clear that you have to make companies that outlast us.
01:49:16
Speaker
You have to make companies that outlast the founders in their lifetimes. So trade companies for the long term, it's okay to take decisions that impacts revenue in the short term with that lens. I'm looking forward to scalar words. I am too. I am too.
01:49:32
Speaker
This episode of Founder Thesis Podcast is brought to you by Long Haul Ventures. Long Haul Ventures is the long-haul partner for founders and startups that are building for the long haul. More about them is at www.longhaulventures.com.