by Y Combinator3/30/2018
Vivek Ravisankar is the CEO and cofounder of HackerRank, which was in the Summer 2011 batch.
They surveyed 40,000 developers on things including their favorite frameworks, what they want in a job, and how they learned to code. You can read the report here.
Craig Cannon [00:00] – Hey, how’s it going? This is Craig Cannon and you’re listening to Y Combinator Podcast. Today’s episode is with Vivek Ravisankar, and Vivek is the CEO and co-founder of HackerRank, which was in the Summer 2011 batch. HackerRank is a technical recruiting platform that assesses developers based on coding skill. You can find them at HackerRank.com. Alright, here we go. Alright, Vivek, why don’t we start with what you guys do, and then we’ll rewind to before you even did YC.
Vivek Ravisankar [00:30] – Sure. I’m Vivek, one of the founders and CEO of HackerRank. Our mission at HackerRank is to match every developer to the right job, with the underlying driver being skill. Resumes are just a very poor correlation to skills, that’s how we got started. What we do as a product, as a developer, you could come on to HackerRank to practice and improve your skills across different dimensions, whether that’s algorithms or artificial intelligence or database skills, and when you’re applying to a company, instead of going through the traditional route of uploading a resume, and it’s mostly going to be a black hole, you take a company’s coding challenge. If you want to apply to Airbnb, you take Airbnb’s coding challenge. If you want to apply to VMware, you take VMware’s coding challenge. If you have the required skills, basically, we automate the whole code and give you a report, then you get called in for an interview. It’s great for developers because it’s merit-based processes made for companies, because now you’re able to attract the right developers for that interview process.
Craig Cannon [01:31] – What might one of those challenges look like? For instance, Airbnb, are we being put into the codebase in any way? What would it look like?
Vivek Ravisankar [01:38] – Yeah, it varies. We built a platform where it’s super custom for companies to go ahead and create. You could have a simple, problem solving challenge to something close to a real world where you could have, you could give access to our GitHub repo and people can go ahead and modify the code, fix a bug on your repo to go ahead… There’s a wide range of problems that you could give on the platform.
Craig Cannon [02:04] – Okay, gotcha. You guys haven’t always been doing this.
Vivek Ravisankar [02:08] – Correct.
Craig Cannon [02:09] – In its exact instantiation.
Vivek Ravisankar [02:10] – Yeah.
Craig Cannon [02:11] – Let’s go backwards. Before YC, what were you working on?
Vivek Ravisankar [02:15] – Sure. Hari is my co-founder. We knew each other from college. We used to do a bunch of things together in college. We started the company, technically, we quit our jobs. I used to work at Amazon as a developer before this, where I did a lot of technical interviews, which is kind of where I saw the problem. The resumes were not a good correlation to skills. We quit our jobs. The way we started to do, we thought that the way to solve this problem of resumes not correlating to skills was to help prepare students for interviews. We started Interview Street. That was our first version. What it did was mock interviews. Let’s say you had an interview call with Amazon or Google next week. You could come onto our site and attend a mock interview with either somebody from Amazon, an ex-Amazon person, who can give you a walk through of, here are the things, we’ll do a mock interview and give you, what are the areas that you need to improve, where do you stand on all of those things?
Craig Cannon [03:20] – The students would be paying these people for their time?
Vivek Ravisankar [03:22] – That’s right.
Craig Cannon [03:23] – Okay.
Vivek Ravisankar [03:24] – It’s going to be hard for people to comprehend the value here in the US, because we were doing it in India, but we were charging 350 rupees, Indian currency, from the student, and we used to pay 250 rupees to the interviewer. 100 rupees goes to our pocket.
Craig Cannon [03:44] – Right.
Vivek Ravisankar [03:45] – You could almost imagine that to be something on the lines of, I mean, technically, it’s six dollars and five dollars.
Craig Cannon [03:53] – You earn different amounts.
Vivek Ravisankar [03:55] – Yeah, different amounts you could say. A student would pay $200, something like that, and an interviewer would get about $150 here. If you then think about it in that way, it was pretty expensive for students, and we didn’t have any good payment gateways or anything of that in India, so we used to set up campus ambassadors in different colleges. Hari and I used to do a trip every month to visit all of these colleges to collect money from these campus ambassadors.
Craig Cannon [04:24] – Cash.
Vivek Ravisankar [04:25] – Yeah, cash.
Craig Cannon [04:25] – Wow.
Vivek Ravisankar [04:27] – Our accounting was all pretty funny. From an interviewer perspective, 250 rupees or $150 or something on those lines here, it’s not enticing, right? If I have to spend an hour of my time doing a mock interview and giving you feedback, is it worth it? I still have the Excel which projects us to make somewhere on the lines of $100 million, or 100… whatever. It’s like, $50 million, something along those lines.
Craig Cannon [05:00] – Yeah, given the trajectory in the beginning.
Vivek Ravisankar [05:04] – Yeah, I think we made about 4,000 rupees. That’s the equivalent of $100 here. After a year, or year and a half. Clearly, you could just drag the cells in the Excel in whatever way you want and get the numbers, but it’s very hard to execute. That failed and we had a lot of lessons that we learned from that. Most importantly, the students didn’t have cash, there was a lot of scheduling, and one of the things that YC stresses on the most, I wish we had, actually, done YC at that point. This was done by Paul Buchheit. Make something that 100 users really, really love verus making a 1,000 people or 10,000 people, they sort of okay like. We should have focused on, you know what? We’re just going to focus on these two companies, mock interviews, versus any tech company, anywhere in the world, and we were scrambling for interviewers from Amazon, Google, Microsoft, all over. We should’ve just focused, you know what? These are the three companies that we’re going to focus. If you want to get a job here, you can come and do, make it really happy, and then figure out a way to scale. Looking back.
Craig Cannon [06:09] – Yeah, of course. But did people like the product?
Vivek Ravisankar [06:11] – Yeah. Whenever people did the interviews, they really liked it. In fact, there were a lot of really interesting testimonials of people who actually attended the interviews, got their job, and they knew some of the areas that they need to improve, they need to prepare, all of those things.
Craig Cannon [06:25] – Yeah, that’s great. I remember all the interviews I did during college. If I could’ve just gotten the email of someone at one of these companies, it would’ve been invaluable to me.
Vivek Ravisankar [06:33] – We used to do that, sort of, informally in college. When we were in the final year of college, we used to train people for mock interviews, this was more like an extension. Why can’t you do this on across every one?
Craig Cannon [06:47] – Interesting. You weren’t necessarily off the mark around finding something that people really like, but you found that you kind of hit a ceiling somewhere.
Vivek Ravisankar [06:55] – It wasn’t very focused either, there were different cohorts of people. Some of them were students, some of them were a couple of years into their work experience, and some of them wanted to do technical interviews. Some of them wanted to do, sort of, a automobile interview and then we had to go ahead and scramble and figure out, can we get somebody from the automobile division to come and take the interview? We had no contacts, so we had to figure out a way to scramble through LinkedIn and Facebook. It was sort of okay, but it really taught us to stay strong.
Craig Cannon [07:33] – At what point do you decide, “You know what? We have to change this, this isn’t working.”
Vivek Ravisankar [07:37] – Yes, probably 18 months. Well, we had applied to YC.
Craig Cannon [07:41] – Oh, you did? Okay.
Vivek Ravisankar [07:42] – We’d applied to Y Combinator with this idea because we got turned down. I think it was about 12, 15 months, something along those lines. I mean, clearly this wasn’t going anywhere. The Excel was completely off the trajectory of the Excel graph and where we were. Then we switched our business model, our idea, to something different. There is a big trend that is still present in India, which is, once you graduate, you have to do your masters in a school here.
Craig Cannon [08:16] – Oh.
Vivek Ravisankar [08:16] – It’s huge. I mean, pretty much, I don’t know, 80, 90% of people would attempt their GRE, which is your standard way of getting here, your TOFL, they will prepare a statement of purpose, they’ll do all of those, and the application form is pretty expensive. You could only apply to maybe five or six universities because it was so expensive that you couldn’t just apply to all the schools in the US. What we did was we connected people who were already studying in these schools who can review your statement of purpose, your resume, and tell you if you had a chance to come here, to apply to the school. That was our pivot. That actually took off a bit, but then after three months or so, the traffic was down to zero. It turns out you can only apply twice a year. We didn’t know what to do for the rest of the year. We actually applied to YC even with this idea.
Craig Cannon [09:09] – There you go.
Vivek Ravisankar [09:10] – We got turned down. This whole thing about mock interviews, applying, helping people apply for their masters, prepare, maybe a year and a half to two, something along those lines that we worked on.
Craig Cannon [09:22] – Okay, and then what sparked you to start working on where you are now?
Vivek Ravisankar [09:26] – Yeah, we were close to bankrupt. There was a real need, and of course we had really good jobs and Amazon was a great company to work for.
Craig Cannon [09:37] – Where you in India working on this?
Vivek Ravisankar [09:39] – Yeah, yeah, all of this is in India.
Craig Cannon [09:41] – Okay.
Vivek Ravisankar [09:42] – Then we realized, okay–
Craig Cannon [09:44] – Just to pause here really quickly, how are you making the connections with people here if you were based in India?
Vivek Ravisankar [09:50] – Sure. What we did was we used to take the directory of people from our school who went ahead and did masters at all of these different universities and who are working at different companies, so we used to go through the directory, keep calling people, email them, “Hey, are you interested? We are from the same university,” Put that university touch in to get them on board. It was just friends, networks, second degree. I used to make 50 calls every day–
Craig Cannon [10:19] – Wow.
Vivek Ravisankar [10:19] – To get this going.
Craig Cannon [10:22] – And just for context for me, people from India want to do a masters here because it’s a gateway to getting a job here, or what?
Vivek Ravisankar [10:31] – I don’t know how it started. Maybe it was just sort of a checkbox.
Craig Cannon [10:41] – Okay.
Vivek Ravisankar [10:42] – I’m not sure that that’s changed to now, but it was definitely, I mean, I graduated eight years back, or nine years back. It was definitely a pride, “Okay, you know what? I’m going to do my masters here.” And there’s, of course, some glamour associated with the great American dream, coming here and working in Silicon Valley. Silicon Valley is sort of this mythical place for all technology developers to be here. It was more like a checkbox. It is the case that you would have to apply. If you had good GPAs and if you had good grades, the question is, why aren’t you going? Why aren’t you applying for your masters?
Craig Cannon [11:16] – That’s funny. Even among the very talented, smart kids going to great schools in India.
Vivek Ravisankar [11:21] – Yeah, for sure. I think it’s, “Why aren’t you applying for your masters? Why aren’t you doing your PhD in Carnegie Mellon or Stanford or Purdue?” There are a lot of, actually, Hari and I might actually be the odd men out.
Craig Cannon [11:32] – Because you didn’t do it?
Vivek Ravisankar [11:34] – Yeah, we didn’t do it, and I’m happy, because I’d applied to six schools and all of them turned me down. I was so dejected when the last school turned me down, but in hindsight it’s great, because I don’t think we would’ve started the company if I’d come here.
Craig Cannon [11:46] – No, probably not. It is the conservative move.
Vivek Ravisankar [11:49] – Yes, and once you graduate, you have to get a job because you have to pay back the loan which is pretty big. You do two years of masters, then you do two to three years of working at different companies to pay back the loan and there’s all of these visa things. By the time you want to start, you’re already, I don’t know, close to five, six years invested in this.
Craig Cannon [12:10] – It’s really common.
Vivek Ravisankar [12:11] – And the friction, or the inner sheer to go ahead and start another company when you’re happy working at Amazon or Google is going to be hard.
Craig Cannon [12:18] – Yeah, definitely. So, you’re basically out of money?
Vivek Ravisankar [12:22] – Yeah, we’re close to bankrupt, yeah. We had a clear choice which was we could go back and get a job. We could’ve got that. But we just wanted to keep going, we just wanted to try this out once more. I have no idea what kept us going at that point in time. There was some conversation, we were just walking around. It sounds obvious, “Hey, if you have to make money, you have to sell something to people who have money.” Otherwise there is no point that you’re going to make money. We were selling to students and others who were already in whatever they call it, student debt, and they had very little money. That’s when we said, “Hey, can we invert this?” Which is, can we build a product for companies to help them identify the right developers based on skills? Of course, we didn’t have the mission written down, like, match a developer to the right job, but looking back, that was the core of every decision that we made. We felt, you know, we’ll still stay true to our core. Mock interviews is one way of doing this, but if you actually give a platform that can enable developers, it doesn’t matter what school they go to, which company you went, but you have a chance to showcase your skills and get the dream job that you want, then that’s great, you know? We should build a platform for that. That’s how we got started, and that got the early tractions so that was the third time we applied to YC. And yeah, YC called us that time.
Craig Cannon [13:54] – Oh, alright. Was it because you were showing significant growth at the time, or the idea was just working out?
Vivek Ravisankar [14:00] – I don’t know. I definitely didn’t want to ask, “Why are you guys calling us now?”
Craig Cannon [14:04] – Is this a mistake?
Vivek Ravisankar [14:06] – Yeah, is this a mistake, are you sure? That was our first trip to the US.
Craig Cannon [14:11] – Oh, man.
Vivek Ravisankar [14:12] – Yeah, and we had some struggles there as well. Hari, my co-founder, couldn’t get his visa to come here. I still think it’s the stance of YC, but YC cares about co-founders a lot. I got my visa, Hari couldn’t get it, so I flew in here just for the 10 minute interview. This was, yeah, six years back. Fortunately, we got in. I remember, I think it was a panel of interviewers. It was PG, Sam, Harj, Jessica, and Paul Buchheit. They were the five people interviewing me.
Craig Cannon [14:43] – Whoa, okay.
Vivek Ravisankar [14:44] – I landed in the morning, and the evening was the YC interview, because I couldn’t get my visa on time.
Craig Cannon [14:50] – You were the last day of interviews?
Vivek Ravisankar [14:52] – Yeah, last day of interviews.
Craig Cannon [14:53] – Probably the last one.
Vivek Ravisankar [14:54] – Yeah, could be the last one as well, could’ve been. You see all these five people, there’s the creator of Gmail, you’re obviously a little stressed, and then you don’t have a co-founder, and PG asks, “Where is your co-founder?” I said, “He couldn’t get his visa.” But it was a good, intense 10 minutes. Frankly, I didn’t think we’d get into YC. Maybe it worked to our advantage, because I was just completely open and honest about everything because I didn’t fake any of my answers or try to impress them. Look, this is the problem. I really want to come here and go ahead and do this. It probably worked out to our advantage.
Craig Cannon [15:31] – Interesting. What you were working on before has this separate difficulty in the application, because you have to educate someone about what it’s like in India, versus here where we have the exact same problem of, “Oh, we’re this company, we have all these jobs, we don’t have enough engineers.”
Vivek Ravisankar [15:48] – Yeah.
Craig Cannon [15:48] – “Let’s sort this out.”
Vivek Ravisankar [15:49] – There is a different way of things. You’re right, the environment is different. I used to sit in the Pioneer Way office, Y Combinator office for the most part. Actually, our first office was right opposite–
Craig Cannon [16:10] – Really?
Vivek Ravisankar [16:10] – The YC Pioneer office. Yeah, yeah, it was RethinkDB’s old office, which we actually took, so it was right opposite. I was used to being in Y Combinator for most part. I remember I took PG’s charger, Mac charger by mistake home. This was the second day or the third day. I didn’t know what to do, because oh my God, what is PG going to think? I started emailing him. “I’m really, really sorry, I took your charger.” He said, “No problem, just keep it there tomorrow.” I said, “Okay, are you going to come in tomorrow?” He said he’s not going to come in tomorrow. The way, at least, I was brought up for the most part in India is if you keep a charger or keep the Amazon box right outside the door, somebody’s going to take it away. Even the tiny things. Then, you have to learn about how the environment is different, how people talk, how people react to it. The dinner time here is 6:30, so I remember sending a calendar invite for a dinner to a person that I wanted to recruit for 9:00 pm, and he said, “That’s kind of my bed time, so I can’t meet you at 9:00 pm.” There are all of these things that you have to learn, and it’s not just the macro part of the developer, how it works, but even the tiny things about how you interact, how you work with people, and all of those things you have to learn.
Craig Cannon [17:30] – Yeah, has that been incorporated into your product? I imagine you have all these international people applying to US companies. How do you help educate them on the cultural norms?
Vivek Ravisankar [17:40] – Yeah, it is a really good question, because right from how you, sort of, the types of sentences or words that you use varies quite a bit between what you would say to somebody in Asia, what you would say to somebody in UK, and what you would say to somebody here. All of those nuances, we need to take care of. There is a different cultural aspect on how you sell, how you talk, how you get people going here.
Craig Cannon [18:11] – It’s really difficult. How long did it take you in particular?
Vivek Ravisankar [18:14] – It’s a forever learning. I don’t think I’ve perfected it. I’ve gotten my dinner time now at 7:00 pm. I’ve shifted a little. But I think it’s a constant learning. I don’t think it’s ever going to stop.
Craig Cannon [18:27] – My friend moved to India for a couple of years and he explained the exact thing opposite. It was tough for him.
Vivek Ravisankar [18:32] – It’s just the way you grow up is different. There’s nothing that’s right or wrong, it’s just different, and you have to get used to it. I remember PG and Sam correcting me from everything to, even the tiny ones. In India, you would probably say programmer. You would add the A at the end, but here, people will not be able to follow. It’s pro-grammer.
Craig Cannon [18:58] – Oh.
Vivek Ravisankar [18:58] – If you have to say, and when I was doing my demo day pitches, all of these little things mattered because it was just a five minute pitch, no, I think it was a two minute or a five minute pitch in the demo day, and every little thing matters a lot when you’re talking to investors. It’s not just the accent, but how you talk, how you schedule meetings, how you follow up, and everything is different.
Craig Cannon [19:21] – Yeah, demo day is tough universally. You’re pitching amidst a group. I think you said your batch was, like, 60 or 70?
Vivek Ravisankar [19:28] – Yeah, I think our batch was 63 or 65.
Craig Cannon [19:30] – Yeah, so that’s enough companies for one to just fall into monotony.
Vivek Ravisankar [19:36] – But interestingly, I really enjoyed it. I’ve never been exposed to such a large group of investors. I couldn’t even sleep the previous day. I was looking forward to it. I wanted to pitch and get a Series A going. I frankly enjoyed it, I was looking forward to it.
Craig Cannon [19:53] – That’s very cool. You raised money after YC–
Vivek Ravisankar [19:57] – Yeah.
Craig Cannon [19:57] – And you’re committed. You’re like, okay, we’re going to stay in California at this point?
Vivek Ravisankar [20:01] – Yeah, I mean, after I came here to Mountain View, I don’t know, maybe that’s partially the reason. The first city, well, I technically now live in San Fransisco, but you take the cab and get to Mountain View, I think everything looked great for the most part. Right from the people that you interact with, the advice that you get, the weather, the food. Sort of the perfect mix for me coming from India. You had the weather similar to Bangalore and the food similar to Bangalore and Chennai, and then you had all the people that I really wanted to talk and get advice from in terms of building a company. It was the perfect mix, at least for me, so I was very much determined to stay here and build a company.
Craig Cannon [20:49] – Man, what was that like flying here and then seeing all the people who you read their blog posts. PG is an easy example.
Vivek Ravisankar [20:54] – Yeah.
Craig Cannon [20:56] – What was that experience like?
Vivek Ravisankar [20:58] – Yeah, it was kind of surreal, I didn’t know it was… As I said, I thought I was going to just stammer and flunk the interview, but like I mentioned, I didn’t think we’d get in. Actually, we were the first Indian based company to get into Y Combinator.
Craig Cannon [21:12] – Really?
Vivek Ravisankar [21:12] – Yeah. That was another quote unquote, I don’t know if it helped or didn’t help, because we thought, okay, they’ve never chosen a completely Indian-based company, so our odds are low. That really helped, because I was just completely free and open, but it was great. It was great to meet all of the people that you admire from far off from reading their blog posts or Twitter or watching their videos.
Craig Cannon [21:34] – What were your greatest learnings from the experience?
Vivek Ravisankar [21:37] – From YC?
Craig Cannon [21:38] – Yeah.
Vivek Ravisankar [21:42] – The intensity was very, very high. There was a dinner every week. I noticed that each founder had their own set of people that they interacted with. Of course I interacted with a lot of them, but you talk to a few people, and then you hear their updates every week and you want to push harder. You have a fixed deadline, three months. Three months, you have a demo day, which you have to go and present to investors, so you have a fixed timeline to make sure that you have a good prototype, you have good possibly paying customers to go ahead and do it. The bar is continuing to increase at every demo day. The intensity, I think the learning is… I still think Hari and I have that intensity for most part, which is good.
Craig Cannon [22:29] – Wow, how do you keep it up? ‘Cause people talk about this all the time. They’re like, “It’s so cool having a cohort that you can kind of benchmark yourself against.” But once you’re out, you’re kind of out.
Vivek Ravisankar [22:41] – Yeah, I don’t know. It’s DNA, or we’re just consistently hungry to win. The challenge is, can you extend that across the entire company? I don’t think we’ve done that. We’ve done an okay job at that. That’s something we still need to get better at.
Craig Cannon [22:59] – That’s a challenge. So, to fast forward massively to now, the reason we started talking about doing the podcast in general was that you guys released this developer skills report. Just for folks who haven’t seen it, can you explain it?
Vivek Ravisankar [23:13] – Yeah, sure. We have a developer community that’s over 3.4 million as of last week.
Craig Cannon [23:22] – Wow.
Vivek Ravisankar [23:22] – That’s slightly over 10% of the world’s developer population who come to HackerRank to solve challenges, to practice their skills and get better. One of the things that we wanted to do was to learn more about our developers, so we sent a survey. It actually had 40 questions. There was a lot of debate. Firstly, will developers–
Craig Cannon [23:43] – It’s super long.
Vivek Ravisankar [23:44] – Then, you’re sending a survey to developers with 40 questions, something along those lines. What are the odds that somebody is actually going to fill this? We sent this to a million of our active developers.
Craig Cannon [23:56] – Whoa.
Vivek Ravisankar [23:59] – It was amazing. We got close to 40,000 people completing the survey. That was huge. I think Stack Overflow had 60,000 and they have 20 million developers or so, so if you compare the ratio, the engagement and how many people completed the survey was giant. Hopefully we can beat Stack Overflow’s number next year.
Craig Cannon [24:20] – Yeah, what was your email subject line? How’d you get so many people to convert?
Vivek Ravisankar [24:24] – That’s a good question. We did A/B tests, but maybe our marketing team, they might have a better idea on what was the subject line. But we did A/B tests, a few things. We also tested on the timing. You know, when to send to which cohort depending on the country, depending on the geography–
Craig Cannon [24:44] – Because you’re fairly distributed, right, in terms of your user base?
Vivek Ravisankar [24:47] – Yeah, maybe it’s about 26 to 28% in North America, about 40% plus in Asia, Asia Pacific, and then you’re talking about 15 to 20% in the UK, and the remaining is just–
Craig Cannon [25:01] – Just wherever.
Vivek Ravisankar [25:01] – Yeah.
Craig Cannon [25:02] – Okay.
Vivek Ravisankar [25:03] – For each of the geos, we also optimized the subject line, when to send them, and how do we actually position this? We did a bunch of things on those lines.
Craig Cannon [25:11] – Was it dialed in for language or all in English?
Vivek Ravisankar [25:13] – It was all in English.
Craig Cannon [25:15] – Okay.
Vivek Ravisankar [25:15] – Yeah. I don’t think we changed any language. The questions were varying from, when did you learn to code to what do you look for when searching for a job, or if you’re a hiring manager, what do you look for in a candidate? We had a wide variety of questions, and some really, really interesting insights. We engaged a firm who could parse the data, who could do real analysis on this, and we’re also likely going to put this up on our website for all of our developers to go ahead and do the analysis with our raw data, so something might happen maybe in the next couple of quarters. It was a really interesting insight, and we published this earlier this year. We got over half a million developers to read the survey. This was a huge milestone and exceeded at least my expectations in both the quality of the content, the design, the reception that we got, and of course, this podcast wouldn’t have happened– It exceeded my expectations in all dimensions.
Craig Cannon [26:17] – Alright, I guess the most important question is, what programming language is the most popular?
Vivek Ravisankar [26:23] – It varies across different industries. The one that people want to learn a lot is Python, and the one that’s starting to get a lot more popular is Go. But it varies across different industries, from what financial services want versus the fast growing internet companies, versus retail. All of these companies are transforming into software companies, so each of them have a different variation of it.
Craig Cannon [26:51] – Okay, so for a developer who’s curious and sees this survey, do you funnel them through a flow on your site, like, “If you’re interested in getting into banking, you ought to be learning X?”
Vivek Ravisankar [27:02] – Yeah, exactly.
Craig Cannon [27:03] – Is that how it works?
Vivek Ravisankar [27:03] – It’s a good question. We are trying to do the rivers which is, what do you want to learn? What language and skills do you want to learn? Then, we’ll help you recommend, where can you most fit? Where does HackerRank see the best fit or the best job for you based on all the data we’ve collected?
Craig Cannon [27:21] – Okay, gotcha. Let’s go through some of the facts, some of the fun facts.
Vivek Ravisankar [27:24] – Yeah, sure, sure.
Craig Cannon [27:26] – I saw in the beginning the report, it’s a lot of stuff about age.
Vivek Ravisankar [27:30] – Yeah.
Craig Cannon [27:31] – When do people start learning how to program, when does it all begin?
Vivek Ravisankar [27:34] – The most fun insight was one in every four developers learned to code before they learned to drive. That was a really interesting insight we were able to get. It’s not super surprising for me as a developer, but when we announced this, I think it was the most popular insight that we even started printing t-shirts–
Craig Cannon [27:55] – Really?
Vivek Ravisankar [27:55] – In our company. That was an interesting one, and it just shows the fact that self-learning is starting to become big. People are learning on their own. You don’t necessarily need to go to a particular school to do your four-year degree and then understand how to code or how to build applications. It’s just going to, I think, if you extrapolate the trend line, I think this will be two in four, three in four developers will learn to code before they drive in the next decade or so.
Craig Cannon [28:23] – Did you collect any data around bootcamps, coding bootcamps?
Vivek Ravisankar [28:27] – Yeah, we did, we did collect. We collected data on how much the hiring managers or others valued the bootcamps and kind of learning. It was mixed, to be frank. There were certain hiring managers who valued it a lot, and certain hiring managers who didn’t. My guess is it probably depends on the quality of the bootcamp. There are a certain bootcamps who say, “You know what? Come here, it doesn’t matter whether you know how to code or not. I’ll help you become a great React developer in 90 days.” That’s not how it works. It almost seems like a negative signal sometimes which is, “Oh, you didn’t know how to code 90 days before and now you’re claiming that you’re a React expert? That puts me off.” It depends upon the quality of bootcamps, and on the other extreme, you have this school which I was personally really impressed by. It’s called 42. It’s a school, actually PG tweeted about it.
Craig Cannon [29:25] – Is that the one in Fremont?
Vivek Ravisankar [29:27] – That’s right, they just opened another one in Fremont where I had a chance to visit. You, of course, have to take a minimum level of coding proficiency challenge or test, and then you’re allowed to explore whatever types of computer science topics that you really like, and then figure out which one you have a natural attraction or a passion for, whether it’s security or machine learning or a frontend developer, and then continue to get better at that, and then we’ll help you connect to the right job. That seemed to be a different approach.
Craig Cannon [30:00] – But it’s massively different, right? If I understand it correctly, they’re giving people room and board.
Vivek Ravisankar [30:05] – That’s correct. And it’s free.
Craig Cannon [30:07] – And it’s free, and it’s over the course of years.
Vivek Ravisankar [30:10] – That’s right, that’s right. It is very different. It’ll be interesting to see how that experiment works. It was started by a big billionaire in France or something who just wants to help the world. I’m not sure how it might work from a business value proposition, but at least in terms of a net improvement to the world to create more developers and to help improve technology, that’s a great start. That’s why the bootcamps are a mixed one from the hiring managers at least.
Craig Cannon [30:42] – Okay, and then what are the differences between startups hiring and big Amazon-type companies?
Vivek Ravisankar [30:48] – The one common thing was everybody stressed on the importance of problem solving skills. Irrespective to whether you were a smaller company or a bigger company than others, the one thing that the startups or the execs cared about the most, because at least in a startup, you may have a recruiter or not, but most likely you’re doing the recruiting on your own, they cared more about your contributions, your open source contributions or your profiles or portfolio much more than what a large company cared about.
Craig Cannon [31:24] – Interesting.
Vivek Ravisankar [31:26] – But at the heart of it, people wanted strong problem solvers, and also one of the things that was present in the survey was how you wanted to dive in deeper on the knowledge of that particular stack, at least for a startup, because of course, you have very low or pretty much no bandwidth to train somebody on the new stack that you’re working on versus a large company where you have a little more bandwidth to go ahead and train, so you’re okay with a generalist who can come and learn a new language, versus in a startup, they like you to know the particular language or the stack before you come in. Those were some of the nuances of the differences that we can see between a startup and a large company.
Craig Cannon [32:02] – In terms of structuring a job offering to someone, or rather just a job posting, what are the things that end up appealing to developers?
Vivek Ravisankar [32:11] – This was very interesting. I assumed differently, and of course, always data over opinion. One of the things that we learned was how work/life balance beats all of the different perks that people offer, whether that’s free food or we have a ping pong tournament every week or things along those lines. I actually didn’t believe this data firstly. I wanted to do another survey for people who did the work/life balance who actually opted in, what would you really mean by work/life balance? What we learned was a lot of developers have their own projects that they want to learn, that they want to build on the side. The environment that you’re working should give a good amount of balance and probably even support if you can do on the things they’re learning on the side or they’re building on the side versus what they’re doing for the company. That was kind of an interesting insight about work/life balance, and one of the things that I like to do at HackerRank and we’re still discussing how you position this in the right ways. If you have any of the side projects that you’re working on, we’ll fund the infrastructure cost for that. We’re still figuring out how to position this, how do you give this away? I’m pretty sure after people hear this podcast they’re going to email me, developers at least. But that’s one way of helping or encouraging people to do that as well.
Craig Cannon [33:40] – Oh, interesting. What are the other ways that companies signal that they have that work/life balance figured out? Or is it just folks come in for the interview, they get a sense for it, and then they say yes or no?
Vivek Ravisankar [33:51] – Yeah, that is one, but I also think if you’re actively contributing to open source, whether that’s something, for example, we are improving a lot of the open source code editors that are there right now. We’ve not done a great job in letting the world know about all the improvements that we’ve been doing, but that’s one way, and of course, interviews help a lot in understanding more about the developer culture.
Craig Cannon [34:14] – Okay, and do you guys have any rules around working on side projects? At larger tech companies, there are totally issues and I don’t think they really flex that muscle very often, if ever. But companies like Apple are infamous for not letting people do side projects.
Vivek Ravisankar [34:32] – That’s probably one of the advantages when you’re small. I don’t think we have any rules. I don’t think we should put any rules, because it’s just freedom of creativity to go ahead and do it. Frankly, if it’s going to be conflicting with your core idea and you’re scared about it, maybe you should take the developer’s idea and implement it. Maybe it’s better. Or otherwise, you don’t need to be scared, and you can just have the developer work on the projects that they want. I don’t think we should put any rules, but I agree. A large company like Apple or Amazon or Google will need to put in constraints. Although Google hacked this by putting 20% rule, where 20% of the time you can work on your projects. I think that’s how Gmail was born, where Paul Buchheit worked on his 20% free time to build Gmail. That’s how they hack it. It’s not really, technically a side project that you do outside. You use all the Google libraries and eventually contribute to Google, so that’s one way you could try and do.
Craig Cannon [35:37] – Interesting. What else is attractive to a developer?
Craig Cannon [36:17] – What else did you learn about how developers are actually educating themselves? Obviously, some percentage are going to bootcamps, some are doing CS, what is everyone else doing?
Vivek Ravisankar [36:28] – It was really interesting. Even though you had a CS degree, a lot of developers said they actually learned a lot from YouTube. Or books, or sites like HackerRank. That’s not a plug for HackerRank, even though I think the title might say this or whatever. That was really interesting because that’s a new, that’s a paradigm shift, right? A paradigm shift in recruiting. If you look 10 years back, you’re going to use GPAs and universities and the companies you worked at as proxies. But now the proxies are changing. It doesn’t matter if you went to a great school and had a great GPA. If you learned on your own and built a lot of things on your own, that is what is needed. That’s all that’s enough. The way that you look for a candidate, who do you look for, how to assess, how do you get people in the door, and how do you remove biases that’s been there for, not 10 years, it’s been there for 100 years. You go to this school, you do work in this company. How do you remove those biases and truly focus on the skill? I mean, it’s a lot of work. But that’s a big paradigm shift, and it’s a very good and healthy trend. If more developers can learn on their own for free, you don’t have to pay $100,000 for tuition, it’s just good overall for the world to do that.
Craig Cannon [37:49] – Interesting. How would someone stand out? What are the kinds of projects that end up appealing to a company? I know if you do a lot of these bootcamps, “Great, you made your own, CMS. So did 10,000 other people.” Did you get any insight as to, you know, when someone does care about these projects, what are the kinds of projects that appeal to companies?
Vivek Ravisankar [38:09] – Yeah, I think it’s hard. There’s one part of it, which is your motivation or your drive, which can really come through. You’re right, if you just forked a repo and made a couple of edits in one file, you can’t claim that you actually contributed to open source or anything. One is definitely, it’s the drive. The second is also, it’s probably hard to just look at a HackerRank profile or a GitHub or any of the profiles. You can’t just look and figure it out. The depth of thinking, and what are all the things that this developer wants to do more? That’s something, probably, you can get from the interview. But the bigger question really is, would you be willing to bring this person on site and go through the interview process based on their skills and not on their pedigree? It sounds obvious, but we’re trying to educate our customers and everybody who are hiring developers, this is the right way to go. We’re sort of leading and pushing the movement for this. Hopefully we can be successful in that.
Craig Cannon [39:21] – Cool. How does this affect a situation where you’re hiring remote developers? Are there any other particular cues that you pay attention to?
Vivek Ravisankar [39:30] – The communication aspect is always a big deal. How do you get everybody in sync? I think some of the companies have done an amazing job. GitLab, it’s a YC company. I talked to their CEO and it’s fully remote, it’s crazy. It just shows how much effort you need to put in to make sure that people are in sync with each other. You’re seeing this happen a lot. GitLab is just an example, but I know a lot of other companies who’ve started. I think the future would be an ability for you to build out engineering teams wherever you want, not just restrict yourself to the Bay Area. A good common connective tissue that makes sure that things don’t break because of bad communication or because of time zone differences and others. I think that’s where the world is going to move, which means ability to know their skills and how they can contribute when not in the office and all of the others is going to play a very important part. Ability to understand both of developers is going to play an important rule.
Craig Cannon [40:40] – Have you guys conjectured or started working on anything that might be able to test someone’s efficacy as a remote developer? It’s one thing to have the technical chops, it’s another thing to be able to work in a coworking space or in your bedroom or whatever when you’re in a different timezone and get shit done.
Vivek Ravisankar [40:58] – There’s one part, which is the IQ. How good are you in all of these different skills, where do you need to improve? And the other part is the EQ, which is what your personality is. We’ve been doing some experiments or partnerships with a couple of them, but we’ve not had a real breakthrough on that yet. I do think that is going to be helpful. The reason I’m saying that is, there are a lot of companies who use HackerRank to build out remote teams in India and other parts of the globe. There is a company which you’re probably familiar, Booking.com, one of the most popular travel companies in Amsterdam. They’re able to hire developers all over the globe using HackerRank, and they do it in a very efficient way because they understand the skills, strengths and weaknesses before you come on site and before you fly to Amsterdam. We’re starting to see this movement happening in a lot of companies on the IQ part of it, and now if we can do a better job or if you can work with a partner or if you can do something even internally at HackerRank, to do an EQ, it can be a very strong solution.
Craig Cannon [42:00] – This is all done through programmatic exams? There aren’t any in person interviews to assess EQ?
Vivek Ravisankar [42:07] – There are some companies who are working on this programmatic way.
Craig Cannon [42:09] – Yeah.
Vivek Ravisankar [42:11] – It sounds very interesting. We’ve not partnered with them deeply to really understand how well it’ll fit in, but it’s certainly a very, very interesting opportunity.
Craig Cannon [42:25] – Okay, so in addition to maybe figuring that out in the future, how else do you see paradigm shifts happening for developers in the future? Obviously they’ll be all over the world. What else is going to be happening in five years?
Vivek Ravisankar [42:38] – There are a couple of things. One is, I think it was there on Hacker News. As much as there are APIs and different levels of abstraction that have made it easier to get things started, to set up your environment, and this is for somebody who’s a newbie, who’s just learning to code, it’s very intimidating to set up something and get it running and compiling. That’s one part where things need to get better. Here’s an environment, here’s where you can go ahead and start to code, because a lot of them give up.
Craig Cannon [43:19] – Oh, yeah.
Vivek Ravisankar [43:19] – Or, I did this thing online, and then I tried it out on my local, I didn’t work because I don’t know if you have this library and there was a dependency failure and you have to go and figure it out and it just becomes a big mess and you give up on the spot. We still have to do a better job if we have to increase the number of developers in the world. You can’t have people giving up after their third lesson or the fifth lesson and saying, “It’s not taking me anywhere.” And also, there needs to be a purpose. I mean, the purpose can’t be you’re going to get $100,000 job. The purpose has to be much more, whatever is in your mind, you can go ahead and build it on your own. You have to encourage the fact that you can build versus I’m going to get you a six figure salary. That’s one part that needs to get better if we really have to create more developers in the world.
Craig Cannon [44:09] – Why does that have to get better? Because $100,000 salary, I agree with you, but to be devil’s advocate on this, that’s a lot of money to a lot of people. Why does that not work?
Vivek Ravisankar [44:18] – It will work for a subset of people, obviously, but in a very longer term, the really great developers are people who did it not because they’ll get a job but because they really wanted to build something, they really wanted to get whatever idea they had in mind out. I worry that if you had, this is just my hypothesis, if you had a monetary value associated with it, you might just focus a lot on that and you’ll reach a ceiling pretty soon. This is just my hypothesis. Versus, if you really fall in love with the idea of building and creating new things, the ceiling is sort of infinite. There’s always new technologies, there are always new things that you can go and build. It’s more about, what is the type of behavior that you’re trying to create? It’s like a kid, you’re trying to get into the developer world, trying to do it. What is the type of behavior that you’re trying to create and why are you doing what you’re doing? That should be taken care. The second big bucket is really the paradigm shift that I was talking about, which is, more companies should come forward and encourage and say, “We’re about skills, we don’t care about resumes, we don’t care about what school did you go to. As long as you have the skills, we’d be willing to.” I still think the proxy of your GPAs and companies exist at a lot of companies.
Craig Cannon [45:47] – Yeah, absolutely. I think that people constantly disqualify themselves before they even enter a race, and we see it with YC all the time. People are constantly asking questions like, “I raised money, can I do YC?” Yes. “I haven’t raised money, can I do YC?” Yes. “I don’t have a co-founder, can I do YC?” Yes. You often encounter this thing where it’s like you can just apply, you can just go for it. Moving things towards a more meritocratic system, that makes a huge difference.
Vivek Ravisankar [46:17] – Yes, yeah.
Craig Cannon [46:17] – What else does a developer value in the hiring process? Not just about a company, what do they want?
Vivek Ravisankar [46:24] – The counterintuitive thing, at least whenever I’ve talked to a lot of companies is, you should also understand the developers are interviewing you. It’s not just a one way thing. Developers are interviewing you in a bunch of different things from the process, how tight and efficient are you? Are you actually doing…? Their peers? And also the challenges, or the questions you’re asking and how deep. Sometimes, I’ve seen this in a lot of companies, people are scared to ask hard questions because it might turn off. It might work the other way, which is, I’m wondering, is this the level? I mean, if you just ask the easy questions are these the kind of peers–
Craig Cannon [47:03] – Oh, man.
Vivek Ravisankar [47:06] – Of course, there’s a different nuance to the hard question, right? You don’t want to ask a complex math problem which has a trick answer and if you didn’t know this particular formula, you can’t get it. The hard in the sense of relevant to the business, relevant to the job and how you go about doing it. The other thing that I’m starting to see happen across a lot of different companies and even HackerRank has incorporated this is the business acumen, the level of business acumen that you have as a developer. Usually it is always developers are builders, you have this whatever requirements or spec, you can go ahead and do it. But if you look at a lot of the great companies like Google or Facebook or others, they’re all developers at the heart of it. The reason, of course, they had great products, but they also had really strong business acumen. Really great developers have that. Steve Jobs incorporated this, the 10X developer has that level of very strong business acumen. That’s something that people want to know, what your strategy is. People want to know how they can contribute, how what they’re building helps in the business strategy. Oftentimes we think it’s just about users who are using, MAUs, and others, but no. I like to know how we’re building a company. That’s something that we’re trying to do a lot, and it’s something that we should do in the interview process as well. That’s my recommendation.
Craig Cannon [48:26] – Interesting. In other words, ask questions about product, not just technical stuff.
Vivek Ravisankar [48:30] – Yeah, and about the business strategy. You should be able to challenge HackerRank or challenge the company on why are you doing this? Here is a different idea, here’s another thing that you can do better.
Craig Cannon [48:42] – Oh, man, now you’re just stressing people out. More stuff to prep for.
Vivek Ravisankar [48:46] – But I think it’s fun.
Craig Cannon [48:48] – I like it, but yeah.
Vivek Ravisankar [48:49] – You can just sit in the, whatever, the armchair critic, you know, you can say, “Why did you enter this geography? I think that’s a bad move.” Or, “Your business plan addition is conflicting with your free user addition.” You can have a good conversation. Frankly, all of these things, the nuances are super important. It doesn’t matter the questions that the candidate asks, whether it’s the right or the wrong thing for the business. You have to judge the quality of thinking. It could be completely off from your business strategy that you would actually implement, because you have way more data and you’ve been doing this for six years or something on those lines, but it’s the quality of thinking, what’s the framework of asking questions that you should be judging people on? I’ve found those questions in developers really be at a much higher productivity at our company.
Craig Cannon [49:47] – What about the brain teasers, are they still in fashion? Do developers like them at all?
Vivek Ravisankar [49:52] – No, I think we shouldn’t be using that. I don’t know who created that.
Craig Cannon [49:56] – I don’t know.
Vivek Ravisankar [49:56] – Some company created that. Clearly I’ll have to go back and look at the history.
Craig Cannon [50:01] – It’s probably McKinsey or something.
Vivek Ravisankar [50:03] – Could be. The underlying goal is the right one, which is basically we are trying to think about the computational thinking, which is, I don’t know, how many tennis balls will fit in a Boeing 747? That’s the question, they’re trying to figure out the quality of your thinking. How do you think about it, what kind of data that you’ll analyze, how do you structure the problem, how do you break it down into sub-problems and go ahead and do it? There are better ways that you can assess the computational thinking or the problem solving skills, which is interesting, because that’s what the hiring managers care about the most, in a much more developer friendly, coding friendly way of helping people do that.
Craig Cannon [50:50] – Cool, okay. In closing, I’m curious about your experience as a founder hiring developers. We’ve covered a bunch of stuff, if you’re a company, this is how you hire your 1,000 and first developer. But if it’s just you and your co-founder and you raise a little bit of money or you’re making enough money that you can finally hire someone, what are the things you’re thinking about? What are the things you’re testing for early on?
Vivek Ravisankar [51:16] – Peter Thiel talked about this, maybe in one of the podcasts or one of the interviews, which is the way PayPal got built. They went and hired people who the others thought weren’t qualified, but actually were really, really good. They didn’t have the traditional or the traditional background in doing it. The first 20 people or so at PayPal were very much on those lines. That’s kind of how we at least built the first five to 10 hires of developers, and it’s really helped us a lot. If you look at it, Hari and I didn’t go to the IITs, which is the equivalent of Stanford. The first five to 10 developers that we hired were people who are very smart, who are very skilled, and it’s almost this quote unquote untapped pool. It’s a secret source that we have. Then, that really set everything right, from their motivation and drive to do great things. The first engineers are still with our company doing amazing things.
Craig Cannon [52:16] – Wow.
Vivek Ravisankar [52:18] – They’re doing some great things right now and leading a team, managing a team. The motivation and the drive to do really great things, the ability to learn, the ability to assess, making sure that the peer of whoever is going to join the company is as good or better than them. All of that has helped a lot. I feel that’s an important thing. I come back to this whole proxy thing. The current ways of doing proxy on GPAs and degrees is bad, and you have to figure out your firsthand people almost from the untapped pool. It helps a lot in your drive and motivation for the long term. That’s just my hypothesis, but I just have one data point from HackerRank.
Craig Cannon [52:56] – I agree with you. The short answer is that you have to do the work to find the great people, and it’s in that arbitrage of you don’t have the right credentials but you have all the skills, so you’re worth it.
Vivek Ravisankar [53:08] – Yeah, also at some level, people should be really passionate about the mission. Again, the nuance comes here, right? You can’t expect somebody to be more passionate than you. That’s never going to happen. But as long as they have a real feel, or connect, you know what? We need to solve this problem, and here are some other ways that you can actually do it. That’s another thing that you need to test for the first developers.
Craig Cannon [53:35] – Awesome. Alright man, well, thanks for coming in.
Vivek Ravisankar [53:37] – Yeah, thank you for having me.
Craig Cannon [53:38] – Alright, thanks for listening. As always, you can find the transcript and video at blog.ycombinator.com. If you have a second, it would be awesome to give us a rating and review wherever you find your podcasts. See you next time.
Y Combinator created a new model for funding early stage startups. Twice a year we invest a small amount of money ($150k) in a large number of startups (recently 200). The startups move to Silicon