Jasjit is a serial entrepreneur with three start-ups under his belt. He's also done stints at Bridgewater and Redfin. He's currently a General Manager with Thinkful, which is on a mission to de-risk education and close the skills gaps in tech through a unique 1-on-1 mentorship platform.

Follow Jasjit on Twitter


Episode Transcript

J Cornelius: What happens when your business model aligns company incentives with costumer incentives? Well, the online code school Thinkful did just that and are reaping the rewards. We'll get the details on how they did it coming up next.

Intro: This is Design Driven, the podcast about using design thinking to build great products and lasting companies. Whether you're running the start up or trying something new inside a Fortune 1000. The tools, methods, and insights we talk about will help you create things people love. And now, your host, J Cornelius.

J: Hey everybody, J from Nine Labs here, coming to you from Atlanta. We're here with Jas from Thinkful. Excited to have him on the show today.

Hey Jas, how you doing?

Jasjit Singh: Hi, how are you?

J: We're fantastic. So what's happening at Thinkful? Tell people what Thinkful is and what you're doing there, what you're excited about?

Jasjit: Sure. We're pretty pumped about ... So we're new to Atlanta, but we have students from across the country. Basically what we're doing is ... We're a new kind of vocational, education company. We're looking to close skill gaps wherever and whenever we see them occur. I think what makes us really special and unique is that we're completely built around a one-on-one mentorship model. So, students who come and work with Thinkful and our programs, they have an individual mentor and the research has shown that people advance more quickly through material and they retain more when they have a mentor, but it also enables a lot more exciting things that I think are pretty revolutionary in vocational education. It allows us to lower our costs and provide access with both the interns and the programs, and in terms of who we can reach and in terms of geographies. And I think the most exciting part of our model is that we align our incentives with our students. We have a tuition guarantee for our web development program, so if people don't get a job within six months of graduating, they get their money back.

J: Alright.

Jasjit: So basically, we lose money every time a student doesn't get a job, which I think a good thing. So everything about the structure of our company is about increasing job placement and the more people we place the more money we make, and right now we have a 93% job placement rate. We are one of only a few of vocational schools that have all our job and outcome data audited by a third party. So, it's something that we're really proud of and we hope and feel like this might be the future of vocational education generally.

J: Yeah, that's interesting that you've set up the model to really keep the company incentivized to make sure that your people get jobs afterwards.

Jasjit: Exactly, yeah. If you think of the traditional school from college onwards, the school makes money when they get you to enroll and then the student is going through that experience of an education, which is already a pretty difficult experience. Learning something new is challenging and frustrating, but then you have this weight on you of "Am I gonna actually get a job afterwards? Is this actually going to improve my condition?" And we've seen a lot of new stories about escalating student debt and people who have $200,000 of loans and maybe finding a $35,000 a year job, and the math just doesn't work.

J: Yeah, that's a big problem.

Jasjit: We're trying to flip that and I think that our model is a big part of that. Right now we have a web development program and then a science program, but like I said, our objective is to go and find those skill gaps that exist and are also accelerating with sort of advances in technology, and build programs through one-on-one mentorship model that close those gaps systematically.

J: Yeah, that's cool. So it sounds like ... We talk a lot about human centered design on this show and it sounds like your business model from the outset is putting human centered design like right in the center of it, right? So you're focused really on the outcome of your customer.

Jasjit: Absolutely and I think- Sorry, go ahead.

J: No, I was just saying you're obviously capturing some value for the business along the way. How's that working out?

Jasjit: It's going really, really well. You can imagine students kind of take a double take when they hear something like a tuition guarantee. It's something they haven't heard before and it's definitely compelling and changed the game when it comes to getting students to enroll. From that perspective it's going really well.

We started in Atlanta only a couple months ago. We already have about forty students. We have twenty mentors who are local, who are with companies like Delta and Home Depot and Hurst Auto, who are working with our students. We're starting to see a community here in Atlanta, and I think over seven thousand students have graduated nationally so it's not small company. But, we're definitely looking to have more of a local presence in a lot of our markets and add more of a component to it, and Atlanta's kind of ground zero for some of those experiments.

J: Yeah, cool. So as you think about what segments to go into and help people learn, where are you getting that research? How are you finding those gaps?

Jasjit: I think that, obviously the first theory is we just look at where the shortages are, where employers are trying to hire people and aren't able to find them for whatever reason. I think there's also some kind of seismic shifts happening at the macro level. I mean, I'm sure a lot of people have heard the phrase, "software's eating the world" and sort of the massive shift that's happening with sort of how software’s replacing so much of business processes. That is creating a lot of shifts. One of those is obviously a demand for technical roles in web development, but you have things like Big Data that are coming down the pike. And other types of roles like coding plus roles, like coding plus marketing or coding plus business development that are coming down the pike and we're pretty excited about some of those things and we think that a lot of the existing infrastructure around education has failed to prepare students for these types of roles and they're training so rapidly.

So a lot of these other companies, or a lot of these other institutions can't be responsive to those needs and we can. So we try to take a broad view of where we see the shortages are and where we see the macro environment in terms of demand and try to make that around where we think the market dynamics work out for a model like ours.

J: Right so, we go through this step process where you validate that there's a problem that exists and you've done that, and then you have to validate your solution as well and it sounds like you've done that. I'm curious like, you can get employment trends from a lot of different places, so are you also going into companies and asking them specific questions about where the gaps are or what skills they need to fill?

Jasjit: Partly. One of the subtle features of our model that I think is actually pretty important ... Because we have the one-on-one metric program, that means we have hundreds of mentors and these mentors are not instructors full time. They're actually working their day jobs in the field, in the industry. So in a lot of cases, they are our feedback loop for what the trends are in the industry. They're on the front lines of any cases. So they tell us not just where the opportunities are for what their teams are hiring for, what their companies are hiring for, but they also help us update the curriculum on almost a daily and weekly way to make sure that we're teaching people the things that companies actually need. You have this old model of education where it takes like five years to get the curriculum approved and by the time students are graduating from it. It's completely outdated. The company has to hire these people and teach them everything new anyways. Well, with our model, because our teachers are the people who are in the field, we kind of flip that on its head as well.

J: What kind of tracking or tools are you using to validate that you are approaching things the right way or answering the right needs and solving the problems?

Jasjit: In terms of tools, do you mean kind of software tools?

J: It could be exercises or ... I mean, software tools, sure, but I'm thinking, you know, everyone is pretty familiar with the Business Model Canvas at this point, so I'm sure you're using something like that. Are you also using like Empathy Mapping or the Value of Proposition Canvas or any of those kind of tools to kind of codify what problem it is you're going to solve and how you're gonna solve it?

Jasjit: Honestly, no. We aren't using those tools. I think there's a general philosophy at the company of rapid experimentation and Lean Startup approaches where we have a sort of hypotheses and we develop an experiment to test them and iterate that experiment forward. For example, I'm kind of working on local markets, trying to build a playbook for our local markets, and developing a personal experience. We have some key metrics that we care about, not just about student recruitment, but also student retention, completion and, obviously, job placement. We prioritize those goals, then we try to run as many experiments per week as possible and be very open, honest with each other about “Are those things working or are they not?” and iterating as fast as possible. I don't know if that follows a particular framework, but we're all pretty inspired about the Lean Startup Method and we try to incorporate that as much as possible.

J: Sure, and that's a great kind of overall methodology to follow and there's a lot of different ways you can approach that. We use a tool called an Experiment Grid, which kind of came out of a similar approach and it sounds like you're doing something very similar. You've got a hypothesis and you test it and if it works, you measure that and if doesn't work, you measure that. Every one of those tests informs the decisions for the next round.

Jasjit: Yeah, I think that's right and I think in our case we have some challenges with the traditional view just because the nature of our business. We often talk about the problem of small numbers because our students are in the tens and the twenties and the fifties and the hundreds, and we don't have users in the thousands and ten thousands. Sometimes it's actually pretty challenging to run really refined measurements around some of these experiments. It's difficult because we have to use a combination ... We have to use that approach, but we have to use a combination of our own gut as well and that obviously leads to some systematic errors. It's kind of a challenging situation when we don't have the types of numbers in this kind of business to be able to run all the types of experiments I was able to run, say, at previous companies.

J: Yeah, because you're not building a software product necessarily, so you can't get really in depth usage analytics or anything like that. You've got to look at it from a much longer time horizon, I would assume.

Jasjit: Exactly, yeah. And so because the time horizon's so much longer, we don't want to slow the volume of our experiments per week because our time horizon's lower, because we're not going to be able to move fast enough. So in terms of the measurement piece of it, we have to have some gut check on top of the data, which again can be problematic. So usually when something's working, within a couple weeks it's pretty clear that it's working but there's a lot of things that are kind of working. We don't know if they're exactly working well enough to justify additional investment and so those are challenging parts of running our particular type of business.

J: Yeah, sure. When you onboard a new student to the time that you've placed him in a position, what's the typical time frame for that?

Jasjit: Sure, so our web development program and our data science program are flexible programs so people don't quit their jobs to do them. We require people work twenty hours a week, so the recommended graduation time is about six months from beginning to end. We have students who complete it as fast as four months, but our goal is six months. 40% of our students actually have a job by graduation, according to our jobs report or audited stats. Then six months from that point in time, it's about 92/93% have a job. So within a year, 92/93% of people have a job in some sort of development role.

J: I see, so you've got two separate things going on there. You've got the six months that it takes to actually perform the education and then you actually keep those people engaged for additional six months on the job hunt side?

Jasjit: Exactly. One of the reasons we can have such strong job placement numbers is because we have career counselors who are actually pretty ... We have, I think, five or six students per career counselor. After a student graduates, they basically get matched with a career counselor who's meeting them like they would meet with their educational mentor, who's making introductions, who's working with recruiters to get feedback for them, who's helping them craft their story and iterate on their story. That's a big part of the program for the people who didn't get a job by the time that they graduated. But that process honestly starts after month one of our program. We're preparing people with technical interviews, we're trying to get a sense of what they're looking for, we have our own kind of view on how people should look for jobs and we have a thesis on that and we kind of educate people in terms of our process in our playbook. So that's a part of, not just the post-graduation experience but the after month one of being in the program.

I really feel like the way we think about Thinkful is as a career transition program as much as it is educational company. Because ultimately when people are making a decision to go to a school, a vocational school specifically, they're doing it to get a job. Everything revolves around how is it going to improve their lives in a financial perspective and from a work life perspective and from enjoyment of their work perspective. We really have to be super focused on that for people to be happy with their experience with Thinkful.

J: I think that makes sense. It's interesting to hear how the business model is really kind of centered around the student achieving their goal. And if you don't help them do that, then essentially there's no money in the business for you, right?

Jasjit: Exactly, and I think there's subtle ways as well where that system makes the incentives good for the students. For example, if I make money as a school when somebody comes and enrolls, I might have ... And say, you know, a lot of third parties are evaluating the quality of my education based on how many people finish, I have a lot of incentives to make the program not as rigorous, right? Because I might want to get as many people through that program as possible. We don't have that incentive because we're on the line. When people graduate, we have to work every day to get them a job in terms of improving our program to make sure that our standards are super high. When someone graduates from Thinkful, they are absolutely job ready. Obviously, that bar changes as competition increases, as technologies change. So every day, every week we are improving that and that's an incentive that we have only because we have this model.

J: Since Thinkful has been founded, I'm not sure how long ago that was, but since Thinkful was founded how has the business model evolved as a result of you running these experiments? What do you think has been maybe one of the biggest insights from running these experiments and how that's changed how you run the business?

Jasjit: I think in general the direction that Thinkful has taken has been from being less vocational to more vocational. Initially we offered one-on-one modules for people to upscale in particular areas without sort of the view on transitioning them into jobs. Then we built full programs for vocational education. Then we realized that access was a problem and that, to improve access, we needed to find ways of aligning incentives better. Then finally we've made a big push within a year towards being more on the ground with people and having more in-person aspects to our experience as a student. We've realized that, while we've started out as this company that was primarily online, for these types of experiences and investments, it's really valuable to have an in-person component. You need that kind of in-person support. You need people to be able to like see you and touch you and feel you and make sure you're real, and not just some ad on Facebook. So, that has been another iteration on the module that has been pretty transformative in terms of how we see ourselves and how we see our value to our students.

J: And did you just learn that from interviewing students or interviewing the mentors themselves or how did you come to that realization?

Jasjit: I think there's always some segment of students who asked us "Is there something in person? Is there some in person component?" That was a question that people had. That's kind of a clarifying question they might have. So there was always an indication, what's going on there? Obviously, there were other coding boot camps or other coding programs that were completely in person so we knew there was a module that was something that people wanted, and in many cases, more expensive than us. We tried some experiments and in doing those experiments, primarily with our own students, we got this extraordinary feedback.

When we went to our in person dinner sessions and events, we saw things happening organically that were really profound. People supported each other, helping each other, getting inspired by each other. People kind of engaging about their journey, the curriculum, where they've come from and their aspirations for where they want to go next. It was pretty extraordinary to see some of that happening. I think that also excited us. I think that talking to mentors or having students tell us that having a mentor that's in their location is really a big part of what excited them about Thinkful as well, also made a big difference. It just makes the experience feel more tangible, I think, to students and I think that helps a lot.

J: Yeah, and that's actually a pretty substantial change, I would think, because most of what you're doing is delivered online and so learning that the face-to-face piece is an important component of it had to change the way you ran the business in some way.

Jasjit: Definitely. I think it's still changing that way. I think it's still early based on that side, honestly. I think it's one of the most exciting things that we're working on and I think that we have this new position from which we're trying a bunch of experiments on this vector, which is sort of how can we kind of take out the in person experiences of this traditional education? How can you pull those out and add those to an online experience so people can get the best of both worlds? There's a fine line there because a big part of what makes Thinkful work is our cross structure, but at the same time there's a lot of business value in having support and in person network instead of being built. There's a lot of value there. There's value that's being created that we just haven't fully tapped yet, as well. So I think that's another part of the experiments that we want to do. How do we tap some of that value? The networks that are being built locally and the fact that we have twenty local mentors at some of these great companies, like how do we tap those networks to help our students? To what extent does having the support improve the number of students who complete their program? What ways can alumni who are local, who come to these events, network with students and help place them, help as well? So there's a lot of untapped value there. We're just sort of scratching the surface of right now, mind you.

J: Are you doing those experiments on student-by-student basis or by location or by cohort or like what's the structure around that?

Jasjit: Our experiments are being run by city. We're currently in a couple cities. Atlanta is sort of our flagship. We're running these experiments on sort of a city-by-city basis, and Atlanta has some unique opportunities to run particular experiments that we're excited about, such as going out to Alpharetta and places that have less access to physical vocational training places. We’ve gotten extraordinary response for some of those ideas. We're experimenting a lot with those types of things. We're experimenting a lot with just generally just moving around and meeting people where they are rather than having them come to us. We're trying different ways of developing the in person component of student experiences to just kind of improve the rate at which they complete and improve the rate at which maybe they refer other students. So there's a lot of those types of experiments that we're excited about here in Atlanta.

J: Sure. Which other cities are you in?

Jasjit: We're also in Washington D.C. and L.A., but we have students from across the country. Those are the cities where we're focused in on really developing out the in-person experience.

J: Right, so if you see something working here in Atlanta, you might take it to D.C., or if you see something in L.A., you might bring it here to Atlanta?

Jasjit: Yup, absolutely.

J: Have you seen anything that works in one city that doesn't work elsewhere?

Jasjit: I think one thing that we're finding is obviously cities vary in terms of cost of living and job market and, I mean, I think structures of cities are different. There hasn't been anything yet that clearly works in one area but doesn't work in another, but it might be too soon to tell. But there are massive differences, you know, in  Atlanta and L.A. in terms of cost or the types ... I think Atlanta has more of tech crowd and tech background than even D.C. or L.A. does. There's a lot of tech companies here. I think there are these differences in terms of the structures of the city, but D.C. has their own Alpharetta, and L.A. is obviously sprawling so there are also commonalities as well.

J: It's interesting. Have you or Thinkful thought about doing the same type of model in other verticals? Something that's not tech, for example.

Jasjit: I don't want to maybe talk too much about what's on the horizon from that perspective. Obviously we believe that we are a vocational education company and we're a vocational education module more than we are focused on any particular vertical. It's definitely something that we think about, but as of now, we just launched a really awesome data science program. It's kind of a new product, a new vertical, a new category. We're really excited about it and we're learning a lot from doing it and our students are pretty excited about it as well. We're pretty focused on that for now.

J: So it sounds like what you're doing is proving that this, I'll call it a hybrid in person, online vocational model will work by using tech and then you might end up, at some point, moving that into another vertical or another field of study.

Jasjit: Yup.

J: That's interesting. So it's kind of the same process that people would use to approach taking any product into an additional market is you make sure it works in one spot and then once it works there, you can easily copy and paste, so to speak, into other verticals or other markets.

Jasjit: Yeah, I think that's right.

J: Cool. So when it comes to the structure of the education itself, how are you testing that what you're doing is effective and that you're teaching the right things? I'm assuming you're talking to the companies that you're placing people. How do you make sure you're testing and getting people the education that they need?

Jasjit: I would go back again to the value of having mentors who have an average of ten years of experience in their field. What's nice about that is that our mentors, when they're not mentoring our students, are hiring mangers or on interview loops. They have a really good intuition for what the bar is for students. When students go through our program, one of the aspects is ... To graduate, they have to build four capstone projects, which become their portfolio. A mentor reviews each capstone, and each capstone has to be approved to graduate. The mentors look at those projects like they would look at anyone applying for their company, and they are the ones who are keeping the bar high because they know what the bar is and they understand that the quality of the projects have to be extremely high. That's kind of how we think about the standards and we make sure that not just the mentors are signing off on it but also their program managers.

That's basically what a student has to have to graduate at the end. That's like the deliverable is the four capstone projects.

J: So are they getting placed at the companies where their mentors work? Does that happen or are there any guidelines around that?

Jasjit: Yeah, it definitely happens. I think if you are a company it's actually one really great way to evaluate and kind of snatch up talent before anyone else sees it, right? If you have mentors who are just really impressed with their student over the six months they've worked with them, they know how they think about coding better than anybody else does. I mean, I have mentors come to me talking about wanting to kind of get their student in front of their manager or their company because they're just really impressed with them. There's also not just the sort of quality of their code, but also their culture fit, like their mentors are in a great position to know if a student is going to be a culture fit.

So yeah, absolutely. That's one of the reasons we have a lot of support from employers. It's a way for them to build a pipeline to some really great talent and to kind of have a right of first refusal, almost, for some of this talent.

J: Yeah and that's another component to the business model, too and part of, I guess, why it works is you've got buy in and incentives from any different directions.

Jasjit: Absolutely.

J: Yeah so that's really smart. It's a smart way of structuring things. So you've got people incentivized ... I guess, all three parties are incentivized to participate in different ways and it sounds like you're not just solving a problem for the students, but you're also solving a problem for the mentors at the same time. And then the business gets to capture some value for doing that along the way.

Jasjit: Absolutely.

J: That's really cool. I like it.

So in terms of the product itself, delivery of the platform for doing the training and doing the education, can you speak to maybe some of the processes around that? About how that product is built and what those iterations look like?

Jasjit: Sure. Our product, in terms of the non-mentor aspects of the product, is the platform itself. I should talk about the experience of the student. From the students' experience, they're doing basically three sets of things to learn whatever it is they're learning, the data science or lab development. The first thing they're doing is learning the concepts. They're going through our curriculum online or on their dashboard, and they're learning some concepts. Then they're practicing that concept with either drills or challenges or small projects- it depends on what that concept is- and then they're taking that thing at the end of a unit, they're building some capstone projects. Then they get to design and scope it out with their mentor, get approved by their mentor, and then go build it and their mentor assists them. In doing that, they'll meet with their mentor three times a week for an hour and they'll review code. Their mentor will pair a program with them and will help kind of guide them through this process and make sure that they are putting together production level code that would cut it at their companies.

That's kind of how we think about the process. So you can imagine that on all of those dimensions, on both the curriculum, the drills, and the challenges, and the small projects and the capstones, we're constantly pushing updates and changes, incorporating feedback, and improving on a weekly basis. I think being a tech company allows us to do that kind of thing. Whereas other types of education companies, that would be more challenging.

J: Using the digital platform enables you to change the way that students are actually going through the process a bit faster. What kind of experiments have you done there and how has that platform changed overtime?

Jasjit: That platform ... I mean, we have entire sections of our curriculum that are rewritten completely from scratch. We, for example, used to teach AngularJS and saw more demand in our markets for ReactJS, so we just wrote a whole new curriculum, and we've had three complete rewrites of that curriculum in the last I think six months.

If you see students struggling, if we get feedback from mentors who'd write notes saying, "Hey, students are struggling with this section or this area", we'll go back to that section and improve it. Mentors will give us feedback because we also have mentor training documents to make sure that they understand, for a particular section, how to go through it. There are these really cool feedback loops with students, with mentors, to help make sure that we're continuing to improve the curriculum. We do it in big ways with complete rewrites and complete additions based on new technologies that we see or having better job prospects. We do it in micro ways on a daily and weekly basis as we find issues or we think adding a sentence here or there could help explain something better. We do it all kind of in a constant way.

J: So it seems like you are able to iterate pretty quickly. If you see something that's not working, you can fix that and get it out and published into the real world pretty quickly.

Jasjit: Absolutely.

J: Yeah, that's awesome. What parting thoughts do you have or what kind of advice would you have for other people who are thinking about how to structure their business models in a way that can incentivize people or capture a value from multiple sides?

Jasjit: In general, I think it really helps to ... When you're defining the problem, there's a lot of value in questioning the most basic assumptions that we just take for granted, and we shouldn't. I think what excites me about what Thinkful has done is that there's this very basic assumption with just identifying the problem that these incentives are misaligned. That you just go to school, you pay, and then you see if you get a job afterwards. That's just an assumption that people make and they don't even think about it for a second. It's a pretty massive problem that ought to be solved.

I think that it takes work to try to observe the world as the complete idiot and just ask why about these things that just exist. I'm not the guy who did that in this case, but I'm really proud of my company for having that approach to things. I think that one of the things that I've learned as part of this process is that it takes work to go through world and your life and try to question some of the most basic things and try to find problems that way that, if solved, could have a pretty massive impact. I think in the case of Thinkful that's what we're doing and I think that it's an approach that hopefully will serve us well.

J: So it sounds like if you look at some of the things that maybe we take for granted and think about why that exists and how it might be better, there's a good chance you'll find a business opportunity there.

Jasjit: Yeah, I hope that's true. I think that's maybe more idealistic, but I think that not a lot of people can do that. It takes work. It's not easy and I think that you have to constantly kind of force yourself to be unhappy with the way things are to find some of those things sometimes. So I think it's human nature to just adapt to whatever changing conditions exist.

J: You're right. That's a pretty powerful thing. Instead of accepting everything as status quo, maybe take a look around and see where you might be able to change something in a way that can have a positive impact for everyone involved.

Jasjit: Yup.

J: Yeah, I love it. Great advice. So, Jas, if people want to get in touch with you or want to learn more about Thinkful, what should they do? Who should they reach out to? How do they get you on the line?

Jasjit: They can email me at jas@thinkful.com.

J: Jas, I think it would be cool to have you back on the show at some point and talk about the way Thinkful's evolved and what you've learned along the way. I look forward to doing that. Thanks again for being on the show today. I appreciate you taking the time and we look forward to next time.

Jasjit: Absolutely, thanks so much.

Outro: That's it for today. Thanks for listening to Design Driven. We're glad you enjoy the show. Have comments, questions, or an idea that you'd like us to cover? Point your browser to designdriven.biz and click "Contact Us" on the top of your screen. We'd love to hear from you. Tell your friends and colleagues about the Design Driven pod. Post on Facebook, Twitter, LinkedIn, or send them an email and tell them to go to designdriven.biz or wherever they find their podcasts. Until next time, remember what Thomas Watson, founder of IBM said: "Good design is good business."