The below is a full (unedited), machine-generated transcript of a Youtube session / podcasting episode I recorded with Brian Leonard, the co-founder and CTO of TaskRabbit in Q1 2020. You can view the video/listen to the podcast on Youtube, Apple Podcast, Stitcher or wherever you get your podcasts.
Erasmus Elsner 0:00
Hi guys, Happy New Year. Welcome to this year’s first episode of Sand Hill road show where I talk to successful startup founders and their investors about the companies that they built an invest in. And the goal like always, is to give you a sense of what it’s like to be in their shoes, or their business take to learn from their many successes and mistakes along the way to get an understanding of how they got where they are today. And today, I’m super excited to be joined by an absolute pioneer of what we now know as the gig economy. If we think about the early companies of the modern gig economy, there are really three companies that come to mind Airbnb, Uber, and TaskRabbit. And that’s why I’m so excited to be joined by the technical co-founder of TaskRabbit, Brian Leonard today.
Brian Leonard 0:40
Thanks for having me.
Erasmus Elsner 0:41
together with Leah Busque he built TaskRabbit from a really small niche marketplace, into a huge platform used by millions of people. TaskRabbit is an online marketplace where people can go and outsource small jobs and tasks to others in their neighborhood. So if you need dry cleaning, picked up your groceries delivered, you can post the job on TaskRabbit name, the price you’re willing to pay, and we have hundreds of background checked neighborhood TaskRabbit is ready to go. The company was acquired in 2017 by IKEA. We will talk about the very early days of TaskRabbit.
I’m super excited to have with me here today. Brian Leonardo, the technical co founder of TaskRabbit. Welcome, Brian.
Brian Leonard 2:04
Yeah, thanks for having me.
Erasmus Elsner 2:06
Before we jump into TaskRabbit, and what it was like to build and scale this company, I want to learn a little bit more about the early pre founder, Brian. So you graduated with a degree in computer science in 2003. And then worked for five years at IBM where you met your co founder, Leah, at any point in time, if I had met you, let’s say in 2003, or 2007? Would I have been able to predict this great founder journey of yours? Or at what point in time that you catch the entrepreneurial bug?
Brian Leonard 2:37
I think it would have been very hard to predict at any point, what what we would have done, but from very early age, I always figured that I would sometime before 30 certainly before 40 be an entrepreneur of some sort. You know, I had various versions of babysitting and lawnmowing and music selling and all kinds of things in my childhood started culminating really when 2008 with the iPhone and the App Store and stuff like that. So even pre TaskRabbit, I started making some apps and things like that. And when we had this idea, it seemed it seemed like a good one. Yeah.
Erasmus Elsner 3:15
So take us back to this really early days. With Leah and I, I remember her talking about it, you sat around the kitchen table in Boston. And she told you about these 100 random strangers who she she wanted to complete all kinds of tasks. So this initial supply side of this marketplace. What was that? Like? Were you instantly convinced that you should go and build this app? Or did you call her the next day and say, well, we have a cozy corporate career here.
Brian Leonard 3:44
Yeah, we worked together at IBM, IBM is such that everyone under a certain age knows each other. It’s not that many. I remember how she quit. She had this idea. And she quit. And I didn’t know what it was, I took her out to out to lunch. And he told me the idea. And you know, still during that lunch, I was sort of in this iPhone phase that I was talking about. And I said, like, hey, let me build the app, because you’re going to need an app, that’s the future. Turns out more or less true, which is good. I didn’t actually build the app, we did all kinds of other things. And then eventually, I built the app, the kitchen table thing I remember, you know, we talked about how the business model would work. Somehow you buy points or credits or something like that. And that’s how it started originally. The clearest memory I have of that is the fact that I stopped by her house. After picking up my wife and daughter from the airport and my daughter was still in one of those like bucket seats, like like, you know, child three month old sort of thing. And, you know, we she was like sitting on the table to while we were while we were discussing this credit system. And so, you know, she’s 11 years old now and in middle school, and so, you know, they, the company and her grew up together. I remember
Erasmus Elsner 5:00
So let’s talk a little bit about scaling it up. And I’ve personally operated a very niche, small marketplace. And I know from that experience, how hard it is to get the v1 out to get it operational and front of people then to get the first supply and demand side inventory in. And then once you have some traction, it becomes really intense because you get so many requests from all kinds of people, so many different profiles. And I found this, this quote of yours from from a crunchbase News article where you said, thinking about starting a business, again, you do a business that requires less people coordinate. Turn up, talk to us a little bit about the initial technical and operational challenges.
Brian Leonard 5:44
Yes, it’s definitely true. I think mostly about this talk I used to give to engineers and product people, and it’s about you, you’re building the code. And you know, the code mostly works, you’ve got this test suite, it’s green, but you can’t in tests when there’s some erratic factor. Because you want to make sure it’s always green, what you do is you step out that erratic factor. And I’m like, you know, if, like, you pretend it’s exactly this time, so that daylight savings time doesn’t screw you up twice a year, or whatever it is, in marketplaces, the ironic factor was like the humans involved, and whether they show up on time, and whether they, you know, are supposed to clean the house every every week, but then they have vacations or, you know, they want you to move it up to Thursday. And the whole talk was that you you can’t step out real life. So the code works. It’s just the people that are the problem. And as a technical founder, and a product leader, you know, it’s more and more learning that it’s not as much about the product working or modeling that real life, it’s features you you build, and then the flexibility you build in that it can ebb and flow as real life happens. And as an example, we tried to build that scheduling, weekly, bi weekly, whatever it is cleaning. And like we end up, you know, rebuilding Google Calendar, and repeating meetings and vacation schedules, and like it’s just too complicated. And in the end, we ripped all that out and just said, like these two people are in a relationship, it’s going pretty well, like, here’s a button where the the Tasker can build the client. So chat, here’s a great chat system, show up, hit the button, like and a story. And in that case, that was the right level to interface with the somewhat exhausting reality that is on the ground up there.
Erasmus Elsner 7:37
Let’s talk a little bit about marketplace broadness versus specificity the mothership of all marketplaces at that time was really Craigslist. And there’s this great graph of Craigslist, all the broad categories that it had, and then all the startups picking away category after category, building, right vertical marketplaces with specific feature sets. Interestingly enough, if you look at TaskRabbit, in that graph, it’s associated with general labor category in Craigslist, I read this interview with one of the x marketing people at TaskRabbit. Jamie, via Gino and she said, while you were a very broad platform, from the very early days on you were deliberate about building a liquid marketplace for the most popular task categories, which were handyman tasks including IKEA assembly, house cleaning, and moving help. So when I think about these popular test categories today for handyman work, there’s thumbtack, then there was home joy for the cleaning. There’s Dolly now for the moving then there are adjacent categories like dog walking, where we have the SoftBank funded wag. My question is there are so many categories, which TaskRabbit could have evolved into, are you sometimes wondering what TaskRabbit could have become?
Brian Leonard 8:50
Yeah, that’s very interesting. We are in the same class and even neighbors with with the founders of Lyft, for example, and like, you know, I chatted with Logan early on and about exactly how this thing was was gonna work. And there was some driving happening on ours, but you know, there was licensing and all that other stuff. And like, yeah, let’s not do that and like, and he did that and like now they’re their public company announces worked out, right. And you know, we didn’t super focus on on delivery, it was like, you know, our fifth ish category, let’s call it doordash or Postmates, or something like that is, is all into that from from restaurants. It’s also a super hard business and debatable about the, you know, profitability margins. And it’s just one of those things we struggled with. But the thing that the resonates most about what Jamie said, is actually the first part like, we were always we had this hypothesis and dedication to the broadest market possible even now, the value still comes from the fact that you’ve got something in your life that just bugging you, it’s just eaten away at the moment. Monkey on your back. So loop in your brain real value comes from an instinct that you see something is just shelf sitting on the ground. And like before you even are done processing all the logistics needed to make this not a problem in your life, like you’ve already like kind of like, tapped your pocket. And you can put that in TaskRabbit it’s on its way to completion. And that has to be broad to achieve that kind of impact there early on days was was kind of wild west in the way that the internet was beautiful. In those days, and Craigslist was beautiful, which is like the weirdest things possible, you can put on TaskRabbit. And like they would get done. It’s like a Crossing the Chasm kind of thing like the early adopters are using it and figuring out all the different ways you can use it. But to get into that early majority, you have we had to we found that we had to ecommerce certify it a little bit. And that means like, you go to a homepage, and it just doesn’t there’s not just an empty text box that says Well, what do you need? It’s like, Well, you know, and at some point, you actually have to suggest what to do to get into that into that early majority. And so, you know, we took what was most popular and leaned into it. But I still see it as you know, pretty broad even household chores is like you add up like hours spent and all that if someone did all the household chores in Europe, and in the US, that’s like trillions of dollars of work. And so us and IKEA not just furniture assembly, but that whole life at home is something that is really great. So the last thing is they pop into my mind is that it’s also really great for the taskers. Some of the most inspiring stories that I have are about people that started out and old lady that started cleaning houses, and then realize you can make a lot more money doing any person work or electrical work and started watching YouTube videos. And you know, she has significantly changed her life because of, you know, essentially being able to change an hourly rate from 20 to 60ish, or 100 this year, or whatever, whatever her rate is these days
Erasmus Elsner 12:08
A related question to this, I saw some of the most grotesque tasks that were put on, including removing a squirrel stuck in an attic and fitting it to a rubber chicken, former high school student at a high school reunion, I mean, the variety must have been huge in the beginning, right? Of what people yes, there.
Brian Leonard 12:29
It’s fun, right? And like, this is what portos like to ask about, like What’s the weirdest, weirdest thing not right? So we actually have more grotesque one sad I would categories is under weird, are odd. People love to ask those. You know, it’s it’s this love hate relationship, because we actually want to get to that general population, most of which do not have rubber chickens that need tuxedos. But all of which have a shelf or a TV sitting on the ground. Those are the stories we want to tell, but they’re certainly less quirky.
Erasmus Elsner 13:01
So let’s talk a little bit about the technical stack. And so you mentioned on Quora, once that in 2008, you decided to go with Ruby on Rails, because it was basically a fresh wave compared to the C and Java that you’re using an IBM with Leah Ruby at this time enabled everything to be done much faster with respect to the current task rabbit stack, you revealed. That was basically AWS, Ruby, react, react, native MySQL, Redis, Elastic Search, and there’s some NodeJS and Python there, as well, as we see a lot of software now moving to open source software frameworks. A lot of the incumbent marketplaces open sourcing some of their frameworks and libraries as well. I remember when I built my, my niche marketplace, we use the Airbnb calendar in our stack. If you had to build a generalizable marketplace framework today, what would be the stack you would land on?
Brian Leonard 13:57
Interesting question, it’s all CTOs job to sort of always understand that. And then make the decision sort of, I don’t know, quarterly, maybe yearly, whether we should actually move to that, or something like that, to stay fresh. And so you know, I’ve thought about that before, lots of companies start out in one language slash framework, and then over time, for a variety of reasons, migrate to others. And so you know, at least one large migration we did, which I would still recommend, then React Native and react in general and then React Native. Specifically, you know, we had a fairly small team pre acquisition 12 engineers, you know, supporting a website, back end. And for apps, you know, two platforms, iOS and Android times, two user bases, Tasker and client, the leverage added by React Native only having to build those mobile apps one time and just in general, the velocity you know, we estimated that five times more productive than than we were before So I continue to recommend that. And we were one of the earliest ones on that. And I don’t make these jumps. lately. That’s like, I jumped to Ruby. And then like, no major jumps very, like, we didn’t jump on to Mongo, for example, we stuck with my sequel, like, you know, we were tried and true technology. And that was the biggest jump, the biggest question is probably would I go with rails and some of the things in that ecosystem again, or I think the choice would be with node j. s, and something from that. And I think it’s, it’s basically about who was involved, I still can make something faster than anything in in in Rails, and there’s a whole set of people that are still in that spot, and still are super ended. And you know, that that ecosystem is, is leveled off, if you look at the graph difference, it’s also about the throughput, like TaskRabbit doesn’t have a major throughput problem. You know, everybody that’s hitting our website is paying us kind of a lot of money. When you put that up against something like I don’t know, Twitter TaskRabbit continues to make this trade off. That’s it’s about productivity and sort of pivot ability, and like adaptability and things like that. I think I think node is a strong contender. They’re newer projects that I’ve made, particularly that have a lot of asynchronous work chosen node. And so I consider that. As for people out there making, one of the things that I just thought about notice is sort of about is the benefit to full stack to you know, JavaScript on the front end the back as your hire. I think there’s a hiring advantage in the J airspace as well. As you mentioned on Quora, once that you had some problem with Elastic Search, we really leaned into Elasticsearch the problem we had taskers needed to look at and filter quickly all the tasks available, just that but also geographically, sort of like in this small window, sort of moving around in a map and then filtering by category or price or all these different sort of things like that. And so we we do normalize that all in Elasticsearch and drove it from there. And so and we use that for still for recommending, so now, if you go into site, and you said I want somebody at this address, we generally serve that from Elasticsearch cache, a lot of use cases that we did an Elastic Search. I’m not sure about the geographical one, maybe there’s post GIS would be a good option these days. Great thing about it was that is sort of infinitely expanded and clustered and all these things.
Erasmus Elsner 17:25
So let’s talk a little bit about Brian Leonard after the IKEA acquisition. And I just learned that you’re now in the process of starting your new company. But let’s talk a little bit about this timeframe after the acquisition. I mean, you were the technologist at IKEA and where everybody knew what a billy is, but very few people have heard about refactoring microservices and DevOps. How was it to transition from being really on the frontier of Silicon Valley Tech to moving into into this Swedish, IKEA furnished? new home? How was this transition for you? Over the last few years?
Brian Leonard 18:02
Yeah, I think I’ve read let’s call it 100 blog posts on various acquisitions and things over the years and 80 plus percent of them like they basically don’t work out. so far. So good with TaskRabbit and IKEA, and I attribute that to a couple things. One, IKEA and TaskRabbit, a very aligned values, they’re basically the same once you throw in sort of translational, Swedish English, sort of things. TaskRabbit says, We’re making better everyday lives for everyday people on both sides of marketplaces, end to end. IKEA says a better life for the many people just very Swedish and beautiful in their own way. And those highly aligned and better life at home is another thing that I TSS, and they, they really do mean, just better life at home not selling more furniture, I can’t even believe so TaskRabbit, you know, does plenty of IKEA assemblies. And we’ve optimized that a little bit as part of a partnership, we did pre acquisition. And, you know, but then also we have a marketplace that does all kinds of other stuff. And so IKEA says, like, you know, we went we went on TaskRabbit everywhere IKEA is which is kind of a lot of places, you know, 30 ish countries. And I said, Okay, well, the fastest way to do that is to only launch the IKEA piece. We got new board members as part of the acquisition and one of the Swedish guys on that says Like, but is that truly making a better life at home for the many people don’t they have cleaning and other things to do? I’m like, I couldn’t even imagine in that moment, this world where this guy is talking me down from my own mission that I’ve been working on for 10 years. Like he’s, he’s like, you know, you really should stick to what you want to do. And I’m like, this is this is amazing. He’s pushing me to achieve something that is not directly and it is like furniture. Revenue thing, but a general sense of creating easier lives for people. And, you know, I attribute most of the success to that. Now it is going through their own just to get closer to your question. And he is going through their own transformation from a door based. One of the things they say is that in the potato fields, like we buy land in the potato fields, and we put this big blue box, and then people come, it’s like an event, we serve lunch and meatballs, and like they buy stuff. And then they go home, you know, they’re going through several different transformations right now one digital, more and more being sold in the app and online, in general, and in their company. Another one, they’re doing lots of interesting experiments and the city center, they’ve got stores in the center of Madrid and Paris and London. The third one is just sort of the corporate culture itself around these things that is necessary like insourcing, more things more development. They’re hiring 1000s, and 1000s, of engineers had the opportunity to preach the Silicon Valley gospel a few times around agile methodologies, and API’s and iterations and things like that.
Erasmus Elsner 21:11
Yeah, I mean, Sweden is actually in Europe, it’s the country with the most unicorns to move on to the last point, I wanted to mention type projects. And I always say you can tell a good entrepreneur by the number of domains registered on GoDaddy, and every entrepreneur has a side project, and then a side project hasn’t been another side project to it until he’s stuck in court with with with the market somewhere. And then I looked at your LinkedIn profile, and I saw that you had two side projects there. There’s wonderlands development, and then there’s the cakewalk app, think about side projects, and why it is important to sort of keep your creative juices going as an entrepreneur.
Brian Leonard 21:55
Yeah, there’s, there’s another class, the most common cases never launched, its completeness thing, if you’re doing something for fun. One of the challenges is, if it’s not fun anymore, do you keep like that last 20% is a you know, it’s like the 8020, it takes just as much time as the first 80%, if not more, you know, and the other is sort of just like open source, sort of things I’ve released over the over time that I’ve really enjoyed doing side projects, all of these, particularly sort of the, the two that you mentioned, that class, which is sort of like not exactly a startup, but certainly like a self contained app for useful thing for me was always about scratching learning, which is a very strong case to be made, that I was able to be a better CTO because then specifically I said, it’s, it’s, it’s my job as a technology leader to sort of understand if there are shifts in the world of technology we should be participating in, how am I going to best understand if React Native is is right, for for us. And somewhere in there, you have to actually try it and make an app, there’s another one, it’s still ongoing, you know, feel free to sign up world called buzz back, use buzz back calm. And it’s for call boxes. So you can, you know, these people were still buildings built in 2019 2020, are still getting these like phone based call boxes, where you have to hit nine to let somebody in. And so it’s sort of amplifies all of that, and you create codes and whatnot. The point is, I use that as a way to explore whether React Native was was good or not. And, you know, that then translated into me being quickly able to, to implement that in TaskRabbit. And so, yes, one of those was about catching up in iOS development, cakewalk app was about that, you know, I made the app in 2010. And then, you know, we hired people, and they continued making apps. And I felt like I was behind. So I picked a project that would allow me to catch up. And Wonderland development, which was, was the transit apps that I made even before TaskRabbit that just kept running. And, you know, over time, the money went to eventually didn’t even pay for the the $99 that Apple charges so and, you know, the old API’s and things like that, so I shut it down. But in general, I think it’s about staying fresh and, and learning to if you’ve got that sparked
Erasmus Elsner 24:30
Okay, so thank you so much, Brian, for being with us. And maybe as a last point, where can people find out more about Brian and follow what you’re up to next?
Brian Leonard 24:40
I’d say I’m not particularly active. But Twitter, bleonard. We’ll put that in the notes or however that works. My my new company is called a group a room. We’re making marketing tools one of these days I’ll get a a think Where you can put your email address and we’ll, we’ll get on on on that. The the idea there actually, like you mentioned before, like my next thing is going to be a little bit more more focused, maybe b2b, we’ve we’ve realized that marketers sort of have a have a tough, tough go in life, particularly, they want to run a bunch of interesting campaigns and they’ve got these goals they need to hit, but they don’t have the data that’s gonna allow them to be effective at that. We’ve seen this in TaskRabbit. We’ve seen this and all the other companies that my co founders have been in and all the people we’ve talked to, and so we’re gonna we’re gonna take a run at it, solving that problem.
Erasmus Elsner 25:43
Do you you have a landing page already?
Brian Leonard 25:44
Yeah, we’ve gotten a little kangaroo on there. It’s called grouparoo. And, yeah,you can check that out.
Erasmus Elsner 25:52
So thank you so much, Brian, for being with us today and I’ll make sure to follow your journey.