UNSCALABLE

CHAPTER #1

James Richards

Founder & CEO, Teleborder

Guo: I know Teleborder was a pivot, but I’d love to go back to the beginning. How did you get started?

Richards: Sure. So I’m a lawyer. Michael, my co-founder at the time was not a lawyer, but he hated lawyers because he’d had to deal with them a few times. And the worst part about dealing with lawyers was always feeling like you’re getting screwed on bills. Lawyers are very, very expensive. They‘re also very opaque when it comes to pricing. It’s rare that you can look up on a website how much a particular service costs. Although intuitively it seemed like you should be able to ask, “How much will it cost me to get my company set up”, or “How much will it cost me to get this transaction done?”, but you couldn’t. It just seemed like you ought to be able to tell somebody how much a legal service ought to cost, at least within some sort of range. But no one did.

So we thought, what if we could package law into these little fixed-price nuggets that you could buy just as easily as buying a book off of Amazon and get all the things that you get when you buy a book off Amazon, like price, clarity, fungibility, certainty, returns, all of that stuff. If you could basically e-commercify law, it would be a lot better, right? Because I wouldn’t have to pick up the phone and call a lawyer and talk for an hour and interview ten of them and get screwed on bills. I’d just click a button and I’d magically get a lawyer and it’d be that price. So that was the idea.

Guo: How did you get your first customers?

Richards: Well, so that was the problem. We actually never had customers, which was why we needed to pivot. It was a two-sided marketplace, and we started by thinking “Well, if lawyers are our inventory, the first thing we should do is go get some lawyers.” And so we went out and we got, in fairness, we did a pretty good job of getting lawyers. I don’t know the exact number, but we had on the order of a couple thousand lawyers on our wait list, and we had on-boarded a couple hundred.

By the time YC interviews rolled around, we had lawyers who we interviewed, who we screened, who basically signed up to offer services on our site and more than just signing up and giving us an email address, actually picked the services that they wanted to offer. So we would have, I don’t know, let’s say incorporation, I can’t remember what our actual services were, but we’d have an incorporation service, and we would set the price, say $500, and we’d set what was included in that package, and then lawyers would log in and they had this dashboard where they could opt in to offering those services. And so we on-boarded on the order of several hundred lawyers and had several thousand waiting to get on. Which I think is what made YC go, “Oh, these seem like smart, determined guys, and lawyers don’t think it’s crazy!”

In terms of how we got them, Craigslist was surprisingly good. There’s an alarmingly high number of lawyers trawling Craigslist. It’s Craigslist, so you get what you get, but there are some good ones there. And an initial seed is good enough to have them tell their friends. Because lawyers, like any service provider, are selling time. And time is the most perishable of inventory because it literally disappears every second. So when lawyers hang out, they all talk about how to get more clients. So if you can get a bunch of lawyers excited about something, then they’ll all just sign up in droves.

But that’s not the hard part. We should have known it, but we were really arrogant. We thought, “Oh, look at all these lawyers signing up, clearly we’re so smart,” when actually they were signing up because that’s the easy part. The hard part is actually getting clients. And they were signing up because they thought we would solve that problem for them. Whereas we thought that by having them, that would solve our problem of getting clients. We were both looking at each other to solve each other’s problem.

It’s so ironic, and in hindsight, so stupid. We launched our customer facing side the morning of our YC interview. Mainly because we didn’t want to walk in and say, “We have no sales, and we haven’t launched our customer facing side yet.” At least we had one of those two things. It was like, “We have no sales, but you can’t blame us because we launched 30 minutes ago.” And we spent all of May, June and July, trying everything to get customers. By July we were grasping at straws.

Guo: What were some of the things that you tried?

Richards: Actually, we didn’t try everything. We tried virtually nothing. We were still in the “we are geniuses” mode, trying to make things super-scalable from day one. We launched every vertical of law I could think of, between 10 and 20 verticals of law. And we had lawyers in the top 50 or so metropolitan areas in the United States. We thought, “We’re going to do this big bang launch,” and “If it’s e-commerce, the way to do e-commerce is simple, you just get people coming to your website, then a certain percentage of them will convert, and that’s your company. If we get a hundred visitors, five of them will buy, right? It’s just math.”

To get visitors, we did a bunch of SEO things, like we had custom landing pages for every type of law, in every major metropolitan area, that we generated programatically. We had a newsletter that no one ever signed up to. The thinking was something like “People will sign up for this newsletter, then we’ll email them stuff and then a couple years later they’ll buy a legal service from us.” In hindsight, it was really bad.

We tried all the things that in theory were scalable. But mainly SEO and trying to get large numbers of people to visit our site. And you know, we were successful in getting people to visit our site; our Google analytics numbers went up, and at first we thought that was awesome. But those visits never converted into sales. Each week our excuse for not having any sales was, “We just need more visitors. Last week we only had a thousand, now we have three thousand. Now we have eight thousand.” And when it got to ten thousand visitors per week, and there was still nobody buying, we said, “Alright, it’s a large enough sample size, now we have a problem. We have a real problem.”

Looking back, we should have picked one area of law rather than 20, and picked one, not even necessarily urban, just one area of the world, geographically speaking, as opposed to the top 50 metropolitan areas in the United States. And owned that, whatever that was. But instead we did the precise opposite, which was to try to go as big as possible, and it was a bit like putting a drop of ink in the ocean, it got diluted to the point of oblivion, it just didn’t matter. All our efforts just didn’t matter.

Guo: So at this point, you decided to pivot. What did that process look like for you guys?

Richards: We’d been thinking about it for a few days; obviously when the numbers are that bad and Demo Day is approaching, you start to really worry. We still thought we were geniuses and that there was some sort of trick to getting customers even though YC had consistently told us there wasn’t. We just didn’t listen.

So we finally decided to go talk to PG about this. And I kind of expected PG to pull a Yoda and tell us, “OK, yeah here’s the secret, go do this. Change the color of this button and the users will come.” Of course, instead he told us, “You guys need to go do precisely the opposite of what you‘ve been doing. You should 1) pick one vertical, and make it something you personally care about, then 2) go really, really deep in that vertical and don’t worry if it pulls you away from this legal marketplace idea. It doesn’t matter if it doesn’t seem scalable at first. Just go really deep in that vertical and see where you come out.”

So we floated two verticals by him on the spot and started brainstorming. Our first one was DUIs, since we figured there are a lot of young kids who get pulled over and don’t want their parents to know. And it was a great fit because they could use our service anonymously and online, without telling their parents and without it costing much money. That may have been good in theory, but none of us had ever gotten a DUI before, and I still don’t even have a driver’s license, so there wasn’t a personal connection with the problem.

Our next idea was immigration, and we certainly had a personal connection to that problem. I’m not a US citizen, and I’ve spent 22 of my now 26 years on a visa. Michael, my then co-founder is Belgian, and was also on a visa at the time. We both found it incredibly difficult to stay on top of our own immigration paperwork. It’s ironic, right? I’m a lawyer running this legal startup on a visa and I find it hard to manage my own visa. That’s weird. Maybe other people share this problem. And PG said, “Yeah, that seems like a good idea, because I was just talking to [a founder of a large technology company], and he was saying that even they spend tons of time project managing all this immigration nonsense. Do you want me to introduce you right now?” At that point, Michael and I asked ourselves, “Alright, are we going to focus on immigration, or are we going to stay the course with the legal marketplace? And if we focus on immigration, what does that mean, and what do we do?”

In the end, we decided that if we were going to kill the legal marketplace to focus on immigration, we should be sure. We couldn’t get any of the random visitors to our site to talk to us, so as a test we paid people off of Craigslist $20 each to come to our office, sit with us and go through our site. We ended up hiring 15 people this way off of Craigslist. We told them, “Okay, let’s pretend you’re looking for a lawyer, and this is the website that somehow your friend has told you to go to. Tell us what you think.” And, oh my God, the feedback was horrible. They basically all said, “Why the hell would I ever use this, I would probably just go to Yelp or ask a friend.” It was just so bad and exactly what we should have done at the start. But after that very demoralizing day, we knew that if the legal marketplace idea wasn’t dead before, it certainly was then. And we literally started from scratch the next day, throwing out the code base and starting over from scratch, knowing only that we were going to “fix immigration”. Which was fun.

Guo: With your newfound knowledge, how did building and growing the new company compare to the old one?

Richards: We basically did exactly the opposite of what we did with the legal marketplace. I don’t think we even launched a public site until the day before Demo Day, because our first goal was to get a paying customer, paying us anything. Even a dollar, we would have taken it. Just any non-zero amount. So we emailed other founders we know asking if any of them wanted help with immigration stuff. And we got a lot of responses. Some people wanted to pay us for a project management tool, like a BaseCamp for immigration. Other people asked us if we could just do visas for them. And doing someone’s visa, that costs at least $3,000. I was like, “Oh wow, three grand up front, I’ll happily take that over $40 a month for a SaaS product.” So I literally did someone’s visa, by hand. It’s still unclear to me why he let me do it, but that was first our paying customer.

Soon afterwards it was demo day and I was fundraising, so we had to figure out where we wanted to go with the business. Clearly we had a business, because we made more money in one day than our legal marketplace made in one year, but we didn’t know how to go from three thousand to thirty thousand, and thirty thousand to three hundred thousand, and so on. The logical next step, which we did, was to just keep doing more and more visas for people. Even though our customers were small startups, a lot of startups have this problem. So we just did all their visas. And the fundraising story was that given enough time, we would either figure out a way to automate the process, so we’d be robots doing visas, or we’d build up the SaaS tool into something sustainable that would, I don’t know, maybe let us stop doing visas altogether.

Eventually, we came up with the idea of building a network. We thought, “Instead of us doing all the visas ourselves, let’s get a network of lawyers to do the visas.” And so that’s how we launched our network of lawyers. Our first lawyer was a great immigration lawyer who had reached out to me over LinkedIn cold. I said, “Hey, you’re an immigration lawyer, right? Would you like to do this visa for us? We’ll pay you.” And he agreed. So instead of us doing visa work for people, we started shipping it out to lawyers and charging a markup over their rates for our software and project management. And as it turns out, that’s a pretty scalable business model. And it’s just so ironic, because it’s kind of like the legal marketplace idea come full circle, only this time with paying customers.

Guo: Did you go out of your way to communicate with these users, or delight them?

Richards: We once picked up two employees who we’d brought in on a visa for the same customer. They came on the same flight from France and we picked them up in a car, had hot towels and drinks ready for them, and then we took them straight to the Five Guys at SFO. And we also got them beard warmers because we had heard they were really into SF hipster culture.

In general though, it felt much more like a consulting job in the early days. So if a customer said they had a problem in any way related to hiring or managing expatriate employees, we would just offer a way to fix it, even if it had nothing to do with immigration. We still do this today. One customer, after getting a visa, also needed a nanny, since finding a nanny in Palo Alto is really hard and he was moving with a young family. So I found them a nanny agency.

Another time, a company told us that some of their employees felt as though their kids weren’t going to good schools after relocating for work. And there’s a problem that we could maybe one day solve, by having real estate brokers be part of our network the same way lawyers are. We’re also looking pretty seriously at expanding into relocation by partnering with moving companies the same way we do with lawyers.

All of these things, you could call them doing things that don’t scale. They’re experiments, no different than Uber delivering ice cream or kittens on random days. You just see what works. And some of them are crazy, certainly none of them are scalable, but some of them work and you go from there. Take taxes. As it turns out a lot of expats need help with taxes as well as immigration, side by side because one affects the other. And we can help them with both because we already have all their data from their visa filing. To return to your question, one day, I think that delight will come from bundling it all; if we solve enough of these problems in one place, then it’ll be a pretty delightful experience for the customer.

Guo: And how was the experience of growing beyond just small companies?

Richards: It was surprisingly not terrible, because we just started talking to bigger and bigger customers, at which point there is a market for the BaseCamp for immigration project management tool in addition to the visa work. So by going through investors, going through LinkedIn, going through whatever channel we could find, we hustled our way into getting [large technology company] as our first enterprise reference customer. And they use us as a complete solution: a project management tool integrated with a network of lawyers who provide visas and advice. One day, we want to add additional services like tax and relocation, and additional countries, so that customers can use us to move their employees to, say, China, the same way that right now they currently use us to move and manage expatriate employees in the United States.

If we keep going, one day it will be this seamless experience to move people all over the world. But it all started with me doing a visa for somebody. And we still have a long way to go. We still manually do some parts of visas; in fact, there’s a lot of work we do in house that we just haven’t figured out how to automate or send to our network yet, but that’s fine, one day we will.

An example is the government forms for visas. They require wet ink signatures, so you have to mail two hundred page documents out to get them physically signed, and the signatures have to be real wet ink. You can’t e-sign them. And because our lawyers are somewhere in Texas, somewhere in Seattle, and our customers are in, say, Mountain View, we’ve had to FedEx these forms to Texas, get them back, FedEx them to Mountain View, get them back, and then FedEx them to the government. It’s crazy, right? So we’re working on an autopen machine and some custom printing solutions to see if we can do this better. If it works, we won’t need to do FedEx anymore. And that saves us at least $100 and one week per visa.

So we always do things manually first, and then figure out ways to automate and scale it once we understand it; just like this autopen or our network of lawyers. It seems to always work way better when you start with the customer and then figure it out from there, as opposed to building something out before you have the customer.