A bona fide pioneer of the mobile ecosystem
Podcast transcription - 24th April
Peter Cowley: Welcome again to another Invested Investor podcast. Today we have Dr. Ronjon Nag, who has had a stellar entrepreneurial career. He's been a wonderful angel investor in Silicon Valley and the U.K. So Ronjon, welcome. Can you give us a bit about your background and how you became an entrepreneur and then later, an angel?
Ronjon Nag: I'm an electrical engineer by training, Birmingham University, and my final year project was in speech recognition. This is like 1984, when I was doing telephone digit speech recognition. And that got me into Cambridge to do a PhD in that area. Not many universities were researching that area.
Today we have Siri and we're quite blasé, but at those times I'd give a demonstration of speech recognition, for telephone digits spoken continuously, usually you had to keep a pause between each word, and people would fall off their chair. "How do you do that?" And I think it was Arthur C. Clarke who said that any sufficiently advanced technology cannot be distinguished from magic. And people would think it was magic.
Peter Cowley: And this was only 10 possible combinations, wasn't it?
Ronjon Nag: Yeah, yes, yeah.
Peter Cowley: Naught to nine.
Ronjon Nag: Naught to nine, but 10 digits in sequence. So it was quite a difficult problem at the time. And I was enthused about it by my watching of Star Trek. And I'm always asking young people today, "Why are you an engineer?" And I tell them that I was an engineer because I watched Star Trek, I wanted to build things that were in Star Trek. And then I suddenly realise they're thinking about Star Trek the movie and I'm thinking about the first generation TV series.
And so, I was at Cambridge and working on, there's a popular technique called neural networks, which is quite popular now. But I was working on it in 1985, '86. And I guess I'm, right now, known as the first neural network company to come out of Cambridge University. Sometimes they call it deep learning today, and it's the sort of foundation of current artificial intelligence.
I left Cambridge and got this fancy scholarship to go to America. Went to MIT and worked on handwriting recognition for bank checks. This is like, again, 1989 or so.
Peter Cowley: That was a master's?
Ronjon Nag: That was a master's at the business school actually, but at MIT everything has to have technology with it, so I ended up with this financial technologies project. And then I went to Stanford.
Peter Cowley: So what moved you to Stanford, another postdoc?
Ronjon Nag: It was the same fellowship programme, and I think that's one of the reasons I got it, because I think I said, "Well, I want to do a bit of science and a bit of management science," and I think they quite liked that. Most people would just say, "I want to do an MBA," or something.
I wrote this letter to the most famous person in the field at the time, which was Professor David Rumelhart, who unfortunately passed away about 10 years ago. But at the time he'd come up with a way of how to compute the parameters of a neural network, which are basically a computational model, a very simplistic model, of how the brain might work. We have about 90 billion neurons in the brain. Unlike a serial computer that we have in our PC, each of these neurons is connected to maybe a few thousand other neurons. And he'd come up with a model on how to create a mathematical model of that process and how to come up with the parameters.
All his papers said Professor Rumelhart, Department of Computer Science and Psychology. And he said, "Sure, come over. I know your professor at Cambridge. Come over." And I thought I'd be in, of course, in the computer science department. But of course it turns out I ended up being in the psychology department. So I was a bit nervous because I didn't know anything about psychology.
But it turned out great. What they were doing, what we were doing, was trying to figure out a mathematical model of the brain. And he'd collected mathematicians, engineers, linguists, physicists, to work on this problem. At the time it was a very popular method. I was finishing my postdoctoral work, it was 1991 or so, it was the end of the Cold War. And you might think that engineers have no problem getting jobs, but actually at that time there was actually a period of weak economy across the world.
Peter Cowley: Certainly in the U.K. it was very weak. Yeah, I remember that, yeah.
Ronjon Nag: Yeah, because of the Cold War just finishing, and so the defence industry, all, of a sudden was cut worldwide, and that had ripple effects right across the supply chain. And of course the miraculous thing about Silicon Valley, of course, is when you can't find a job what do you do?
Peter Cowley: Start a company.
Ronjon Nag: Start a company. And so I was working at the time, my background was in speech recognition, mainly, at Cambridge. And I'd noticed that people were trying to build what they called PDAs: Personal Digital Assistants. And these were little tablets. We call them tablet computers today.
Peter Cowley: The Apple Newton was an early one, wasn't it?
Ronjon Nag: Yes, Apple Newton, then there was the PalmPilot. And people were trying to create these tablets. Some of them were small, some of them were big. And the model there was that it was going to replace paper. This is quite different to today, but the idea we would have tablets that would replace paper. And interestingly enough, no one was really working on the handwriting recognition part, despite having that vision. And people were using what's called print handwriting, where you write one letter at a time, with a space all around the character. I'd read an article at the time it was called Byte Magazine.
Peter Cowley: Oh yes.
Ronjon Nag: They said, "Cursive handwriting recognition's impossible, and so we can only do print right now." And I said, "Well, it didn't work that well, but we have techniques like this in speech recognition to remove the spaces and know where one word starts and one word ends." I said, "Can we use these similar techniques in handwriting?" So, I started working on it, with Stanford and Professor Rumelhart, we started working on cursive handwriting recognition. And it very quickly got off the Stanford computers because we didn't really want to be contaminated, as such.
There is one story, we were working away, and the Microsoft people found us. And it was me and my cofounder, who was a graduate student at Stanford, a psychologist. He couldn't get a job because he had a psychology PhD. He was better in computer science than I was. The Microsoft people did background checks on us and found out that we had the same professor. So they visited Professor Rumelhart and said, "Oh, can you give us all the data and all the code?" And Professor Rumelhart didn't believe in patents. He thought, he's sort of old school.
Peter Cowley: Yeah, open source everything, yeah.
Ronjon Nag: Open knowledge, all knowledge is free.
Peter Cowley: For society, yeah.
Ronjon Nag: We're a university. I'll tell anybody anything they want. But he called me up and said, "Well, these Microsoft guys, they've asked me for my code and my data." He said, "What do I do?" And I said, "Well, I think it's your code," and he was working on a different concept, "it's up to you."
Nine months later we met the Microsoft guys and they said, "He said he could send us his code and his data and never did." So he was kind of like helping us, trying to help. And so, we were a classic Silicon Valley start-up company. We were two guys in a garage. We finally hired one other person, a Rhodes Scholar, actually, from Oxford. And there's now three of us.
Peter Cowley: And capital? Did you get any investment?
Ronjon Nag: The total capital put in was $500, which for me, by the way, was expensive. And so we manage to get our first contract from Apple. And Apple, though, was working with our competitor. There was only two companies in the world who could do the cursive handwriting recognition problem, where you could just join up the letters. And as you start writing the word "minimum" or "potato" or something like that ... sometimes you don't know whether it's an I or an E or ... so it's quite a difficult problem.
So there's us, which is two guys in a garage in Palo Alto, California, and our competitor was a hundred people in Moscow. This is, again, like Cold War, borderline. And so you can imagine the poor Apple product manager who's in a no-win situation. We both had roughly even credibility. Who do I bet on? And so they gave us some money to get a little prototype. We ended up working with all of Apple's competitors. And they ended up working with the Russians.
Peter Cowley: On the Newton?
Ronjon Nag: On the Newton, yes. And just to back up, the Newton was a PDA and it had cursive handwriting recognition, but it was oversold. It was oversold. Apple had oversold it and said basically it would do everything, you know, make toast and slice bread and everything.
And then there were jokes about the handwriting recognition errors that were made by cartoonists, and sort of put a damper on the market somewhat. But we kept going. We kept going.
And then I was lucky enough, you know, I had a friend who worked at Xerox PARC. He said, "There's this guy at Fortune Magazine. He's visiting us. I told him to go and visit you because you're trying to make a business out of this artificial intelligence stuff." So he came over and he wrote an article, and then one day he called me up, he said, "Well, it's a rather boring issue this week, so we've decided to put you on the cover." And unfortunately for my two cofounders he'd cut them out of the picture. I kept saying, "It wasn't me." But I guess they wanted an international looking face on the cover. And two months later, Motorola bought the company.
Peter Cowley: Bought your company?
Ronjon Nag: Bought my company.
Peter Cowley: But you'd morphed a bit. You weren't just doing cursive handwriting recognition, were you?
Ronjon Nag: No, we were, that was our main thing, doing cursive handwriting recognition, a little bit of speech, but mainly cursive handwriting. We were still only three people, bought for $12 million.
Peter Cowley: But that was to hire you, surely. Or was it some tech? The IP?
Ronjon Nag: No, the way I look at it just from an angel investing perspective, first, of all, they only had a choice of two companies in the world. What creates value? One is uniqueness, right? And they looked for the other company. Culturally we were a better fit with Motorola. And the company name was called Lexicus. So, the investment was $500.
A little anecdotal story, just towards the end we basically had run out of money. And we did two things: one, we asked Motorola, "Do you think you can lend us some money while we're negotiating?" And they kindly said yes. But of course that gave us more negotiating power. I think while they were negotiating, I think corporate headquarters didn't quite realise we were only three people. Obviously the corporate dev department knew we were-
Peter Cowley: Yes, because they'd come to see you.
Ronjon Nag: The price is going up and up. And then finally they said, "Okay, they know you're only three people. No more money. This is plenty."
And then towards the end I'd asked my mother, I said, "By the way, can you wire me £25,000 please," which was like 98% of her net worth. And my dad said, "No, no, no, don't send it to him," right? But of course she was always very loyal to me. And so she goes into the bank, and it was a Co-Op Bank. And in those days I guess bank managers knew their customers. So the bank manager walks out, and then tries to dissuade her from sending this amount of money to her good for nothing son 6,000 miles away, which apparently, I guess, happens where the children fleece their parents. But she goes and does it. So, I figure, well the worst case, I'll have to get a job somehow and pay it back.
Peter Cowley: Exactly, yeah, exactly. Course you will.
Ronjon Nag: Then, two months later, Motorola buys the company for $12 million.
Peter Cowley: How long were you tied in for?
Ronjon Nag: Technically three years, three years.
Peter Cowley: Okay. Did you work through those three years?
Ronjon Nag: I did. I actually stayed longer. I ended up staying six years. And Motorola treated us very well. They treated us very well, and very gently. They were very concerned because we were almost the opposite of the rest of Motorola, which are hardware people. We were very young. Certainly, when you get acquired, if you are young, you get accelerated through the promotion process. You get to be vice president, where most other people would spend 20 years in the company. And here I show up. I was a kid basically.
Peter Cowley: A lot of resentment?
Ronjon Nag: A little bit. I think culturally they were quite good, and very careful to protect us until we could work out the lay of the land and look after ourselves.
Peter Cowley: So the cursive handwriting became K9?
Ronjon Nag: No. No, no, no. So the first thing they did, Motorola said, "Okay, we need you to do Chinese handwriting recognition." And I said, "I don't speak, read, or write Chinese." And they said, "We don't care. Just get me Chinese handwriting. Buy a company, hire people, just go do it." But fortunately I speak mathematics, and we did hire some Chinese engineers and we created, essentially, the first Chinese handwriting recognition systems.
Peter Cowley: Used on which product?
Ronjon Nag: On the Motorola pen phones, and this is another story. Just before I left Motorola we built a phone. It had a touch screen, had handwriting recognition, had speech recognition, had a full screen browser. It had games. It had applications. We sold millions and millions and millions of them in China. And of course I said to my colleagues at Motorola, I said, "Well, okay, let's just ship this in the U.S. now, in Europe." And they said, "Look, Ronjon, you don't understand. In Asia they think pictorially, gestures, and so this will work in Asia. It won't work in the West. Look at BlackBerry. They're eating our lunch. We need to create a keyboard phone." I said, "Okay, whatever."
So no, it's the end of the '90s and the Internet boom is coming, and I'm saying, "What am I doing here? Why am I doing corporate? I should be out there doing stuff." So I then started a second company, Cellmania. Because I had seen all the five year roadmaps at Motorola, I knew that Internet was coming on the phone.
And so Internet on the desktop was all about search engines and mobile directories. Yahoo! was the earliest pioneer in this area. So I said, "All we need to do, then, is kind of like a Yahoo! but for the phone." And people were just building these little small, tiny websites, and they called them WAP sites, Wireless Application Protocol, to fit on a phone. Well, pretty much over the weekend, we typed in every single WAP site in the world, which was like 20,000 of them. It's not like millions. At the time it was only 20,000. And so over the weekend we'd built ourselves the biggest mobile directory "search engine," just over the weekend.
Peter Cowley: How did you find all these names?
Ronjon Nag: Well, you've got the web, right? So you just use regular Yahoo! and you find the URL for the actual phone version, and I had a team of content people who would just sit and find all these links. And the only problem was that those early days ... all these people were writing these sites, and later on people wrote apps as opposed to sites, there's no billing system. So, unfortunately a lot of them, there's no way to monetise it. And a lot of the mobile operators were almost doing the opposite, as in telling these little tiny companies, "You pay us to be in the directory."
So we made the independent version. Unlike Lexicus, which was bootstrapped, $500 in, $12 million out, this was more, "Okay, Motorola plonked down $1 million"
Peter Cowley: For shares?
Ronjon Nag: Yeah, for shares. We ended up raising about $13 million total.
Peter Cowley: But not in the short term. The first round was a million dollars?
Ronjon Nag: The first round was a million dollars. And then the second turn, it was basically another, maybe, $9 million.
And so the first valuation is at $6.5 million, and the second valuation was $100 million in 2002. And so, we spent a lot of that doing various things, ramped up to 80 people. And the third valuation, you know, we'd spent most of it. And then of course 2002, dot com bust, didn't matter who you are, how good you were, no one was writing any checks, right?
And so we had to make the decision. I had to talk to the wife and say, "Look, we've either got to take the winnings from the first company and put all of it, the whole lot, into the second company, or we have to close it down." So she said, "Okay, dear. Well, I trust you. Off you go. Do what you need to do." But she didn't quite understand the ramifications until after I'd done it, which meant basically I was in the doghouse for five or six years.
Peter Cowley: So you put several million dollars in?
Ronjon Nag: You've gone from student life, pre-Lexicus, then all of a sudden there's a million dollars in the account, and so we lived a nice life, and then, "Okay, take it all and put it all in the next" They call it double down in Las Vegas. And the third round valuation was at $10 million.
Peter Cowley: So down from 100 to 10?
Ronjon Nag: So 6.5, to 100, to 10. And at that point we had $10 million of preferred overhead. That means you have to sell the company for more than $10 million before anybody, any of the employees or any of us, would make any money. And so we were essentially valued at negative $10 million.
Peter Cowley: Yeah, yeah. That was a preference share of some sort, yeah?
Ronjon Nag: Because the preference, there's maybe a subtlety that often, I think, a lot of entrepreneurs, maybe even today, don't know. You read these things about preference shares, I never really understood what it was until you actually get to the "Oh, that's what preferred means." It means that all the investors have to get their money out first, and then it's given out to everybody else.
Peter Cowley: It's very common though, isn't it, at VC level.
Ronjon Nag: Yeah, which is fine if you sell for many, many multiples. No one really cares.
Peter Cowley: And you bought preference shares yourself when you doubled down.
Ronjon Nag: That's right, I ended up owning a lot. I doubled down and bought preference shares myself, but then I had these shares from the old round that are a hundred million pre. And we didn't have kids at the time, and so I said, "Well, these shares are basically worth nothing, literally nothing, because there's $10 million of preferred overhead. And they're worth nothing." And I sold them to my nieces and nephews for $1.
Peter Cowley: In the West Country of England?
Ronjon Nag: Yes, correct, that's right. In the U.K., in the Midlands, East Midlands. And I forgot about it, right? And they didn't know what they were. They really didn't know, I said, "Well, here's a birthday present. Here's your shares."
And so for the next five, six years I was in the doghouse. But slowly and surely, we'd invented the first mobile app store in 1999. No one knew what that was, the first mobile directory. Our main customers were mobile operators, you know, Orange, and AT&T, and Telstra, these kinds of companies. And we would basically private label them. And we would do the work. We'd test the apps, find the apps, pay the developers, download the apps, make sure no one can steal them. And we'd run it, and we'd get a small percentage.
Peter Cowley: Of the sale of the app, effectively.
Ronjon Nag: Exactly.
Peter Cowley: Yeah, yeah.
Ronjon Nag: And then in the beginning when there weren't really any smartphones and not many people download, it wasn't very much money. But as it went through the cycles, it started off, you know, ring tones were a big business. That went to an $8 billion market for ringtones. And then people stopped buying ringtones. They bought games and they bought apps. And slowly but surely our revenues went up. We stabilised.
Peter Cowley: So when did you break even, then, in '06?
Ronjon Nag: No, no. We were forced to break even, it wasn't by choice, forced to break even for about a period of two years between the second round..
Peter Cowley: You had to cut costs to do that?
Ronjon Nag: We had to cut costs. Went down from 70 people to seven.
Peter Cowley: Really?
Ronjon Nag: And this is what basically sorts the men from the boys. In entrepreneurs, often if you take enough time horizon, often there's these upturns and downturns. I'm always telling my entrepreneurs, "It's never as good as it looks. But it's never as bad as it looks either." But a lot of entrepreneurs, particularly in that dot com bust era, just gave up. And for a period of two years we had $5,000 in the bank, and we had $100,000 burn rate. But we had $100,000 revenue. We had like $95,000 of revenue every month. But our costs were $100,000, but we somehow made it through, and then to a final, very small round, just for a few hundred thousand dollars, which for us was luxury.
Peter Cowley: And which valuation was that?
Ronjon Nag: This was the 10 million.
Peter Cowley: Oh, that's the 10 million. Okay, fine.
Ronjon Nag: But we did that round with just a few hundred thousand dollars.
Peter Cowley: Yeah, it makes so much difference.
Ronjon Nag: Relax, yeah. We felt, yeah, it was luxury.
Peter Cowley: You've got some money in the bank every month, yeah.
Ronjon Nag: And then we got this key contract with Sprint. And this is the other lesson I give companies: "Look, exist. You have to exist. If you believe your concept is there, but there's a market timing risk, and often there is because by definition you're innovating and the market may not realise they want your product or your company, you have to sort of pace yourself and exist. Come hell or high water, you have to exist until the market wants you and is ready."
We got big contracts from mobile operators, built our cash base, but we were so traumatised from having no money we didn't spend it. And we said anything we do, we want 10 times our money within two years on any project.
Peter Cowley: Even hiring somebody as well?
Ronjon Nag: Yes, exactly. And so as a result we didn't actually do much diversification. Maybe if I had a wider board they would have said, "Well maybe, Ronjon, you can take a bit more of a risk and try these things."
But we kept going and then around about 2008 we met with BlackBerry. At the time they were called Research in Motion. And we said, "Look, we've spent nine years at this point working on mobile app stores." Apple had just launched their app store in 2007, a year beforehand. So we said "BlackBerry, do you want some help, because we've done everything you can think of: selling videos, selling music, selling apps, subscription billing, PayPal billing, adding it on to the carrier bill, clubs, you know, games clubs." Like, "Do you want some help?"
They said, "No, that's okay. We've got 50 people working on it. Thanks for visiting." And we went back and kept working. And people were afraid for us because Apple had just come out with its app store. They thought it would cripple our business, but actually our business went up because all their competitors wanted an app store too. And people suddenly realised what it was. They went "Oh yeah, we want one, we want one."
2008 was the financial crisis, as we all know, so no one was going to buy any companies. 2009, leftovers of the financial crisis, no one was going to buy any companies. 2010 there's a window. So this time I hire a bank. And I'm thinking, "Why do I need a bank?" Now I know all the wireless people. At this time it's been 10, I know everybody in mobile. And I hire a bank and I find out the reason to hire a bank is not because I'm stupid. It's because when the CEO calls up, a partner, potential acquirer, it automatically goes into sort of, "Okay, how can we partner? How can we licence?"
When a bank calls up, they know exactly what you're calling for. You might get a quick no, but you don't beat round the bush. Their client knows this is probably going to get sold. If they want to do it, then they need to act. And the banker, had all these lists of companies to meet with, and RIM was one of them, BlackBerry was one of them. And so I rung up, "They've got their own guys. Why would they want to buy us?" And said, "No, no, no. I think we would go and meet them." So, I say, "Fine." We’ll go and meet them. So, we met them and then RIM says, "Oh, right, we need to buy you. We need to buy you right now," not because their guys were stupid, but because the market was moving so quickly. Their corporate development team had realised that they haven't got time to build these things.
Peter Cowley: But they'd been working on it for two years already, that's a hundred man-years of effort.
Ronjon Nag: Well, we'd spent a thousand man-years of effort, right? We spent 10 years working on it. And we'd done every single variation. They bought like five, six companies very, very quickly, news interface companies, mapping companies, and they were going to go head to head against Apple and Google. And their argument was, "Well, we don't want to do Android because everyone can just buy the same part. What advantage would we have?"
And I think probably internally we thought that the Android system was sort of undervalued. It was a Canadian company, probably a little bit humble, and they didn't undervalue the power of the BlackBerry brand, may not have realised that, all things being equal, people would have chosen BlackBerry because they had a third of the market, third of the smartphone market. And I think today Samsung has that. They basically took that approach in that area. So they ended up buying Cellmania, and so I made about 10 times as much.
Peter Cowley: The second time round?
Ronjon Nag: The second time round. So it took 10 times as long, but 10 times as much. Sold it to BlackBerry, worked for two years, got the phone out, and then this time I thought it's time to do something else.
Peter Cowley: Did you sell for cash or shares or both?
Ronjon Nag: Cash. Yeah, all the companies I've sold so far, they all think their shares are undervalued, and they all want to give us cash instead. And actually BlackBerry, when they did buy us, their share price doubled about three months later. So my argument, we were free basically, because of course it was all me, wasn't it?
Peter Cowley: Of course. What happened to the other customers? Did BlackBerry switch them off, these app stores?
Ronjon Nag: Basically, yes.
Peter Cowley: They must have hated that, some of them.
Ronjon Nag: Yeah, some of them were BlackBerry customers, we had to manage that.
Peter Cowley: To negotiate that, yeah.
Ronjon Nag: But the trend, I mean, BlackBerry, RIM, they were only half joking. They said, "Look, you better sign quick because there's only going to be four app stores. I know you've got 20 mobile operator app stores out there, but you better sign quick because there's only going to be four: Microsoft, Google, Apple, and BlackBerry. And it's basically ended up pretty close to that. It's really almost, now, a duopoly. And if you're not one of those guys, what I said to them, "Look, we're, at this point, 12 years old. If anyone can adapt, we can." And, you know, we had competitors and they did adapt. They did TV app stores, car app stores. There's still space, you know, you can adapt. If you're in technology, by definition the market will change. This is not like building coffee shops. The market will change.
Peter Cowley: Yes, you need to, adapt to it.
Ronjon Nag: And you need to adapt.
Peter Cowley: So you sold to BlackBerry, then, Ronjon. And how long did you stay there to earn your, earn out?
Ronjon Nag: Two years. Actually a bit longer than that. I think one of the things when people give you lots of money, I think one should be loyal. I know some entrepreneurs leave on the day, but I stayed till the project was finished, for the new phone they were going to try and compete with Apple and Google. I then left, and then advised Vocal IQ which did speech recognition again. They got sold to Apple. So my story is, sold one to Motorola, one to BlackBerry, and one to Apple.
Peter Cowley: Well done.
Peter Cowley: Thanks for listening to another Invested Investor startup experience. Ronjon has had two fascinating journeys [inaudible 00:26:50] in to Motorola and what is now BlackBerry. A key tip to entrepreneurs is you have to persist and exist, and when the market wants you, you're around to be eaten up. Look out for part two next week.
Peter Cowley: Thanks for listening to another Invested Investor podcast. You can subscribe to all future podcasts via our website, investedinvestor.com, or via a number of podcast platforms online. Remember, you can order our book online. And be sure to follow us on Twitter, LinkedIn, and Facebook to get the most up to date, interesting, in insightful content from the Invested Investor.