I founded Tryllian in 1998, in Amsterdam. I had been working in AI for about ten years then, I was a pioneer in the use of the internet in business in the Netherlands, Tryllian combined both areas of expertise: we created chatbots that roamed the internet to communicate with one another. “Roamed”, as in we had servers where they’d meet, and they’d come to your computer to chat. The AI was limited: neural networks were just beginning, deep learning hadn’t been invented yet. As these bots were chatting and socializing, we called the application “Gossip”.
From a friend in Silicon Valley I learned that it is important to plan your cap table carefully. For those who don’t know what a cap table is, I’ll explain.
In a cap table, you forecast how much stock investors get for how much capital they invest, in the lifetime of your company, or until the exit. The primary goal is to set it up in a way that after each round of investment, everybody is happy. It is in investors interest that the founders are happy, that is, that in the end the founders have enough stock left to remain in control of the company. The first investors (angel investors and friends and family) get their stock cheapest, because they run the highest risk. Investors in the last round before exit don’t run a big risk, so they pay premium for their stock.
You can also plan stock or stock options for staff you hire. In the US that is common, in the Netherlands it is difficult because tax laws prevent it from being worthwhile. Instead, for employees that join at the start, I’d create an initial round of investment that allows them to buy stock cheaply, so they become cofounders.
I learned about cap tables from my friend in 2000. This was two years after I founded Tryllian. At the time, I didn’t realize that my cap table was already broken. Starting in early 1998, two angel investors invested in our startup, which in the autumn of 1998 consisted of five developers. Angel investors invested €160.000 each, in the course of a few months. When we formalized the contracts, they ended up owning 50% stock (25% each). At the time, I had no experience with founding a company with investors. I had started PayMate in 1995, a company that enabled internet payments in much the same way Paypal did a few years later, only we had two factor authentication. We talked to dozens of companies, large and small, all of them said “there is no future in internet payments”. It is like Henri Ford doing market research and learning “customers need faster horses”. For PayMate, a large insurance company paid for prototypes and research, but when it came to formal investment agreements and serious money, they backed out. PayMate never went anywhere, so Tryllian was my first experience with real investors.
In 1999, the Dutch government created the Twinning incubator in Amsterdam. They provided office space and investment, as well as help with finding VC capital. I applied for Tryllian, we got to rent offices in their new building at Amsterdam Science Park. Twinning invested a couple 100k and helped me get VC company NPM to invest another three million. However, as in the Netherlands investors usually make pre-money valuations, they received about 50% of the shares, leaving me with 25% of the stock.
The people at Twinning, none of them having significant VC experience on their resume, had toured Silicon Valley where they had talked to VCs and entrepreneurs. One of the things they were told is that a startup company needs a “professional CEO” to grow the company. Hence, soon after Twinning and NPM had invested, they came up with their idea of a “professional CEO” for Tryllian. They found a candidate, a man who had worked for a bank for 12 years. I remember going through his resume and my first thought being “no way”. I told the board. They pushed, I met the candidate, a man in a gray suit and a tie, and again I said “no way”. However, because I was a minority shareholder, despite the fact that I was the founder of the company and still the CEO, they pushed through. That was how Tryllian ended up with a CEO whose only qualification was “guy in a gray suit and a tie”. He did not have significant experience in either management or software.
In 2000, I thought we shouldn’t want to wait until we’d run out of money because I realized, once you badly need new investment, you’re not going to get it at favorable conditions. So I started talks to VCs in the Netherlands, UK and Belgium. In 2001, we finalized negotiations with a consortium led by a Belgian VC, and we secured an €11 million investment. As we had only a dozen or so employees, mostly developers, one devop and an office manager, 11 million would take us far, even without revenues.
I was traveling a lot, promoting our technology, speaking at Demo2001 in Phoenix, AZ, and in New Orleans at a large internet event, having a booth at JavaOne in San Francisco where we won a “cool app” award, presenting in London, and I was invited to the World Economic Forum in Davos as a Technology Pioneer. In the mean time, the company hired dozens of people, not just developers, they were all kinds of people. We had four people in finance, three in HR, a “building manager”, marketing people, sales people, and others. Also, we moved to a new office space. We were spending a ton of money on other things than making software. Every Monday morning, I’d make a list of employees, marking which employees contributed to creating or selling the product. I found that every week, the percentage of people that contributed to the product dropped. The CEO kept insisting that this was ok. We went on like that for months and months.
Then finally, after a few intense conversations with the finance guys, I went to the board and said “this CEO needs to go!”. I had 12% stock left, I had no significant power in the board, I couldn’t just fire him by myself, or I would have done that already. It took a while, before finally I convinced the chairperson of the board that the CEO needed to go. By that time, she could see the man had already spent a significant part of the 11 million without generating any visible result. He was fired the next day. I spent a few more months with the company, but I realized it was broken beyond repair. The CEO in the gray suit had discontinued the only product that gave us visibility in the market, our little roaming Gossip chatbots. He had just pulled the plug on it. I left the company, and that was it.
It took me a few years to reflect and realize the mistakes that I made.
First, I do think that if I’d been a guy in a suit, it wouldn’t have happened this way. The cap table would have gone wrong, probably, or maybe less so, but they wouldn’t have hired a clueless CEO just like that, the company wouldn’t have spent so much money in so little time, and we’d have had time to further develop our product for years and years. We were AI-based social media, if we had still been there when social media took off, we might have been wildly successful.
Second, I learned that as an entrepreneur, you should not let other people tell you what to do, not even when they have invested millions in your company. I could have said “you hire the guy, I’m out”. In retrospect, I don’t think they would have hired him. But I had no experience in these matters, I felt intimidated, I gave in. Instead, I should have kept believing in myself and not allow insecurity to take over. Because whatever happens, it’s better to fail on your own account than because of mistakes others make.
Third, use post-money valuations in investment rounds and set up your cap table accordingly. In order to push the company forward, you need to be in control of it.
This happened 20 years ago. I still meet people who say “hey, I remember you, you founded Tryllian, these little Gossip bots, what happened to that?”. I don’t know what to answer them. Yes, we had those Gossip bots. And no, they’re not around any more. The company didn’t make it, and I find it hard to explain how or why. Well, now I explained. In the few years the company existed and rose to fame, I made two grave mistakes, and I paid for those two mistakes. I have moved on. With a dent in my ego.
