My "I almost got rich at a startup" story

Silicon Valley, home of a million dreams. I've been in 7 or 8 startups, and I've never come close to getting rich. Steve Jobs used to classify bonuses as follows:

• We're going out to dinner!
• We're getting a new car!
• We're getting a new house!
• Fuck you! I'm quitting! (The proverbial "fuck you" money).

Now, I've had startups that have exited well. Radius went public, Chegg went public, Skyport got acquired by Cisco. But unless you're a founder, startups are kind of cheap with the equity. Your hope is that your grant will be worth $10/share. The most lucrative outcome I've gotten was working for an established company, making money, that paid me 80% of a years salary as a retention bonus after the acquisition, and then a further 25% because I was a grade 12 engineer. But it was also the worst company I've ever worked for. Daniel Pink says that people are motivated by Autonomy, Mastery and Purpose? I had none of any of that. I was often told NOT to fix things, which for an engineer is like being told not to breathe, I wasn't learning anything new, and the purpose wasn't there. But hey, making double salary. When they laid me off, my settlement for that gave me over 2x take home for the year.

Conversely, one of the best jobs I had for motivation was at Chegg Inc. I was working on the eReader team and we discovered our best selling textbook was how to use Microsoft Office. The eReader had a little trouble getting traction because it was more expensive than renting the physical textbook. Our Product Manager did a survey, and we discovered our main customer were waitresses trying to get an office job. They could read the textbook on their phone on their break. That's the Lord's work right there, huge Purpose. My boss, Dax E. gave me enough freedom to go around and be a bit of a free electron, so I stood up a metrics system that saved the company $4M the first textbook rush, so huge Autonomy, and I levered the metrics system to help with other projects like Tetris (cleaning up the financial backend so we could go public). Mastery entered in as a proceeded from lead, to manager, to finally Director.

But neither of those jobs made me rich.

The job that almost made me rich was Marketocracy. Marketocracy was the idea of this guy Ken Kam (he's dead now, sorry to say). Ken was a maverick who had started a mutual fund in Silicon Valley, not on Wall Street. He hadn't gone to the right school, and he wasn't born to the right family to become a mutual fund manager. So he and his partner did it anyway, and their pitch was that they were in the valley, so they could track and understand tech better than someone 3,000 miles away.

It worked, Tech Value Fund was obscure until they had the #1 track record in the country for 3 years. They went from $10M to $1B in AUM (assets under management) in a year, then they had the best record for 5 years, and they reached $4B in AUM. Since the mutual fund manager makes 2%/year, ?that's $20M/year per $B in AUM.

So what to do next? Ken thought, well if I can do it, so can others. So he set out to make a website that would give people $1M in Monopoly? money to virtually invest. He had Bruce Horn , the guy that invented the Finder on the original Mac team, as a friend, and he and Bruce reached out to me. At the time, I was both traveling around the world as a software consulting doing Y2K upgrades for my fathers company, and making $160/hour as a WebObjects consultant (WO was this framework Apple had for building webstites-as-computer-programs). But the Y2K thing was drying up, and while the WO work was interesting, scrounging up clients was tedious. Bruce wanted to use WO to build the site (which was a good choice) so it was a good fit.

Plus the idea appealed to me.

So now a brief digression on how Marketocracy was not supposed to work. So there's this thing in finance called the "Efficient Market Hypotheses". The theory says that everything in a stock price is already priced in; because the market is perfectly efficient.

Obviously, nobody actually believes this, because that would mean there would be no point in investing in anything, because the market would have already priced it correctly. The reality is that we can't tell the future, so the only people that believe the market is efficient are academics. What the EMH actually showed was that:

1. Wall Street wasn't that good at picking stocks, they couldn't beat a dumb index of all the stocks.
2. If stock picking ability exists, its rare.

Marketocracy's original plan was to wait 3 years, then anoint the best person on the site as a fund manager and use them to launch a new fund. Sort of a long, complicated interview process. After awhile, unable to convince the VC types that this was working, the company pivoted to looking into taking the top 100 people on the site and blending them somehow.

This is where I came in. I'd arcitected most of the site and written about half the software, but I was running out of challenge. So my boss/CTO Blake W. asked me to look into it. I built all the software for simulating the performance of a group of mutual funds, selecting the top 100, blending them into a master fund, then generating trades to track that master fund in multiples of 100 shares (you had to trade in those sizes at the time).

It did great at first. We could beat the Market by 15% a year, which was huge, and we had 3,000 stocks in the portfolio, so we were much less volatile. In financial terms, high alpha and low beta, supposedly impossible. But what I found from looking at the data was that #2 above is true, investing ability is about 2 out of 1000 users. With our 30,000 active users, that meant we had about 60 good people we were using. The other 40 were basically matching the market so they weren't hurting anything.

Life was good. We were growing our AUM naturally, and we were about 6 months from being profitable naturally. We weren't going to bother going public or getting VC money at all, we would just be able to do this complicated stock split/reverse split magic that would essentially pay a dividend without it looking like a dividend so it would be taxable as a long-term-capital-gain.

But then we invaded Iraq. And the Market got super emotional. We started trailing the market. Management panicked. Ken was happy with my picking algorithm as a black box when it was doing well, but he didn't know how it worked, so when it wasn't doing well, he started thinking maybe he should cherry pick the people. Which would have been a disaster, picking mutual fund managers is different than picking stocks.

So Blake stepped in and took over the algorithm work while I was working on something else, and tried some more momentum-based strategies. It seemed like those were going to work. Then they decided to take 30% of the fund and put it in SPY (S&P 500), 30% in the momentum picks, and 40% in the regular picks.

In real life, it did terrible. It took me awhile to figure out why, but the problem was the blending. Let's say you have funds A-Z. One algorithm predicts A, B, C. The other algorithm predicts A, D, E. The final algorithm predicts B, E, F. Blended properly, that would mean 2xA, B, C, D, 2xE, F. But we didn't do that, we would blend that as A, B, C, D, E, F, which means we were returning to the average.

Anyways, after that brutal mistake, we weren't going to be profitable. Ken brought in someone who was kind of an evil vizier type, he brought in a yes man to run engineering, and Ken got what I call "CEO disease" which is what happens to you when you're surrounded by people that will only agree with you. He and the Vizier would sit in his office and plot against the rest of the executive staff. Gradually, the entire executive staff left one at a time because they couldn't stand the evil vizier. Gradually, one at a time, after the yes man got put in charge, each of us engineers left as well.

So I didn't get rich.

Things I learned though:

From Bruce Horn I learned a lot about UI design, and he learned about web UI design from me.

From Robin Stevens I learned a lot about what a really top tier executive looks like. She was jokingly the VP of Everything Else, because she was a go-getter and early on in the company she volunteered for everything that wasn't assigned to someone else.

From Blake W. I learned how to herd a bunch of programmer cats with good humor, even when one of them could be a bit of a loose cannon (me).