I’m an entrepreneur, investor, engineer, math puzzle fan, squash aficionado (the sport, not the vegetable), and enjoy anything dealing with AI, technology, payments, finance, medicine, commerce, advertising, and marketplaces.

My (long) background:

I was born in Palm Beach, Florida, where I went to Palm Beach Day School, and got exposed to BASIC on an Apple II and HyperCard on a Mac. I got a Mac as my first computer, in (I think) 1988 or 1989, and was frustrated with the dismal software selection for Macs…so got a book on C which included a basic C compiler (THINK C, for those who remember).

I started writing my own software, and then uploading it to various BBSs as “shareware” (“try before you buy”), and scored my first sale (check in the mail) for $5 for MiniScreenSaver. A few days after that, I sold 4 copies and got a $20 check in the mail. I think I was 11 years old.

This blossomed into a little business earning ~$100/week throughout grade school and high school. I went to a boarding school, Andover (Phillips Academy), and was running my little business from my dorm room, and using AOL (America Online) as my “internet” + email.

In December 1996, AOL stopped charging by the hour (used to be $3/hour), and instead switched to unlimited use for $20/month. Laws of supply and demand being what they are, it became impossible to connect to AOL, and AOL would just dial each number a few times and then give up.

So I wrote a program called AlwaysONline, which dialed each number 999 times in a row, and then kept you online forever to prevent AOL from kicking you off. I was 15, and then the checks started really flowing in. So much so that it was hard to handle all the mail, which is when I figured out how to process credit cards “online” by having people email me their encrypted credit card numbers, and then charging them in the credit card terminal in my dorm room.

Forbes (back when it was big!) did an article about me, and I learned that any publicity is good publicity if you’re trying to get distribution of a useful product.

I followed this up with more and more products…and also upgrades of existing products, sold to my existing customers (SaaS before SaaS).

I went to Harvard where I studied Applied Math with Computer Science, but kept running my little “business” with more products and always striving for a little controversy or some unique distribution edge. One of the more popular ones was called TypeRecorder, a key-logger which I built after I lost an English paper I was working on (hadn’t saved it yet…) when my computer crashed…but turned out to be useful for spouses trying to catch their partner cheating. That led to a collaboration with a company called SpectorSoft, where I built the Mac version of their “see what’s happening on your PC” product.

When I graduated from Harvard, I just…worked for myself, in Harvard Square. It was an odd and non-traditional choice in 2003 but my business was working.

But I was dating a girl (now my wonderful wife!) at HBS at the time, and during my Senior year, I got introduced to this guy Chris Dixon, who was at HBS. My girlfriend’s friend’s boyfriend was friends with Chris, and it was one of these “oh you like software and the internet? I know somebody else who likes software and the internet!” moments. We met for coffee at Au Bon Pain, a coffee shop in Harvard Square, for several hours.

Chris had a side-gig building software for an email marketing firm, and he mentioned how they embedded tracking pixels (a 1×1 image which, when fetched, showed that the email was read) in emails. I said that would be a great idea for a consumer product, and it would be very controversial.

And thus DidTheyReadIt was born, and eventually launched in 2004…to great controversy since you could see if an email you sent had been read, and where it had been read, and even if it had been forwarded. It was a recurring subscription and did very well.

Meanwhile, Chris went to work at Bessemer Venture Partners, and he and I talked regularly. We cooked up an anti-phishing product called FraudEliminator, and eventually that became SiteAdvisor, a venture-backed company (by Hemant Taneja at General Catalyst, and Rob Stavis at BVP!)…which one year later was acquired by McAfee for about $75M.

This windfall caused me to re-asses whether I wanted to keep doing my little shareware projects. And for my shareware, I had come across a very interesting idea to get more people to pay. Only about 1% of my users would ever pay. People just don’t like paying for digital goods like software or music, versus tangible/physical goods like food or furniture. How about I let them get my products for free if they buy/try something else?

Get my product for free if you sign up for Netflix, or shop at Gap, or get a Discover Card. This idea came from a conversation with my friend Greg Tseng (very fruitful breakfast!) and I thought it could be applied to payments. Soon most of my revenue was coming from this, and when SiteAdvisor was bought, a big chunk of FraudEliminator’s sales were from this method.

So I started a new company called TrialPay in April of 2006, paired up with somebody that Chris Dixon introduced me to (Terry Angelos) and a good friend my college (Eddie Lim). Many years of ups and downs, some of which I’ll try to explain on this blog, but we eventually sold the company to Visa in 2015.

TrialPay’s biggest market ended up being social games (don’t want to pay for virtual coins? get them for free by doing X). We worked with all of the big social gaming players except Slide, and I think at the Allen & Company conference I connected with Max Levchin, who ran Slide. I was unable to persuade him to use us (partially because he sold the company to Google thereafter!), but I was able to persuade him to become and advisor, and we became friends. I credit my studying Russian in high school with helping facilitate that and standing out from the noise!

Max was looking for something to do outside of Google, and we cooked up something we called Expedite, which we incorporated in 2012. It was hard to pay for something on a mobile phone in 2012 — too many fields, too much time. What if you can pay with your identity, allowing merchants to convert more customers, and then just pay us back later? The original idea was log-in with Facebook (Facebook Connect), and if you have enough friends and pictures, you’re *probably* a real person, so let’s make you a loan.

We couldn’t buy the domain name Expedite.com (we tried!), so were expedite.cc, and then a naming consultant gave us three options: Thumbprint, Affirm, and Trustable. Affirm it was.

I helped get things off the ground and coded the first demo (for our first customer, 1800Flowers), Max became full-time CEO, and eventually the company started really working when we went from the first bad idea (social loans for low value merchants) to serving Casper and Purple mattress — high value merchants with improved yet traditional underwriting.

I thought we were going to sell TrialPay in 2012…and came very close with Google and PayPal, but no…so I was still doing that.

When we finally sold TrialPay and I was working at Visa (but still helping out at Affirm, TXN, and Point), I got lots of calls from venture firms saying I should join them. I asked Chris Dixon for advice, he said I should obviously work at a16z, and eventually I agreed!

I’ve been at a16z since 2015, where I first started off doing Fintech (and where Affirm was an existing investment, as Max and I brought it to Jeff Jordan when we were raising an early round!), leading deals/serving on boards in Mercury, OpenInvest, Wise, Earnin, Plaid, etc. But once we “verticalized” the firm my focus is on “Apps” (think anything B2B or B2c, versus developer tools/infrastructure).