This HN thread by the founder of Github on side projects got me thinking. You could make a strong argument that my most successful projects--NamesDatabase (my last startup) and DuckDuckGo (my current startup)--essentially started as side projects and, perhaps more interestingly, evolved from other failed side projects and startups.
I thought it would be illustrative to enumerate my projects. I'm limiting the list to projects that I started (as opposed to consulting/affiliate gigs) and ones that I spent a significant amount of time on (as opposed to just a few weeks/weekends).
- BBS (1993-1995). Ran a BBS out of my bedroom. Forgot the name. Yes, people would call into it in the middle of the night and it would make that modem sound. The best feature was I scheduled anonymous chats with my Dad, who is an ID doctor covering HIV/AIDS.
- Silkscapes (1995-1996). Made an online store for my Mom's offline clothing business. Launched it, but didn't really attract any customers.
- Velvet Donk (1996-2000). An e-zine featuring essays from me and other people I knew. Picked up a lot about HTML layout messing around with it.
- Donk mailing list (1997-2001). Essentially a fark.com over email, curated by me. People that I didn't know would ask me to be added to it. I thought about putting it online, but didn't. Should have.
- Apathy (2000). One of the only one of these projects still online. The idea was to create a "Web" album, analogous to a music album. I spent much of my last semester in college working on it. Had plans to do more albums, including a Startup one, but haven't gotten around to it.
- learnection (2000-2001). My first "real" startup, which was an attempt to increase parental involvement in primary education by using the Internet to give more info to parents about what was going on in the classroom. So many reasons for failure here: bad hiring, bad customer development, misunderstanding the sales cycle, etc. etc. Lessons learned though.
- The Public Inbox (2001-2002). Essentially posterous. You would email a general address, i.e. firstname.lastname@example.org and it would put it on the Web and send you back a Web page. Failed to iterate and stick with it, but learned a lot about SEO and email.
- Journal (2001-2002). Essentially a blogging platform. You would email a general address, i.e. email@example.com and it would blast it to your friends and then also archive it on the Web for you. Failed to iterate and stick with it.
- Zoofoo (2002). A coupon site. Attracted a decent user base via SEO and some word of mouth, but never really made much money from the affiliate stuff.
- Simple Email Client (2002). In retrospect, this should have been a Web app, but it was installed software. Spent a bunch of time on it, but never launched. Should have at least launched it.
- Yahoo! Store data analyzer (2002). This would download all your store data, make a database out of it and then let you run queries against it to discover things you couldn't do easily in the store tools. Spent a bunch of time on it, but never launched it either. Should have at least launched it.
- Fifty Rules Used by Highly Successful People (2003). I wrote an ebook and sold it online, with some encouraging early success. I took it down when NamesDatabase took off and because it felt a bit silly, but perhaps should have stuck with it.
- NamesDatabase (2003-2006). My so-called "last startup" because it was the one that had a successful exit. It was essentially classmates.com, which is who we sold it to. It grew out of The Public Inbox weirdly enough. A lot of people would send in emails looking for other people, which would get indexed and sent a decent amount of traffic. So I myself put up pages of names, which got indexed and sent a lot of traffic. And NDB was born...
- Twenty Questions (2006) An interview show. Built a small studio in my house and made two episodes before running out of steam. Still my dream job.
- Tldscan (2007). Made a bunch of sites that did deep crawls of structured information and put them on the Web, e.g. sites behind ip addresses, streets in the US, sports statisics, etc. Got a bunch of SEO traffic (50K uniques/day) and decent revenue ($500/day) and then Google blacklisted all of my sites. In retrospect, I made them too cookie-cutter even though I stand by that they were each individually useful. I should have spent much more time on the ones that were the most useful though.
- I've Got a Fang (2007). Essentially Squidoo or Mahalo -- I wanted to capture links in people's heads for particular subjects because I thought they were better than Google results. Launched but couldn't get people to upload enough stuff to where it made sense to continue.
- The "Wall" (2007). Facebook for introverts, focused more on group dynamics and life events. Never launched due to my partner on the project getting tangled up on other things.
- Kangadoo (2007-2008). Essentially posterous for grandparents. Parents would email or upload photos from their phones and it would be auto-sent to a homepage for the grandparents or other family members. Part of the idea was to sell it in stores as a software box, which we made and brought to the 2008 annual Toy Fair. Failed when partner walked out. I didn't want to continue it myself after that due to the heavy sales work.
- Groupomatic (2007-). Meetup.com, but more decentralized. I created it mainly out of a frustration with Meetup when moving to Philadelphia and trying to start some groups. I still use it, but only to maintain my own groups. Failed in customer development and iteration.
- nth Club (2007-). A golf social network for private club members. It's organized around leagues where people get paired up and play at each others' clubs. Worked in Philly but didn't spread beyond for various reasons.
- DuckDuckGo (2007-). A search engine. Directly grew out of I've Got a Fang & Tldscan and proximately from a weekend side project mashup of delicious and Wikipedia.
- Angel Investing (2009-). The jury is out as to how well I am doing.
- Traction Book (2009-). Started out as something like Founders at Work, but has evolved to a handbook on how-to get traction. Somewhat on back-burner because the evolution has caused it to be a much bigger project and I don't have the time.
Update: more great comments on HN.