make a web site that makes money

I decided that I could use more information about the web site in order to make my big decision, specifically some factual information about target markets, how to reach them, and why-or whether-it’s really in their interest to throw down some cash to use the site.

So, I threw down some cash of my own: $500 on a marketing consultant for a small feasibility study. She told me what sort of people and what sort of companies would find genuine value in paying for the software. She addressed my fears about starting this business, by telling me specific things I can do to address them. She told me the negatives and risks clearly and specifically. Bottom line, I got exactly the information I needed to make my big decision, with enough clarity and specificity to trigger my imagination and weigh what I’m up against.

I just posted a wiki page explaining the concept of the web site:

http://decisionero.com/decisionero/Exploring_vs._arguing

With about 50 options and factors posted, the page with my decision about whether to work on the site instead of the Ph.D. was making more than 7,000 SQL queries. It was taking about 7 seconds to load—yikes!

It turns out that Django’s caching of query results didn’t work the way I thought. So I spent the last week restructuring the code, and did an all-nighter last night. Got it down to 120 SQL queries, and a passably fast site. Whew! (With a little more work, it could be way faster than that, even.)

As I was considering whether to cancel my Ph.D. admission this year and go all-out on the web site instead, it occurred to me that I don’t have enough money to last a whole year. There is some risk that I’d be under financial strain, and end up going to grad school next year with little or no money in the bank—a bad position.

And then, out of the blue, a friend I’d shown the prototype page to said he and his brother are eager to invest $15K.

I said I wouldn’t accept investor money for at least a few months, but having some cushion like that makes me a lot less worried about paying the rent.

Sometimes I’ve found myself worrying that this new web site won’t make any money. Better-positioned and better-funded competitors could come along and the site might never take off. Come to think of it, there are already competitors.

This kind of worry comes up especially when I consider delaying the Ph.D. for a year to do this site. If the site never makes enough money for me to pay the rent, then I will have wasted a year of my life.

And then I realized: I want this web site to exist, badly enough that I’ll put it up whether it makes money or not.

Paul Graham mentioned at Startup School that this is actually a great competitive advantage: if you don’t give a damn about whether the thing makes money, you are much tougher than people who are only trying to turn a buck.

I am crazed wiki zealot, and this project is simply an opportunity to let that zealotry run wild. Maybe it will become financially self-supporting, maybe not.

I’ve been working on web sites full-time for about six weeks now (since quitting my day job). For one site, my pair programmer and I hacked out a prototype in 6 hours one Friday, and we haven’t updated the live site since then. We’ve been coding every weekday. The idea has continued to evolve.

Progress is not as fast as I’d hoped. I was thinking more like two weeks to go live.

It’s great, though, to get immersed in a project of my own. It’s taken a while to get really ramped up, but now I’m waking up every day thinking about what to do next on the site.

My pair-programming friend and I just launched http://flashburrito.com. Hooray!

I tried for months to convince people where I work to pair-program with me. I’ve had only very small success. Usually people object that it’s not efficient, since two people are doing one job. I disagree, particularly on the types of job we are doing, where you spend a lot of time stuck because you lack information that someone else has. This is not the kind of thing you can meaningfully argue about, though. You have to do a little pair programming to understand how amazingly effective it is. I’m sick of trying to persuade people to try it.

So I placed an ad on craigslist for someone to pair-program with me on private projects. I got about six responses, and started pairing with one of them. It’s been going well. We’ve been meeting two or three times a week for three hours at a time, and we’re about half way through a small web site that could potentially generate a small income stream.

Two people contacted me out of the blue, wanting me to make web sites for them. I and the pair programmer are both interested in doing these gigs. This looks like it’s turning into a business. We haven’t gotten any revenue yet, though. We’ll see, but this has been fun and exciting. Plus, I’ve been experiencing the joy of pair programming—blasting out remarkably simple, solid program code, cutting through problems like a hot knife through butter even when it’s stuff I know very little about.

I just got my second check from Google, again for about $100. So http://greenlightwiki.com is making a little less than $100/year.

I would say that this goal is not yet complete.

I just got a check from Google for for a web site that I run right now: the Green Light Wiki. That site was never intended to make money, though, and it took 20 months to rack up enough money to even be worthy of a check.

A web site that made, say, $500 per month would be an incredible boon to a college student, as long as it didn’t require much work to maintain. That would almost pay my rent.

I’ve got a few ideas, and some of them require as little as a few days to set up. Given the demands of school, doing one of the ideas will still require some serious discipline with time.