Note added in 2021. When I wrote this 10 years ago I had no idea that what I then saw as a fascinating side project would become my company, and sustain my family for years. I was wondering, back then, who would want these wonderful maps I was making. The answer ended up surprisig me beyond anything I could have imagined.

23 visits a day

Star map example

I’ve been obsessed for the last 24 hours, ever since I put on-line the version of that I thought was the first one to deserve promotion. I knew that I shouldn’t really care; I knew that, most likely, nothing would happen; I knew that I should expect to be ignored in Hacker News. And I’ve done it because it was fun. I’ve had a good time figuring out how to build the puzzle of Common Lisp, Python and Javascript that serves the star chart PDFs. And hosting is free. But still. It’s hard not to give in to what I would naturally do at this stage: add features at a frantic pace, assuming that the next thing is the one that will make the difference.

And it is not.

Ancient history

It started about 5 years ago. Working as I do as an R&D engineer at the division that builds large format printers at HP, I thought it’d be nice to take advantage of the huge and good printers around me to print a really good star chart. Good as in thousands of stars, with colors, in their right position, as seen from a given vantage point at a given time. And big, really big.

It is the curse of the engineer: if it can be done, it will be done. I learnt how to compute the position of stars and planets, I found out about the HYG database, I learnt enough PostScript to get by, and I put together a Common Lisp program that produced a nice ps file with thousands of stars and all the planets, with constellations and grids (it helped when they demoted Pluto from its planet status, as it was the hardest to compute for). I made the program so that it would draw pictures of the planets instead of their symbols, a feature loved by my non-astronomer friends.

It was an interesting project, and successful in a way: I made several charts as birthday presents, and gave a specially big one to the Fabra observatory here in Barcelona, where it went on display. But when I tried to run it in the server where my web sites are hosted it brought it down. I guess it should have been possible to optimize the program, but I never did it, so it became the start of an off-line boutique shop that was mostly found by word-of-mouth.

Back to the present: Amazon Web Services

Until a month ago. A friend of mine was looking for a birthday present. I offered her a personalized star chart and she thought it was a great idea, and this time I found myself hacking on my program again.

It may be because taste evolves, or because your expectations change when you grow older: but I ended up spending quite some time fine-tuning the output to make my maps prettier.

And then I thought that, maybe, servers are now powerful enough to run my program without a sweat. And I wanted to try out AWS, so it was a great excuse. Sure enough, the free-for-a-year version of EC2 can run my program and produce a PDF file in about 3 seconds, as fast as my iMac. Amazing. So I decided to give it a try.

I already owned the name, and was using it to host articles on math and science for kids. I decided to re-purpose it and build a site where you could buy a personalized PDF file of the sky (so why, again, would you empty a cauldron full of perfectly good onion soup to put sestertii on it? Obelix would ask).

Building the product

So I’ve spent all my free time during the last couple of weeks building Javascript in the browser, Python and Common Lisp in the server. Also trying to produce a design that was not too ugly. After several iterations, I got to something that my non-esthetically-challenged friends thought was almost acceptable, so I thought it’d be good enough. And yesterday, after a weekend spent adding the last features I thought were key, I shipped. As in I actually told other people it existed, I posted it to Hacker News, and published it in my G+ and Twitter.

And nothing happened.

As it should be. Because it’s possible that nobody wants what greaterskies offers today. And most of the features that I’ve been adding for the last week are probably irrelevant. And if somebody cares about it is not the Hacker News crowd and, most interesting, I don’t know who that would be. Mothers who want to present their teenagers with science-sounding decoration? Who knows.

A minimum viable product

Note to self: you need to realize that you don’t really know what people want, and set out from the start to experiment and figure it out. The goal is not to build a product, not even to sell it: it is to find out what is it that you can offer that people actually want. Something that’s not too far away from what you are set up to do, and that people will actually buy.

The most difficult thing in this business is deciding what constitutes the minimum product that will help you learn what you can offer, what they call the minimum viable product (and you can maybe see at this point that I’ve been reading The Lean Startup, which I would whole-heartedly endorse if it was not a compendium of management-speak and unnecessary made-up words and acronyms.)

But here’s a rule of thumb for next time: list all the features that you think are essential, then cut by half. And ship it.

The future

So I’ve worked too much. A trivial amount, I know, but still too much. And I’ve released too large a product. Now it’s probably time to do exactly the opposite of what Steve Jobs says should be done, and try to find somebody who wants it. Maybe astrologers, or university professors, or wedding photographers, or soccer moms.

And, more interesting, start experimenting. I’ve set up Google Analytics to measure the proportion of visitors that go from the first page to the second, where they should input the data required to prepare the chart; I also know how many fiddled with the system and didn’t buy. These are the metrics I will focus on, and I plan to report them weekly. Stay tuned.

Juan Reyero Barcelona, 2011-10-03


blog comments powered by Disqus