Here I am, once again, in a familiar situation. This time, it’s with RubyPilot.com, a Ruby resources aggregator and my latest side project. As I approach the finish line, questions begin to surface. How will I drive traffic to the site? Why would someone choose it over browsing Reddit? Perhaps I should narrow the scope and create a directory of apps built with Rails. But then, who would be interested in that, and how will I monetize it?
I reassure myself that, lacking a better idea, I decided to build something I would personally use, hoping that inspiration for further development would strike along the way. This is partly true. I did stumble upon a new idea during this process, and I enjoy refreshing the site to see if anything new has appeared. But the truth is, I work on these projects with the aim of making money. Somehow, I always end up with hobby projects, clueless about how to drive traffic and earn money from them.
The only time I felt differently was during my programmatic SEO experiment. I gained access to a government API that provides information about new streets in Poland. After some preliminary research using free tools, I discovered that people are searching for “new streets in [city name]” (in Polish). Doing some quick math, with approximately 100,000 towns in Poland and assuming I could rank for 5% of them with a traffic range of 0-10 visits per month, this could potentially generate 0-5,000 visits per month. If so, I could monetize the site with AdSense ads and it would be profitable.
This led to two experiments. The first was to see if I could actually attract a few thousand visits per month with programmatic SEO for the “new streets in [city name]” keyword. In reality, I managed to attract 1,500 visits per month. The second experiment aimed to determine whether this traffic was sufficient to make the website profitable through AdSense. Unfortunately, it was not.
I learned a lot and quickly validated my idea, allowing me to move on. Even though it didn’t work out, I was relieved to have a definitive answer. I believe this is how it should work: come up with ideas, prepare experiments to test the main hypotheses, execute, learn, and either build upon the idea or move on to the next.
No more clueless hacking; let’s experiment. I’ll try to apply this approach, and this blog will serve as my laboratory notebook.