First Results: Listening to Analytics
I posted <2008/11/01/making-analytics-sing/">earlier about my desire to listen to analytics data. So far I have barely scratched the surface of this, but just getting the system up and running is pretty inspiring. Now it's time to dig into the data processing and analysis. By the way, to get this to an audible frequency I have to interpolate some values, so every time I do any processing, I go back to the source, process, and then interpolate. Anyway. So far what we have is this:First, hourly data for a website stretching back to 2003. What I like about this is that it clearly shows some seasonalities, through repeating rhythms, in the data--this site gets big boosts around Thanksgiving and Christmas. Interesting, if you ask me.Second, I realized that this had a bad DC offset that could become problematic later on. To avert this I decided to take the derivative of every two samples. The resulting audio has very similar characteristics in terms of the overall rhythm (and it should), but the sound is "thinner".There's a bug in the normalization algorithm that's maintaining a small DC offset, but that can be fixed.The "noise" you hear in these files indicate days when the traffic spiked or tanked. You can probably hear that it gets noisier toward the end (not louder, but has more anomalies), which corresponds with this company's introduction of online marketing campaigns--especially email campaigns, which tend to drive a ton of traffic on launch days and not so much on other days.So again--this is just the foundation of what will come. With some luck, the monkeying around could lead to a new way to absorb analytics data. Who knows. Without a doubt, it will lead me to some things I haven't thought of before.