By Evan Levy
May 16, 2011
More Articles by Evan
Collecting data and making it work well for you can be extremely powerful. Here are some tools you could use to do just that.
Gathering data: First thing you need to do is find the data. Hunch.com/developers/v1 is building a search graph. You connect with your Facebook and Twitter account, and it starts asking you questions about yourself. Followalong is used to find people in your industry or someone specific you’re looking for. This tool helps you find specific interest sets in people. Mozenda- builds a scraper for you.
Statistics power a lot of the modern world. As humans, we’re good at looking at random raw data and picking out trends, but randomness often looks like a trend. There’s a book called how to lie with statistics that is a good read to make sure your data doesn’t become convoluted.
Don’t forget that a 95% confidence interval is wrong 1 in 20 times. Don’t blindly mine data for patterns. You’ll definitely find something, but have a hypothesis.
Visual = powerful. If you’re trying to tell a story, publish visual data. Try a tool called Geckoboard: build a culture of data and widgets using API’s.
Visualize website changes over time. This is a powerful way to see what changes were made to a site, and what happened at specific points in time that can be correlated with analytics trends.
Google docs and xpath is a powerful way to automate processes that you might want to run a few times, but don’t need a full blown tool for.
Analytics filters can be very powerful. You can use “regular expressions” to find very specific data quickly. If you wanted to look for numbers in site pages, use brackets [0-9] to say look for numbers zero through nine.
Django and Stack overflow are great tools to build scripts.
A program called “R” (www.r-project.org) can be great for plotting and simplifying noisy data sets. Show trends, identify outliers, display seasonality, all with one program.