Analyzing One Million Robots.txt Files

Where is the insight? You have a bunch of clusters, but what do you actually do with the clusters once you have them?