Hello all. I am currently doing a small pilot survey of satanism sites online for an article in order to get more data on ephemeral networks of seekers. One of the things I suggest is to use "links to" as an indicator, and I have happily used alexa, marketleap and site analytics to get some raw data. But I am at a loss to explain the large variability in the results. For example, alexa lists about 600 links to Church of Satan, while marketleap has 25000. It is especially visible with popular sites: YouTube has 700000 links with alexa, while marketleap list tens of millions. How do I explain this? Are the search samples much smaller in alexa? Or is the marketleap number including duplicates and such? Or is it related to search criteria and the definition of "link to"? Any suggestions will be much appreciated. All the best, Jesper Petersen, NTNU, Norway.