The first graph shows a breakdown of total visits vs. new visitors. This is Google's best guess as to what percentage of traffic is new. I am averaging 660 visits per day, and Google estimates that 85 of them are new. Of course it's not as simple as that, but I am not going to delve into the mechanics of Google Analytics. Others do that far more effectively.
The second graph is Google's estimation of the absolute unique visitors to the blog over the past month. Again, how it determines this is not straightforward enough to delve into here, so the number I am showing may seem a bit puzzling. The answer to the question, "How many people visit the blog?" seems to be AT MOST 3500 in the last month, which is obviously a lot larger than the number of semifinalists. It is conceivable, though highly unlikely, that every 2011 semifinalist has visited this blog in the last month. The most plausible explanation is that some of you have cleared your browser cache or space out your visits to the blog such that Google's tracking mechanism no longer recognizes you as a returning visitor, instead counting you as a unique visitor. My guess is that there are some several hundred of you lurking, with at most a hundred actively commenting. I have no way of confirming this, however, since I've limited my ability to collect that sort of information.
OK OK I'm eating my 10 person words. however, I've logged in from at least 4 unique IPs.
ReplyDeleteGoogle Analytics tracks with cookies, though, so I would imagine that the number is determined more by how many devices you've used than what the IPs were (though it could have an effect too). My guess is that the average semifinalist will have had 3-6 counts by Google as unique, having been issued that many separate cookies. One or more for a university-connected computer, one for a home computer, and one or more for a mobile phone. Start adding in things like iPads and browser cache clearing, and I think we can safely bet that 5 is a reasonable estimate for a month's average. That would indicate about 700 actual users, most of whom are merely lurking, and I stand by my estimate of maybe 100 active commenters.
ReplyDeleteIf anyone wants, I can attempt a better breakdown based on things like browsers, operating systems, and the like.
What we should be asking is how many computers Hakeem uses to check this site.
ReplyDeleteYea, how many OPM computers have logged on?
ReplyDeleteIn the last month? 44.
ReplyDeleteCompare that to visits from U.S. State Department, which shows up 129 times in the last month. Surprisingly, I am showing a number of other government agencies as well. I guess that shouldn't be surprising, since including myself there are at least two other current/former PMFs that have posted here.
Er, I should say, that's the number of visits I've had over the last month from these sources. I don't know how many people that represents.
ReplyDeleteThank you for this!
ReplyDeleteReally interesting!! I am especially interested to know where everything is coming from... I like the agency breakdown, do you have a geographic breakdown??
I have the geographic breakdown as well, but I am not sure that's as meaningful as what we see in the nominees, semifinalists, and (when it's released) finalists lists. The reason for this is that not every visitor is actually interested in this program. Some people arrive here searching for other meanings for PMF (private military forces, for instance). As an example, in the last month I saw two visits from the third largest city in Russia, Novosibirsk, but I don't see any indication that the current group of semifinalists includes anyone who attended school there (the university names start with or include either Novosibirsk or Siberia). I am suggesting it might be more meaningful to map out the US and non-US locations of the nominees, semifinalists, and finalists based on the school names. I am cleaning up some data to attempt this as we speak.
ReplyDeleteThat will be great to see!
ReplyDelete