Until today, Alexa has been the place to go to get a semi-authoritative and free gauge on traffic. But with the release of Google Trends, there is another must-see site. It is along the lines of BlogPulse, but much more authoritative, because it is tied into the most used search box in the world. It is however interesting to note that it has a similar problem as Alexa, in that all data is relative (not absolute). In other words, they’re not giving actual traffic numbers. They’re just comparing relative search volumes across times and between words. This is so similar to Alexa’s relative rankings, and I cannot help but feel is a general strategy that they use to keep themselves from being called to task on their numbers.
Whenever you’re looking for hard and fast numbers on Web traffic, based either on keywords or on site visitation, you have to turn to such companies as comScore and Hitwise, who each have dubious methods. comScore essentially runs software on enough individual’s PCs to get an average cross-sampling. Alexa too runs special software in the form of their toolbar. comScore has the superior method of cutting deals with ISPs to get a direct and pure sampling, but potentially skewed by region due to ISP service areas.
Alexa is the only one so far that has turned around and put its data out for popular general use. But their numbers can be challenged if put in absolute terms. The solution is to make it relative, and I am never done with explaining their Traffic Rank, Reach and Page Views. Of these numbers, Reach is the funniest. They call it reach “per million viewers”. They are essentially saying “parts per million”, measuring traffic that visits a site as one would measure ammonia levels in a fish tank. The trick is, you can’t tell how much ammonia is in the fish tank without being given the total volume of water. So, you can’t know how much traffic is reaching a site in absolute terms unless you’re given the total number of Internet users.
Similarly, while the Google Trends tool is amazingly useful in comparing keyword traffic on a word today versus last week, or between keywords A and B, it’s not telling you traffic on a keyword in absolute terms. This is still a process of piecing it together from the clues.