Poynter definitely knows its stuff. I am new to the whole idea of web analytics and panel methodology to measure access to websites, but this article, “What web analytic can – and can’t – tell you about your site’s traffic and audience,” by Dorian Benkoil, gives the straight facts and shows how the two methods are similar and different.
Benkoil explains that some factors may affect the hit count and drastically over- or under-estimate the number that sites want so desperately to be able to share with advertisers. Cookies, codes that websites put on a browser to count users, are one of the main causes. According to a study that he cited, cookies can inflate the hit count by up to 2.5 times what it actually is. That could be a huge difference to advertisers and sponsors.
The issue caused by cookies is that they count the times a web browser system hits a website. They are not able to count physical people, and they cannot account for people who use more than one computer. With the existence of iPhones and iPads, many people can browse the web while they are on a train, eating lunch or waiting at the dentist (a much better alternative to reading old golf magazines).
My brother is an IT junkie, so over the years I have picked up a few things from him. Thank goodness he taught me how to set up an internet router, or I would be completely lost. The point is that he has often told me to clear my cache or cookies, and I blindly did it, not completely understanding it, but this time, Benkoil gave a good description of what cookies actually do.
The idea of panel methodology also seems like an interesting concept. Researchers study participants’ internet usage, and the data is based on real people, not arbitrary numbers. I’m sure that web analytics have their own unique purpose too, but it seems that panels could work really well for smaller regions.
The main question that comes to mind, since this is a class based on community journalism, is this: What is the best way to measure websites in smaller communities? The article mentions that analytics are not necessarily reliable in areas where many people use a computer, like public libraries, but it also says that smaller sites are not as easy to measure in a panel study, since the sample size decreases. Some companies will not even consider measuring a site until it gets a certain number of hits per month.
Even though they aren’t chocolate, warm-from-the-oven cookies, cookie codes seem like the easiest and fastest way to access information about web hits. We could argue that the over- and under-estimates probably balance out in most cases, so until we find a better way, cookies seem to be the best option.