Scott Buresh, the co-founder and principal of Medium Blue Internet Marketing is a specialist when it comes to an in-depth analysis of a business site's statistics. He reports that many companies that he works with have access to traffic statistics, but they don't have anyone in charge of regularly checking the data and measuring changes.
"Typically the interest in this data is highest around the site launch and then trails off, which makes it impossible to gauge the effectiveness of ongoing site changes and traffic-building initiatives," says Buresh. "It is also very common for companies or site owners to look at the sheer number of hits or page views as the benchmark for success, not taking the time to understand the quality of the traffic or how well the site is addressing the interests of its visitors."
Given that scenario, what is needed to capture, and then analyse, basic traffic statistics?
"Usually, the data comes from the server logs, which come from the Web host, unless the company uses its own servers. There are also products, such as WebTrendsLive, that can track traffic remotely by placing a little piece of code on each page of the site. Again, almost every Web host offers access to this data for free or a modest fee, but it may take a special request. To analyse basic statistics requires no special knowledge, just a willingness to regularly dedicate a small amount of time to the site."
According to Buresh, it is not necessary to spend a large amount of money in analysing traffic statistics. "Although there are many experts who will charge large sums for in-depth traffic analysis, which breaks down everything on the site in painstaking detail, some of the most compelling data is easily understood by just about anyone. Once you understand the meaning of the most basic information on traffic reports, relating it to the site visitors is fairly intuitive."
An unfortunate aspect of online business is the myths that have grown up around the number of "hits" a site receives.
"Most Web surfers have come across sites that boast about '20,000 hits per day' or something similar, and most people believe that this means 20,000 visitors," says Buresh. "In reality, 'hits' actually refers to the number of requests for information the Web server receives. To use an oversimplified example, if a homepage has 20 separate graphics on it, each visitor to that page will account for 20 hits. If that site is boasting of 20,000 hits per day, it is really only talking about 1000 visitors. Obviously, this statistic is not a fair indication of actual site visitors."
Buresh points out that the true measures of activity are the average number of visitors over a given period, usually daily, weekly, or monthly. He adds that without access to this data and the ability to look at visitor history, it is impossible to tell if traffic building initiatives, whether online or offline, are working. It should be noted that the more traffic increases, the more accurate the rest of the data becomes, since trends in a larger sample are more telling than trends in a smaller sample where a small number of atypical users can skew the results.
Buresh says that another aspect of traffic analysis is determining if visitors are really connecting with a site. "To see how well a site is connecting with visitors, it's important to look at the average time people spend on the site and the average page views per visitor. If the average time that people spend on a site is small, for example less than a minute, or the average visitor only visits one or two pages, it may indicate some sort of problem. Perhaps the site is attracting the wrong traffic, with visitors abandoning it quickly when they realise it isn't what they were seeking. Perhaps visitors are confused by the navigation and decide to look elsewhere. Maybe the site just gives off an inexplicable bad vibe. Whatever the case, an awareness of the time people spend on the site and the number of pages they view can bring a potential problem to the surface."