Web Analytics – Statistics to Track
Not all statistics are worth tracking. The key to figuring which ones are is by answering the question of what will you do with the information? If a number is high, low, decreasing, increasing, or staying flat, what action will you take? If the answer is nothing, then there's no point in noting the statistic. Don't waste your time with things that are interesting, but not valuable. If you do, you will easily become overwhelmed with the maintenance of all your reports.
Asking these questions is particularly important if you're providing reports to clients. Clients are notorious for asking for data they don't act on. Although not malicious, they WILL waste your time. So push back a little and ask what they're going to do with the data.
Here are some of the calculations you'll want to consider making if you run your own website. Remember, not all of these will be worthwhile for all sites. Read my posting on Web Analytics – Definitions for the meanings of terms in this post.
Repeat Visitor Share
This a measure of how interesting your visitors found your content to be and how easy it was for them to find what they were looking for. The goal is to aim for a rising repeat visitor share which would indicate that more and more visitors are returning to your site over time. A decreasing number may indicate that your site is doing a poor job of convincing users that there is any need to return.
Calculation: Divide the number or repeat visitors by the number of unique visitors.
This is the percent of visits where the first page viewed was the only page viewed. In most cases, this means the users didn't like what they saw and so they left the site. I say in most cases, because the site may have in fact met the user's goals by providing everything be sought and hence there was no use in continuing. A high reject rate on top entry pages probably means trouble though as entry pages such as home page often don't have much in the way of content and are unlikely to have satisfied the user's needed.
Calculation: Divide the number of one page visits by the number visits.
Probably the most important statistic for e-commerce sites. This number indicates how many people performed the action you wanted them to perform out of all the visits to the site. Each industry will have a different rate that is good and bad. If you don't know yours, then focus on whether the rate is increasing or decreasing. When you make a change to your site, track the change in conversion rate.
Calculation: Divide the total “actions” e.g. order by the number of visits.
Cost Per Visits
A useful way to measure what each visit is costing you. Combine it with other numbers to determine whether you're paying too much for conversion.
Calculation: Divide marketing expenses by the number of visits.