Practical Analytics: Understanding Visitor Summary Data

The Visitor Summary table, which appears in the Visitor Dashboard or the Overview Dashboard, is one of the most used features of WebTrends Analytics products. But it is also one that can lead to confusion, especially when your data users want to compare the numbers shown in this table with other “visitors” numbers shown elsewhere in their WebTrends reports. In particular, when they see that the numbers shown in this table do not match the ones shown in the New vs. Returning Visitors report, they inevitably ask, “Which number should I use? And why are they different?”

Visitors Who Visited More Than Once

The Visitor Summary table shows Unique Visitors numbers for all the web pages included in the profile based on the sessionization method in use. The most commonly used sessionization method  is a First Party Cookie (FPC), because it gives you the most accurate data short of requiring all your visitors to log in. If you are using a FPC, but the particular visitor’s browser rejects the cookie, then WebTrends will resort to using IP-User Agent for sessionization of that particular visit, and the visitor count will reflect that. If you are using IP-User Agent for sessionization by default, the Visitors number will be based on that, instead. But all of those visitor numbers, even those based on IP-User Agent instead of cookies, are included in the Visitor Summary table.

Visitors who Visited More than Once from the Visitor Summary table represents Repeat Visitors during the time period selected. The WAA definition for Repeat Visitors is:

The number of Unique Visitors with activity consisting of two or more Visits to a site during a reporting period.

The numbers shown beside Visitors who Visited More than Once is dependent on the time period selected. So, if you are looking at a month’s worth of data and you had 68,962 visitors who Visited More than Once, that means 68,962 of your unique visitors that month came back more than once during that month.

I consider this metric one of the basic, no-frills, engagement-oriented metrics WebTrends can provide out of the box. I like to use this to calculate a metric called Repeat Visitor Share, which is simply Visitors Who Visited More than Once divided by Total Visitors, and then I trend that number over time.

It’s worth noting here that the numbers shown for Visitors who Visited More than Once should equal the total found in the Visitors by Number of Visits report (found in the Marketing chapter under Visitors) LESS the visitors who had only one visit.

Another thing to know is that Visitors totals shown in the Overview Dashboard are counted at the end of the visit, but Visits shown in the Overview Dashboard are counted when the visit starts (what we call an open visit). Counting Visits at the start of the session ensures that WebTrends captures as much data as possible, while counting Visitors at the end of the session  gives WebTrends time to correctly attribute that visitor as a Returning Visitor or a New Visitor and to capture other types of information about the visitor correctly. We’ll talk about one way this difference in when Visits and Visitors are counted can impact your reports further down.

New vs. Returning Visitors

The New vs.  Returning Visitors report, found in the Visitors folder of the Marketing chapter, is telling you something quite different from the data in the Visitor Summary table.

The first thing to keep in mind is that this report is reporting visits, not visitors, as is shown in the Visitor Summary table. The New vs. Returning Visitors report is reporting the number of Visits by New Visitors, the number of Visits by Returning Visitors, and the number of Visits by visitors who couldn’t be put into either of those categories because they were not accepting cookies. It shows you the percentage of your traffic by members of these three categories. I like to trend Returning Visitors over time and use that as one of my basic, retention-based benchmarks.

The second thing to note is that WebTrends decides which category to put a visit in based on whether or not the visitor had the sessionization cookie (usually, the WebTrends FPC, but it could be some other cookie that you are using for sessionization). Therefore, this report will not display meaningful data if you are using a method of sessionization weaker than cookies, such as IP-User Agent. In that case, all Visits will be categorized as Visits by  New Visitors.

However, assuming you have a strong method of identifying your visitors, Visits by New Visitors are those where the cookie was not present. The total visits shown for Returning Visitors are those by people who did have a  site cookie. Because this report is visit-based, rather than visitor-based, the visits of a user who visits the site for the first time Monday morning and then returns to the site Monday afternoon would show up in two places in this report. The first visit would show up as a visit by a New Visitor, and the second visit would show up as a visit by a Returning Visitor.

From this report we don’t know how long ago our Returning Visitors visited from this report, nor how often, simply that they had visited us at some point in the past. If a user visited six months ago but this month they visit only once, their visit will show up in this month’s reports as a Returning Visitor visit, because they had your cookie from their previous visit however long ago. But since they only visited once during this month, they would NOT show up as a Visitor who Visited more than Once, because they only came to the site one time.

The total number of Visits shown in the New vs. Returning Visitors report should be close to the total number of Visits reported in your Visits Summary table in the Overview Dashboard. But it may not be exactly the same. Why is that? Remember when I said that the Visits shown in the Overview Dashboard are counted at the start of the visit? That is true of the Overview Dashboard, but not for the New vs. Returning Visitors report. This report and some custom reports counts Visits at the end of the visit, what we call a closed visit. For this reason, the numbers may be slightly different, depending on how many open visits there are compared to the number of closed visits.

Some newer versions of this report also include a  “Daily Visitors” column in this report — which is a custom measure based on our Visitor History table. Visitor History relies on the use of cookies, but the Visitors total shown in the Overview Dashboard does not — it will count a visitor even if that visitor is not accepting cookies. For this reason, the number of Visitors Overview Dashboard will usually be slightly higher than that shown in other reports.

As with anything, different ways of measuring things will almost always result in different numbers, even when those differences are necessary. When you look at your WebTrends numbers, always keep the context of the data in mind. And now you know a bit more of that context.