Products
Webtrends Analytics 9.x
Webtrends Analytics 8.x
Issue
All visitor measures within Webtrends Analytics report on unique visitors. However, there are a few nuances that you may need to understand when viewing select reports or creating your own custom reports.
Resolution
The following describes some best practices for using and understanding reports with visitor measures.
Dimensions that allow visitors to be in multiple members (rows in the report)
Example reports: Pages of Interest, Key Metric Summary
A visitor can visit multiple pages of interest. Consequently summing the visitor measure in the pages of interest report can show an inflated number of visitors.
Visitor looks at Page A = 1 visitor
Same visitor looks at Page B = 1 visitor
The visitor numbers are accurate if looking at Page A by itself (1 visitor) or Page B by itself. Totaling this report would show 2 visitors though when we only had one. The recommended approach is to disable the row total for visitor measures in reports where a visitor can participate in activity that places them in multiple members of a dimension. While Webtrends has already done this with packaged reports, but this is a configurable setting in your custom reports.
Another example where this behavior may be present is in drill-down reports. Drill-down reports achieve their totals by summing all levels below. If a visitor is in two different sections within a drill-down report, the total visitor numbers will be inflated when summed. When needed, you can create a one or two dimensional report to accurately roll visitors up to a level within the drill-down report.
2-D reports are not affected by this behavior since the top-level (1D) takes into account when visitors visit different 2D sections and will not inflate the total numbers.
Custom date ranges
Webtrends Analytics reports on unique visitors for the following standard time periods: Daily, Monthly, Weekly, Quarterly, and Yearly. When possible we recommend using these periods for reporting on visitors.
Webtrends Analytics can report on visitors for other time periods by utilizing the custom date range feature. Since all visitor counts are aggregated into these time buckets, Webtrends Analytics applies an approach that sums the count of visitors from the largest and next largest complete time buckets within the selected period. This ensures a much higher degree of accuracy, especially over larger periods of time, than summing daily visitors, but is still less accurate than selecting standard time periods.
Take for example the following three scenarios:
- When looking at a report for a given day, all of the information for that day has been aggregated. This means, if a visitor comes to the site, leaves, and comes back again, we have only counted them as a single visitor.
- If we look at a weekly report, and if the example visitor has been to the site on Monday, Tuesday, and Wednesday this report will have aggregated that data so that they appear only as a single visitor.
- If the visitor returns three days during the following week, their count would be distinct across both weeks when looking at the entire month. In the case that the next week crossed into a new month, we would need to look at a custom date range since Webtrends doesn’t have a standard two-week time period. Instead of summing daily visitors to get a count of six, the largest full periods (week 1 and week 2) are summed to get a total of two.
Event based data (page views and hits) are not affected by custom date ranges since they are merely a count of requests to pages.
Non-Additive Metric
The sum of the Unique Visitor totals for the days of a month will not equal the Unique Visitor total for that month. This is because if the same visitor came back 3 times within the month, they will appear in the daily totals 3 times, but in the monthly total only once.
The unique identifier for a visitor changes
Webtrends Analytics reports on visitors and visits based on the session tracking definition set for the profile. By default, Webtrends Analytics reports on visitors using first party cookies. If a visitor removes his cookies from his machine, uses a separate machine, or uses a different browser he will appear as a new visitor as he will receive a new cookie identifier.