This article explains how Help Center Analytics may show different results when comparing these to the results from other analytical tools such as Google Analytics and Zendesk Explore. For example, when comparing the results for "Article View" in Help Center Analytics versus Zendesk Explore, this may show as find different counts for the same time range.
Different Results in other Analytical Tools
Why are the results in the Help Center Analytics different from Google Analytics and Zendesk Explore?
It is common to see certain discrepancies between different analytics tools. Zendesk have explained some of the differences you can expect between Google Analytics and Zendesk Explore in the Zendesk help article Understanding discrepancies between Guide and Google Analytics data in Explore.
Each tool uses its own logic for collecting and aggregating visit events. For example:
- Help Center Analytics counts the article views as the actual views of an article, not the visits to all the Help Center pages.
- Help Center Analytics aggregates all languages of the same article, both together and separately.
- Help Center Analytics uses the UTC time to report on visit events.
- Help Center Analytics reporting is calculated for each brand/help center, while in Explore (unless you apply a filter) the reporting is for all the brands in Zendesk.
As we have no information on how reporting in Google Analytics or Zendesk Explore is implemented, it is not possible to provide more details on any differences.
Article Views
How are article views calculated and what is the difference between article views and unique visitors?
A visit to an article counts as an article view in Help Center Analytics.
What is excluded from the overall metric are the views of other Help Center pages that are not articles (for example, the home page, category page, section page, new request page).
For the Unique Visitors count, visitors are tracked by generating a random unique identifier for each visitor. This identifier is stored in the browser cache of the visitor to the Help Center. The unique identifiers are then used to count the Unique Visitors.
The unique visits are counted in intervals of 1 hour and the visits are then aggregated (summed) for the dates selected in the Analytics filter. The total number would therefore be slightly higher than the actual number of unique visitors.
For example,
- If the visitor were to visit the Help Center multiple times within 1 hour, this would only be counted as 1 in the Unique Visitors.
- If you are looking at the Analytics for a period of one month, and a single user visits 3 times that month for less than 1 hour each time, the aggregate count for the Unique Visitors would be 3.
If a user visited 3 times in a single day, but the visits were 2 hours apart, then the Unique Visitors count would also be 3.
The unique identifier is cleared if the visitor clears their cookies or browser cache.
Translations
How are the metrics of all translations of the same article collated together?
Help Center Analytics aggregates all languages of the same article, both together and separately.
For example, if a visitor views an Article in English, this counts as one view. If they then visit a translation of that same article, that then counts as a separate unique view. The total view count would therefore be two visits.
The views are not unique, only the "unique visitors" are. So if the same person visits two translations of an article, it counts as two article views and one unique visitor.
Brands
Is there a way to see which Brands are included in the Swifteq analysis? For example, is it only a single brand, or can you select multiple brands at the same time?
It is only possible to analyze one brand at a time in Help Center Manager. You can not aggregate all brands into a single view.
You can switch between brands by clicking the Profile button in the top right menu, then click Help Centers and select Add Help Center. Refer to Connecting and Selecting Additional Help Centers.