Average page views per search, (search depth), is a metric value that is collected in Google Analytics search reports. Search depth includes the number of subpages that are viewed by a user after a search query. It is, therefore, a metric that provides information on the use of the internal search of a website and indicates the optimization potentials in the information architecture. Any term entered in the search field can be seen as of interest to the user. When users find what they are looking for, they reach their destination and therefore the user experience is satisfactory. If not, the data from the site's search reports can be used for optimization to provide users with the content and products that interest them.
Calculation
The search depth value is the average number of pages that are displayed after a search query. It is calculated as follows:[1].
Average pageviews per search = total pageviews after performing a search; is calculated as the sum of all search_depth values in all searches / (search_transitions + 1)
If a user enters three different search queries in three different sessions and then searches two results pages after each search query, the results pages are added first. The next step is to add the sessions on which a search query has been performed. In summary, the first sum of all the results pages is divided by the sum of the sessions, since search transitions are defined as individual searches in a session.
An example:
(2 [product] + 2 [product] + 2 [product]) / (1 [session 1] + 1 [session 2] + 1 [session 3]) = 2.0
Site search reports can, of course, refer to other collected metrics. For example, the number of all searches, sessions, or the number of page views. The lower the average page views per search, the more likely it is that the user found what they were looking for.
How does it work
If a user enters a definition in the internal search of a web, this entry is registered through the search tracking of the site, which must be configured in advance[2]. Google distinguishes certain dimensions in order to separate the phases of a search.
- Search page: It is the subpage in which a search is carried out.
- Search results page: This is the page that shows the results after a search.
- Call up the results page: If the user clicks on one of the results, this click is counted as a call to the results page.
The dimensions allow a more detailed view of the use of the search function. How long do users stay on the web after a search? What subpages do they visit and how many? The answers to these questions can be used to examine the effectiveness of a web search. However, for a more detailed analysis, the data from the site's search reports should be considered in combination. Only an overview of the different metrics and dimensions allows a substantive analysis of the findings to be able to formulate concrete recommendations for action.
Practical relevance
Google recommends five questions that simplify the interpretation of metrics in site search reports:[3]
- How often do users use my search box and what are they looking for? The use of site search includes searches made during a session. Keywords or keywords can be viewed under Behavior> Site searches> Search terms.
- Where do users start searching and what do they find? Subpages that a search was initiated on are displayed under Behavior> Site Searches> Pages. When a landing page is selected as the primary dimension, it is feasible to sort the results pages based on the frequency of the display. This means that the relevance of the results can be examined. The landing page must meet the needs of the user.
- Are users satisfied with what they find? The percentage of search exits in the site's search reports is an indicator of user satisfaction. This value is similar to the bounce rate for a visit to the site. In this context, the number of clickable results pages is also essential. If the web provides what you are looking for, this value will be low, which indicates a high level of satisfaction. If the value is greater than one or two, it is very likely that the user did not find what they were looking for.
- How do different user groups perform searches on my web portal? Using segments, for example, you can define different groups of users to better interpret the acquired data. Thus, it may be relevant to separate search queries by country or other aspects of the data. This is particularly useful for multilingual websites.
- What business results do the searches that users carry out on my web portal provide? The submenus in the site search reports provide a more detailed view of the data. Here you can see if and to what extent the internal search behavior of users has an effect on parameters such as conversions or ROI. For this, product categories must be chosen as dimensions, in order to later see KPIs such as objectives, e-commerce or sales.
Relevance for web analysis
Search depth and other KPIs in site search reports can be used to review the internal search usage of a web portal. This data allows deep insight into the interests of users and user groups, especially in the case of larger online stores and content-heavy publishing websites. On the one hand, the search terms indicate these interests and, on the other, the data is used to know the behavior of the user in terms of the user experience and the usability of the web portal. Successful and unsuccessful searches can be analyzed as well as content search terms.
The search terms entered are also used for optimization in a next step. Certain search terms that do not produce results can be used in content creation or in the layout of the information architecture (for example, the hierarchy of product categories). Search terms that are typed many times can be used as indicators of interesting products or content ideas to increase user satisfaction and conversions. When users reach their destination on the web page and find what they are looking for, the internal search can be considered effective and this naturally affects the sales of a web portal. Based on experience, up to 25% of users use the internal search function[4].
Web Links
