How often is searched and for what ( search terms )?
Where does the search begin (page) and end ( exit/conversion )?
Do searchers find what they expect ( search result bounce rate/refinement )?
Is there a distinction between the searchers ( technology , e.g. smart device vs desktop)?
What business results come from the searches ( ecommerce transaction , but also other micro-goals such as newsletter registrations )?
Analysis of search behavior can lead to conclusions and recommendations for:
new topics for content pages on the website (with SEO-friendly texts);
topics of discussion for the purchasing department regarding additions/changes to the product range;
adjustment of navigation, information and filtering elements on the website;
adjustments in the timing of topics of the annual online marketing agenda, also for channels such as e-mail marketing and SEA;
etc.
The method
For our client we have made the on-site search behavior over the period May 2011 to April 2012 (12 months) insightful by collecting the search terms and arranging them per month in a top 50 ranking based on use and on conversions . The following image illustrates a top list, I have replaced the search terms and amounts with fictitious terms:
Google Analytics Site Search - Sample Composite Internal Ranking List
A pattern quickly became apparent in the search terms, after which the individual terms were categorized into 6 groups, namely search terms with the following theme:
place names;
accommodation requirements (in relation to available facilities in the accommodation/room);
environmental requirements (in relation to time spent in the area around the destination);
accommodation names;
destination requirements (in relation to facilities provided by the accommodation provider);
otherwise.
After all search terms have been categorized, they can be combined into a keyword cluster in which the mutual relationships to the theme become clear. This is the moment when the labor of collecting data slowly starts to pay off in insights!
A practical example
For example, look at the cluster 'place names'. In this overview, the search terms are divided per month by theme and the position on the top 50 search list based on "number of times used". By determining an average and following the position with conditional formatting in Excel, the first patterns begin to unfold visually.
When this data is processed into a line graph, the seasonal influences in this segment of the travel industry quickly become apparent.
Google Analytics Onsite Search - Average ranking per keyword in cluster
Because we are convinced that steering on conversion data improves the return on a website, the turnover data resulting from the search term is also included in the analysis:
Google Analytics Onsite Search - Include Business Goals in the Equation
Based on these three overviews, the following conclusions and action points can be drawn, for example:
London: is searched onsite the entire period and contributes substantially to turnover. There is clearly sufficient demand throughout the year, but searches are made from various pages and the uruguay phone data required information is therefore not found in the regular navigation.
>> Action: How does the current coverage of this location compare to the supply? Could we perhaps highlight 'London' more clearly in months with low coverage with a better call-to-action?
Paris: searched in certain seasons, but converts relatively little. Relatively many searchers, but clear change in season. High bounce on results pages, many exits.
>> Action: What does the current results page for this search term look like? Are there reasons to find that this is not converting in the current range, availability and price proposition.
- Board index
- All times are UTC
- Delete cookies
- Contact us