How to dig data for top search keywords from site search

How to dig data for top search keywords from site search

Top Searches Report is the simplest report provided by Unbxd dashboard to Category and Product Managers who need to optimize website performance. In this post, I will be talking about how to collect this data and how to use it. In the the next post in this series, I will talk about the insights product managers can use to understand their visitors and optimize conversions. The value of this post is in the fact that you can use the points in here even if you do not use Unbxd dashboard. Collecting Search Queries For the purpose of reporting, there are two sources of search queries:
  1. Keywords that are embedded inside SEM campaign URLs. These keywords ultimately arrive at your e-commerce site where they are treated as inputs for the site search box.
  2. Keywords that human visitors enter into the search box manually.
The difference is that the former keywords are entered into the SEM campaign by a merchandiser who may or may not be aware of the actual language used by visitors searching on the site through the search box. True visitor intent can only be gauged from the latter and not the former. In order to get accurate data on keywords used by visitors, you must differentiate between these two sources of search keywords. Counting Search Queries Once you’ve collected the search queries from the correct source, we need to determine how we want to count each query. Let me give a few examples of what this means:
  • Would you like to consider “kenstar iron” same query as “iron kenstar”?
  • How about “ipad case” and “cases for ipads”?
  • Is “ipad3” same as “ipad 3”?
This is important because in the table of top queries, treating such queries either same or different can make a huge difference to the numbers for each query. Depending on the purpose of the report, be it SEM campaign or just product sourcing, it can make a huge difference to the success of the e-commerce operation. I will cover this difference in more detail in the next post, where I will talk about SEM campaigns and product sourcing. Ultimately, we also need to ensure that site search present the same results for “ipad3” and “ipad 3”-like queries. In the next post, I will talk about all the insights that can be derived from a top search keywords report.

More To Explore

Learning To Rank with eCommerce data

Learning to rank with eCommerce data

What, Why and How of Learning to Rank: Learning to Rank (LETOR) is used in the information retrieval (IR) class of problems, as ranking related