Leveraging site search analytics

Customer intent is a powerful and under-utilised analytics tool

Trent Billington
Search.io
2 min readMar 5, 2018

--

Analytics is central to making good business decisions. Unlike regular analytics, search analytics are very powerful as they capture the exact state of user intent. Accordingly at Sajari we collect extensive data around site search, from search and result interactions through to advanced content scoring, anomalies and more. These reports allow our customers to improve their content and in turn better serve customer intent.

Valuable insights

Search insights can be pretty amazing. We have customers tracking random site search metrics such as:

  • “cancel”, “disconnect” and other negative queries to understand sentiment
  • “outage” and “error codes” style queries to track physical outage locations by the search location
  • “change address” queries to track self service ease

But these basic metrics are just the beginning. Content based sites look even deeper into things such as reinforcement learning confidence scores for queries and results. These help businesses track the performance of every result ever seen for a given query to help spot content performing abnormally and to investigate the causes behind queries with poor user interaction. Looking at this data from the URL side shows how it’s performing across all queries, enabling poorly written content to be re-optimised.

Exporting

Maintaining and utilising multiple analytics systems is a pain. Accordingly for a long time we’ve had requests to export our analytics data so it can be ingested elsewhere.

For a long time we’ve supported Data Studio and BigQuery, but we’re now allowing all analytics data to be exported from our admin console in CSV format. Exportable analytics includes:

  • Trending queries & their click-through rate (CTR)
  • Top queries by volume
  • Learning & Result confidence levels
  • Content issues including trending queries and ineffective queries
  • Usage by API endpoint

Exporting to CSV manually is the first step in enabling automatic exporting of data via our API to drive external analytics processes and show search analytics inside CMS’ and more. If you want to take your site search to the next level, check us out today!

Sajari is a search and matching technology built from the ground up to incorporate analytical feedback and machine learning at a fundamental level. We provide web search and custom matching solutions for all shapes and sizes.

--

--