Univé deploys Elastic Cloud to transform website search experience for customers

In the past, site search at Univé was powered by an on-premise search database, which stored web pages, PDF files, and other documents. But adding database content was a time consuming and manual process. As a result, new content was sometimes delayed or even lost. Maintenance was also a drain on resources. Every month in-house developers spent up to three days adding new features or resolving technical issues. As the volume of Univé’s digital content expanded, these challenges also grew and the IT department began to investigate alternatives to the on-premise solution.

“We had several criteria for a new search platform, but most of all we wanted to transfer responsibility for installation, development, and maintenance to a partner organization, reduce burdens on our in-house team, and lower costs compared to an on-premise solution,” said van Mier.

Univé sought the advice of its long-time IT partner, Atos, which recommended Elastic Enterprise Search deployed as an Elastic Cloud SaaS solution. Atos also advised using Elastic’s built-in dashboard tool for data analytics and visualization.

As a result of deploying Elastic Cloud, Univé saved on average 20 days of developer time in a period of 2-3 months compared with developing and deploying the solution on-premise and reduced day-to-day maintenance time by 95%.

Univé uses Elastic Enterprise Search to track and analyze website searches and then applies the data to continuously optimize search performance. For example, end users often enter search terms that are relevant to their needs but aren’t recognized by the search database. Univé can now track these ‘no result’ searches and add unrecognized search terms to its synonym library. Over time, this increases the accuracy and relevance of results and enhances the overall user experience.

The Elastic Enterprise Search web crawler also helps improve search results by indexing documents and adding page titles, descriptions, keywords, and URLs to the search database.

Univé takes advantage of Elastic’s search API to fine-tune search results and alter the weight attributed to specific search fields. Known as relevance tuning, this helps return results that most accurately match a user’s intended search.

Van Mier emphasizes the benefits of working with Atos as its technology partner, particularly due to their Elastic Cloud expertise.

Ultimately, improved search means more efficient internal operations for Univé and a better website experience for customers.

In the first six months after the Elastic Enterprise search deployment, Univé saw an 87% reduction in ‘no results’ searches. Customers were also more engaged. Within the first month, the top five search results bounce rate almost halved. In the future, Univé expects to enhance search personalization by implementing custom tagging and adding the Elastic autocomplete feature. Van Mier and his team are also looking forward to the planned PDF document indexing functionality.