An Iterative Approach to Search

Search is an imperfect science, which is why Blend helped a national children's hospital build and adjust a new search solution one piece at a time.

Search takes many forms, and we saw this when working with a children’s hospital to hone and perfect their own search functionality. Blend’s role was to work with design and usability recommendations from one group, pair it with the editorial needs of the hospital, and then provide a solution that made the most sense while still being mobile friendly, totally accessible, and provided accurate results.

First, we built and implemented a custom search solution, building in predictive text and, with help from Solr, a tweakable search model that ensures hospital patients and parents can find exactly what they want, right away.

Then, as data became more complicated and real-time results came in, we were able to adjust and iterate the search experience, both in terms of complexity and simplicity. Today, the search solution is much more spartan and usable, while still maintaining many of the connections we integrated early on in the process.

Project Results:

  • A sprint-focused, iterative process that allowed for testing and feedback without disrupting user flows.
  • Effective predictive text that has helped users "know their unknowns" when searching for medical information.
  • Mobile-friendly and accessible, meaning a search built for everyone.

Related Stories