Before Elastic Site Search, search was not a prominent tool within the Twilio online knowledge base — primarily because their previous solution did not work well. Aside from the fact that the search bar returned results that were not relevant, any improvements the team wanted to make to their search algorithm required heavy amounts of back-end development work. As a result, the knowledge base was an underused asset and left their new users without a solid place to easily get started. As a result, Twilio needed a solution that would provide powerful search without a lot of work.
Twilio had their new Elastic Site Search (formerly Swiftype Site Search) engine up and running in minutes.
With Elastic Site Search service, Twilio easily added search boosts, synonyms, and other adjustments to return the exact results that users were looking for.
Twilio exposes a globally available cloud API that developers can interact with to build intelligent and complex communications systems. Thousands of developers use Twilio on a daily basis, and Twilio has developed a comprehensive knowledge base to provide answers for users who are getting started.
Once Twilio made the switch to Elastic Site Search, they had their new search engine up and running in minutes. Now that search worked well for the knowledge base, Twilio decided to feature it prominently on the page. Within the company, people were immediately excited by the autocomplete drop down results. The change was easy on Twilio’s end and employees could quickly see how much better this made their user experience.
After getting their search engine up and running, Twilio’s support team used their built-in Elastic Site Search dashboard to customize the results for the most popular queries with the goal of creating a better end-user experience. With the drag-and-drop result ranking tool, the Support team was able to rearrange results for the top 50 queries to make sure the best results were easy to find. Perhaps more importantly, they were finally able to add in results for queries that were returning null values in order to minimize the need for users to file support tickets for answers that were easily covered in the documentation.
These dashboard analytics not only helped Twilio support customize existing search results, but also helped them determine what new support articles needed to be created going forward. For instance, prior to deploying Elastic Site Search, the Twilio support team had a sense that there was inadequate documentation surrounding session initiation protocols (SIP), but never had time to develop articles for users who had SIP questions. However, when looking at the analytics dashboard, they could instantly see how many users were searching for help on SIP. With this data in their hands, they were able to easily justify the time spent on creating a new searchable SIP FAQ section and in turn now spends less time answering support tickets on SIP because users can quickly find the information they need through the knowledge base. With the Elastic Site Search Service, Twilio was able to provide its users with powerful search for self-service support, all without the need to involve engineering or long-term consultants..