When working with hundreds of thousands of customers, with millions of messages, a search function is a no-brainer. Search at this scale serving businesses needs to be fast, reliable, and smart. As the product and business scaled, search became weaker and weaker, and was in great need of a revamp.
Historically, Sparkcentral's search had been been plagued with classic search problems: not returning results that absolutely do exist, not surfacing the most relevant results, and the search queue interface was squeezed into the original queue space, making it hard to distinguish. The need for a better search grew by the day, so our team set out build and excellent search experience.
As Sparkcentral continued to build out the new platform, search was recognized as an essential feature in the legitimacy of this new platform. Rather than carry over the lackluster search feature from the old platform, I wanted rethink search to be something to be proud of. We wanted users to be confident in, and able to rely on, the in-app search.
At Sparkcentral, I was grateful to have incredible product managers and engineers to work with, who were usually within shoulder tapping range. The moment product managers began ideating on a new feature or product, I was invited to the conversation. Over my years with the company, I'd built up quite the backlog of wireframed concepts and notes from conversations, all of which helped guide our conversations.
The engineers respected design and made an effort to keep up on the latest design and tech. Together we'd build out a style guide/component library which greatly expedited the development process.
Customer Success was also a crucial collaborator. Not only was our product meant for people in roles like customer success, our CS team had a finger on the pulse of our customers and knew what they wanted out of the product.
Before digging into the details of a new search experience, the PM on my team and I sent out a survey to five of our heaviest search users. It included simple questions such as, “What do you use queue search for? Please be specific about which queue you search in and what you are looking to accomplish through searching,” and “What is most frustrating about queue search?” Through these surveys, we gleaned valuable information. Some of it validated our assumptions, and some aligned perfectly with the usage data we had captured. The common theme though was that users didn’t like search. They didn’t use it often, because when they did, they experienced bugs and found the feature to be unreliable.
I created an isolated experience around search. In the old Sparkcentral platform, when a user performed a search, the results would overlay the existing work queue. This created a few problems: a queue of search results hardly looked any different from the regular work queue, users could be working a case while search results were overlayed on their work queue, and the filters and sort options that were available to the work queue *appeared* to be available to the search results, but they in fact did not affect the search results at all.
Creating an isolated search experience with a unique visual style intuitively informs the user that they are not in their typical work queue. In their typical work queue, every second matters, as there are several time-based metrics being recorded against them as they work to resolve customer inquiries. The user is free to explore and find what they’re looking for.
This new context also removes any of the tools that were only used for the work queue, such as those filter and sort options, and provides new tools that only apply to search. Users now have advanced filtering options that allow them to filter by: channel that the customer reached out to, date range, current case status (New, Pending, or Resolved), and incoming and outgoing messages.
A common workflow that our users perform in regards to search is to reply to or tag conversations in search results. In order to do this, we provided all of the tools necessary to work a conversation within the search experience.
In the weeks following the release of the revamped search experience, we saw an increase in usage of search and a decrease in support requests regarding losing messages and customers. We found success in optimizing our customer's operations and decreasing the workload on our own customer success team.
If you'd like to know more about the process, research, and results of the project, please reach out and I'd be happy to share.