BACKGROUND

One of the most crucial pieces of software utilized by Asurion is named Helix (Asurion’s knowledge base system). This is the platform our “Experts” use when customer’s call in for support regarding device issues. If the Expert did not know the answer to a customer’s question off-hand, they would use Helix to search our internal database for a solution. However, the search engine did not always return the most relevant content since it is keyword matched based. On average, agents spend 5+ minutes just looking for the right article. On average they click on 3 articles, and revise their search 3 times before they find relevant content. This issue was costly for the company and delivered a poor customer experience.

ROLE

Lead Product Strategist & UX Designer

User Research, Interaction, Visual Design, Prototyping & Testing

My Role

I was brought into the project to simply make some enhancements to various elements of the platform. However, one of the most critical flaws we as User Experience Designers’ have, is that we can never just think small. We view interfaces and experiences as one cohesive journey that is mapped out in the best way possible. Therefore, I wasn’t happy giving one enhancement to an overall flawed product, so I took the initiative and recreated the entire platform - start to finish. What started as a small project turned into the next four months of my life, and I couldn’t be happier with the results.

Previous state of Helix

Index Helix & community to return exact matches of words typed

Pros: 

• Super fast - real time results 

• Scalable - solution for all (Mobility, Soluto and PSDS)

Cons: 

• Little ability to influence or customize search results

• No performance improvements over time

• Relevancy is the hands of the searcher

Results

We introduced a new predictive auto-complete function similar to Google search.  Once the agent starts typing, we populate a dropdown list with predicted searches based on historical data.  This not only saves them time typing, but also suggests to them potentially better search phrases. We focused on customer value and significantly reduced wait time.

After we piloted this new capability to more than half of our Verizon agents across three sites, results showed a 42 second benefit on calls using Helix (about 30% of all calls).  Contributing to that savings is a reduction per helix call in the number of searches, the number of articles clicked, the time spent typing, and the frequency of tech-lead escalations.

Pilot Key Learnings

• Experts like using and say “its like google”

• 70% of Helix sessions use auto complete

• Major reduction in the level effort

• 26-138 seconds of reduction

  • Less time thinking

  • Less time spent typing

  • Less articles clicked

  • Less searches