Call Center Application
Call Center Application
Customer Service Agents suffer cognitive overload from
context switching among multiple software systems to
service a customer - creating a workflow bottleneck and
making calls take too long. The main software they have relied on for over 20 years is difficult to use, outdated and expensive to maintain.
United Airlines
2023
Tools used
Figma, Sketch
Project type
Legacy tool transformation; single page web application
What I did
Requirements gathering, stakeholder workshops, heuristic analysis, task analysis, visual design, usability testing (concept and validation), QA, timeline creation
Timeline
8 weeks (4 two-week sprints) with iterative handoffs
Old & Current
Who does this help? 7,000 globally distributed reservation agents
70% non-native English speakers
Full time employees have more tenure (>20 years) than vendors (>2 years)
68% of agents have drastically more familiarity with technology than the remaining 32%
Tenured agents have change fatigue from the amount of new software they have been introduced to
How might we increase agent adoption of Navigator?
The problems:
Information overload; not enough hierarchy
Race to release new features without maintenance to existing ones
Still minimum viable, so agents still rely heavily on the old tool
Need to accommodate for information density
The solution: to solve agent pain points by improving the itinerary card.
The itinerary card houses the most important content for agents, yet, is the most difficult for them to decipher.
Qualitative data revealed that it was worth investigating the effectiveness of the itinerary. I conducted a heuristic analysis which revealed that the trip summary is tough to understand, there is redundant content, and it is spatially cramped.
Through content testing, we were looking to gather insight on information architecture and where we can avoid redundancies.
What did we learn?
All in 1 place
Agents were excited and relieved to see explorations that allow them to view the entire itinerary on one screen.
Less upfront is more
Agents preferred the most basic data at the surface level to allow for increased scannability.
Keep relevant data proximal
Qualitative data provided hierarchical clarity in 1st level vs 2nd level data, as well as data groupings.
Medium-fi explorations
We observed agents conduct task flows using our prototypes (2 rounds)
Agents do select flights to change
Confirmed key agent workflow interactions that had not previously been validated.
Meaningful hints for multiple user types
Intentional tooltips are helpful for new agents, and not too disruptive for the more tenured ones.
Appreciation for second-level data groupings
“See how detailed that is? That’s really good…We need all of that, that is great, even the award miles...perfect!”
Post-release feedback
Navigator is now the preferred and used tool driving efficiency for over 7,000 agents. Agent anecdotes from the company-wide newsletter:
Problem solving based on user-driven insights helps to increase tool adoption.
whats next?
Steer product strategy toward addressing agent usability issues to continue to increase agent use
Sharpen agent focus with enhanced navigation
Address split feedback from the different agent demographics by exploring preferences
What could we do better?
Gather more quantitative data with post production testing and analytics
Continue to align product patterns to design system