Final design of Dashboard (anonymized)

Customer Service App

Primary tool for an airline's CS department
10.01.21 - 11.15.20


As an airline Customer Service Specialist. I need to be able to help customers cancel items related to their travel (trips, passengers, special services, etc.), so I can resolve the second highest reason customers call.


  • Product Manager
  • UX Lead
  • UX Designer
  • Program Manager
  • Product Owner (Client)


  • Decreased Average Call Time by 35% (exceeded goal by 10%)
  • Increased Customer Satisfaction scores by 14%

My Process

case study highlights are bolded


  • Stakeholder Interviews
  • Contextual Inquiries


  • Task Analysis
  • UX Strategy Blueprint
  • Bullseye Diagram


  • Ideation
  • Review & Critique
  • Concept Map
  • Wireframing


  • Visual Specs
  • Design System
  • Functional Spec


  • Analytics Review
  • Focus Group


This airline handles thousands of customer issues each day by phone. Hundreds of customer service representatives (Specialists) all lean on the same set of tools. However, the primary tool is over 10 years old and carries many inefficiencies.

How might we update the tool and improve the usability and user efficiency without diminishing other critical functions?

For this project I was responsible for UX research, UX strategy, Interaction Design, Visual Design, and leading and managing the UX team.


I started this project by gathering stakeholders and asking a set of questions I comprised several years ago. These questions involve the product, strategy, users, business, and surrounding system(s). Having answers to these questions allow me to identify knowledge gaps and recommend a research plan that meets the project objectives. We identified some crucial research activities:

  • Contextual Inquiries. There was little knowledge on the Specialists and how they used the tool, their behaviors, and attitudes. Because Specialists' time was limited, we chose to do 1 hour "side by sides" as opposed to diary studies.
  • Online Surveys. To get quantitative data to support our qualitative, we launched an online survey targeting demographics (name, role, etc.) and psychographics (tool satisfaction level, points of frustration, other tools used, etc.)

Since the existing tool was seen as "hard to use" and "repetitive", I thought it would be best for us to build a set of Task Analyses to measure and mark a few user flows before and after design was complete:

  • time on task
  • total steps
  • complex tasks (tasks that required external tools OR deviations from current path)
  • repeated tasks
Example Task Analysis for "Cancel A Single Passenger from Flight"

From this phase we were able to uncover several problems. The top two were...

  • Finding alternate flights was very difficult. The filters were very small and tightly packed, so users often found themselves using the wrong fields.
  • Cancelling anything less than the entire trip, for everyone, was difficult. If Specialists chose specific passengers to remove, future steps were confusing and rattled their confidence.


While synthesizing the research, I also wanted to draft a UX Strategy Blueprint* for the entire product. This would allow the team to clearly outline the product strategy and establish design guiding principles. These principles were essential to the subsequent tasks in design, and we constantly used them when grading concepts. These were our top "Guiding Principles":

  • Make tasks efficient while not sacrificing the ability to convey needed information over the phone
  • Reduce the need to leverage other tools in the airline's ecosystem

*Thank you Jim Kalbach for creating this template!

UX Strategy template


At the start of the design phase, we focused on conceptual approaches to solve the problems at hand. Lo-fi, quantity over quality. To keep each other fresh and motivated, I set up daily "collab" times where a designer could take the stage, propose a problem to solve, and then kick off the group into a wire framing session. We invited the entire team to take part in the process. Afterward, each person walked through their designs and talked about their rationale, which intentionally laddered up to our "Guiding Principles".

Initial concepts for removing a passenger from a trip (Figma)

One challenge to highlight was the complexity of the process driven by unstructured customer goals and a complex set of institutional rules. To surmount this, we created cheat sheets and set up lunch n learns to preface design work. We also created conceptual maps before any wireframes to make sure we could articulate how the solution worked and prove it could address complexity.

Conceptual map for "cancelling a passenger's seat upgrade"


We arrived our final design by iterative testing with real specialists via Our final design was tested with 5 Specialists from 5 different customer service groups. We found no thematic usability issues and measured our original metrics in task analyses. We built a prototype of the current tool so performance would not be an issue and found...

  • Decreased time on task by 55%
  • Decreased total steps by 30%
  • Decreased complex tasks by 15%
  • Decreased repeated tasks by 90%

As the design evolved, the Engineers beaome more involved. For this project, the engineering team specifically requested to stick close to design system components with minimal changes. I conducted weekly status meetings which became more of a "component map" review as time went on.

In this role, I also served as the Lead over the design system in Figma. It was our goal to make design files easily consumable by the engineers, so we built components just like they were going to build them. We also employed a set of best practices - topping that list was "Stay DRY (Don't Repeat Yourself)".

In tandem, I created a comprehensive "States Doc" (a list of conditional statements) that would accompany the visual specs and assure the team we were covering all the use cases.


2 months after deployment, the team measured success and found the following:

  • Decreased Average Call time by 35% (exceed goal by 10%)
  • Increased Customer Satisfaction scores by 14%

In Retrospect...

One thing I found particular success with is tying design rationale back to the UX strategy. The Client really appreciated the correlation and it gave the strategy more weight as the project progressed. Another practice that the development team really appreciated was the "States Doc". They really loved having full transparency to the functionality of the proposed design (not just "happy path" flows).

Next Case Study: Analytics DashboardBack to Case Studies