Change Flights Online

Help Expedia flight customers and agents to easily change flights online.

 
 

CONTEXT

Calling support agent was the only way Expedia's flight customers could exchange their flights, leading to frustrating 1.5 hour phone calls with the agents on an average and a very poor NPS score of -12%.

 

BUSINESS GOAL

  • Reduce Average Handle Time for support agents

  • Minimize operational costs and agent error costs

  • Retire Legacy Flights tool and build on Expedia Voyager used for lodging to handle Flight transactions as well.

 

GOAL

Improve flight change experience for:

  • Customers: to make it fast and self serve

  • Agents: to streamline workflows

 

Impact

  • $3.9M cost savings per/annum

  • $1.5M incremental net revenue/annum

 

MY CONTRIBUTION

  • Lead the design efforts

  • Conduct research and usability sessions, create interactive prototypes and mocks, provide redlines

stakeholders

  • Training team, Supply, Finance, Fraud, Service Delivery

Team

  • Product Manager, Developers, Content writer and data analysts

 

research & Insights

Interviewed agents on how they changed existing flights. 
Agents hopped between multiple tools to handle a scenario making process inefficient and error prone.

Command Prompt for Global Distribution Systems (GDS)

Legacy Flights tool for agents

 

Project inception with cross functional team 

Worked with PM to list out all the flight change scenarios which were complex and spanned across teams/tools. It was a large problem space.

  • Partnered with PM to bring together different teams and stakeholders to understand the system’s complexity, identify the roadmap, key workflows, what we want to keep from existing system and what we want to change

  • Stakeholders voted and agreed on the high impact, high volume scenarios that should be tackled in the given timeframe which were:

    • One way and round trip changes for all travelers, ideal for customer self serve experience

    • Agent assisted flight changes for less than all travelers

    • Airline initiated schedule changes and refunds

Inception with stakeholders

 

Key Issue 1 | Customer scenario

Misleading results and surprise fees result in long, frustrating customer calls eroding trust

  • Free form search before call returns policy-ineligible options, misleading customers.

  • Customers must call support to assess selections and available flight options, leading to long, frustrating calls.

  • Surprise agent fees and penalties cause dissatisfaction and erode trust

 

UX DELIGHTER 1

1. Surface right policy information at the right time

Iteration1

Iteration2

Final

 

2. Conducted A/B test to inform how pricing should be displayed to customer to easily decide and change flights.

Variant 1: shows full price of each new flight option, which is later reconciled against what the traveler has already paid.

Variant 2 (Winner): shows only the additional (or reduced) cost relative to the already purchased itinerary.

 

3. Consistent flight-change experience across devices

 
 

RED LINES

 

Key Issue 2 | Agent SCENARIO

1. Fragmented agent workflow drives errors, delays, and costs

  1. Agents hopped across multiple tools making the process slow and error prone
    Complex nested hierarchy added clutter and confusion
    Ineligible search results led to agent errors and airline penalties for Expedia

 

UX DELIGHTER 2

1. Streamlined consolidated agent flow

  • Designed a guided workflow

  • Showed relevant flights only, with clear cost difference

  • Existing progress indicator caused usability issues

Iteration1

2. Defined standardized scalable design patterns

  • Designed and iterated on various steppers

  • Worked with design system team to finalize it

Final

Key Issue 3 | Agent scenario

Lack of hierarchy and missing error signals slow resolution

  • Calculations unclear and hard to follow

  • No clear visual hierarchy

  • Agents hit dead ends when errors surface too late

 

1. Change flow in Voyager tool

  • Simplified information hierarchy

  • Leveraged existing design system

Iteration1

 

2. Improved usability with new patterns

  • Worked with design system team to simplify payment summary

  • Easy lookup of fare rules

  • Clear Amount due summary

Final

 

Prototype

 

Learnings

  1. Collaborate with support to first hand understand user needs and challenges

  2. Regular standups and buddy testing with devs allow better collaboration

  3. Complex, policy-heavy flows demand end-to-end orchestration

Portfolio
payment model