8,000 crew members. One broken app.
CrewPal is IndiGo's operational app used by all 8,000+ cabin crew members for shift management, duty tracking, and fatigue monitoring. When I inherited the project, it had 14 screens, constant complaints about usability, and a scheduling error rate that was costing the airline real money.
Operational UX is high-stakes design. A crew member checking duty assignments at 4am in an airport lounge needs information instantly, with zero ambiguity. Failure means flights delayed, compliance breached.
The app that crew dreaded opening
Shadow research revealed the real problem: crew members were screenshotting their schedules and sharing them on WhatsApp because the app was too slow and confusing to use at 4am. A mission-critical operational tool had been abandoned for a workaround.
Root causes: 14 separate screens for information that should be on one. No proactive alerts for fatigue threshold breaches. No offline capability for airport dead zones. Compliance requirements buried under layers of navigation.
Shadow research at 4am in terminal lounges
I spent 3 nights conducting shadow research — observing crew members using the app in their actual work environment. The results were more damning than any survey would have revealed.
Key insight: the most critical information (next duty, fatigue status, schedule changes) was on screen 7 of 14. Crew were navigating 6 screens every time they needed the one thing they checked 8x per day.
"I know my roster by heart because the app takes too long. I only use it for sign-off." — Cabin crew member, 6 years at IndiGo
Restructure, then redesign
The first decision was architectural, not visual. I restructured the 14 screens into 4 core views based on frequency of use: Today's Duty, My Schedule, Notifications, and Profile/Compliance. Everything else became secondary navigation.
Prototyped with 40 crew members across 3 months. Each round of testing happened in context — airport lounges, hotel rooms, on mobile with fatigue-level lighting conditions. By round 3, task completion time for the primary action (checking next duty) dropped from 47 seconds to 8 seconds.
One screen. Everything you need.
Today view as the home screen. Next duty, departure time, aircraft number, and fatigue status — all above the fold. No navigation required for the daily check-in.
Proactive fatigue alerts. Rather than requiring crew to check their fatigue status, the app surfaces alerts when thresholds are approaching. Compliance became passive rather than active.
Offline-first architecture. Worked with engineering to cache duty data for 72 hours. The app works fully offline — critical for airport dead zones.
The WhatsApp workaround disappeared
Three months post-launch, the informal WhatsApp schedule-sharing groups had reduced by 80%. Crew were using the app because it was faster than the workaround. That's the real measure of success.
What I'd do differently
Shadow research is non-negotiable for operational tools. No survey or interview would have revealed the 4am lounge behaviour. Physical context changes everything.
Architecture before aesthetics. The biggest impact came from restructuring information hierarchy — not from visual polish. Always solve the structure first.