Research Phase:
The issue didn’t stop with my friend. The problems were common and showed consistently low usability.
Without access to Amtrak users or internal analytics, I gathered feedback from Amtrak forums and created my own benchmark data using SUS surveys to quantify the issue and measure improvement.
Pain Points
Interviewee needed outside tools to complete task.
During the contextual interview, I observed my friend visit 51 pages, search for route info in “At the Station,” check the trip plan map and route pages, download PDFs, and search Google with multiple tabs open. Even then, he was confused trying to piece together a trip and find information the system assumed he knew.
Usability was the weakest point.
Usability consistently scored lowest of 4 SUPR-Q categories, often tied with the Loyalty category. Interestingly, I found some older users made more errors in usability testing but rated it less harshly; however, even with those results factored in, usability still scored an F average.
There were frequent issues, dating back years.
Forum mining showed users repeating the same complaints over the past 6 years: invalid dates, station errors, and unfamiliar travel norms. Users didn’t know what part of the search was the problem (dates? location?), and resorted to trial-and-error just to move forward.
Analysis
Users expect trip booking to feel like a negotiation. Instead, it behaves like a password check.
Design Direction
Users expect alternatives instead of "no." This can be accomplished using 3 design principles:
User Control & Flexibility:
Always present an alternative path forward. Let users adjust results based on what they do know.
Error Prevention:
Errors don’t need to exist if the system proactively guides users toward available stations, routes, or dates.
Recognition, Not Recall:
Don’t make users rely on memory or prior knowledge. Present options clearly so users never have to guess or Google.
Exploring Different Designs
I explored four approaches to help users understand routes and navigate the booking flow:
Option 1: Map Toggle Within Search Page
✓ Uses existing UI patterns
✓ Keeps users in familiar app booking flow
✗ May clutter the interface when not needed
✗ Harder to make responsive
Option 2: Expanded "At the Station" Feature
✓ Answers "Where can I go from here?"
✓ Matches user behavior & search patterns
✗ Doesn't directly address route confusion
✗ Page-by-page rollout = slow
Option 3: Color-Coded Trip-Planning Map
✓ Visual route guidance directly in booking
✓ One map = everything in one place
✗ Overwhelms simple booking scenarios
✗ Requires lots of development work
Option 4: Redesigned Info Architecture & Route Labeling
✓ Makes existing assets more findable
✓ Flexible for different users
✗ Doesn't directly address route confusion
✗ Would require broader site overhaul
Final Decision: Hybrid of All 4. A separate Map page (used only when needed) supports flexible, complex bookings without disrupting simpler flows. It combines:
• Search flexibility (Option 2)
• Spatial map (Option 1)
• Contextual route info (Options 3 & 4)
No new features. Just reorganized data in a way that scales to user needs.
Iterations + Reasoning
Driven by user testing and feedback
Added a route selection screen for more complicated use cases, to help customers visualize
Added an internal UI users pick their route visually before prices. Reduced cognitive load, fewer dead ends.
Simplified the map to reduce errors and speed up comprehension.
De-emphasized train names, boosted geographic clarity. Faster orientation, more intuitive discovery.
Split fares and seats into separate steps to reduce cognitive load and prevent errors.
Moved seat selection to a new page, with breadcrumbs. Clearer flow, no lost progress, more control.
Key Final Designs
High-fidelity prototype
Search locations as broadly as needed, but only on available dates
Browse available destinations reachable from search
View all possible routes between selected stations
Confirm details and visual path of chosen route
See all fare options for chosen route
Confirm trip itinerary and general price before selecting seats
Style Guide
Using parts of Amtrak's official brand guidelines I could find.
Outcome :
Helping Amtrak Compete on Experience
These improvements to the search and booking experiences help reinforce the company's “Retrain Travel” marketing. Amtrak can become a “less stressful alternative” to flying or driving, starting from the booking experience.
Impact Beyond Marketing
Estimated using industry averages:
Daily call volume 35K × $10 avg call cost × 1% reduction × 30 days
(Not internal Amtrak data).
~3,600% ROI on designing to avoid errors:
Based on transportation industry-standard benchmarks, even a 1% reduction in daily support calls could generate an estimated ~$107K/month in recovered value.
The solution supports Amtrak’s long-term goal to double ridership by 2040:
The redesign helps more users to book successfully by reducing common friction points like user errors, slow times to task, and incomplete flows.
Key Takeways
Document each design pivot.
I now document in each Figma version page the reason for any design changes so if I change projects, I can quickly recover context and move forward without wasting time re-creating my reasoning from scratch.
When in doubt, reuse patterns
If I could start over, I would first audit for and document Amtrak’s existing design patterns beyond what I could find from their existing brand guidelines document. Reusing what already works reduces UI design fatigue and keeps solutions more feasible for engineering.
Don’t overdo surveys
I learned that more concise surveys produce clearer insights. Asking the open-ended questions earlier than the Likert scale questions and consolidating similar questions would reduce users' fatigue which could provide better feedback.