Amtrak Re-Design

User Testing
The Problem Space

The Amtrak website breaks all ten of Jakob Nielsen's 10 general principles for interaction design. The website is difficult to use and creates an unpleasant experience for the user. In order to determine what aspect of the site we wanted to focus on for our re-design, we conducted user testing and created mid-fidelity prototypes. 

User Testing Plan 

We will have two people conducting the interviews at a time. One person will document the important things they see, the other will facilitate the interview verbally. The screen will be recorded so we can go back and analyze some of the screen engagement. Additionally, we will voice record the interviews as permitted by the interviewee. 

If the interviewee comes to a roadblock in the process of completing a task, we will ask them what steps they would take from that point if they encountered this problem in real life, rather than telling them how to proceed.

Tasks:

  1. Book a flexible coach seat for San Diego to San Francisco on December 1st in the morning.

  2. Purchase a monthly multi-ride ticket from Los Angeles to San Francisco starting a day in January 2020

  3. Find the interactive planning map

    1. Find a train station as close to Boulder, Colorado as possible

    2. Once you find that station, book a one-way train from San Diego to that station for November 10th

Top Usability Errors
  1. Lack of feedback when users run into problems purchasing multi-ride tickets

  2. Difficulty selecting a seat type and moving through the ticketing process

Problem Area #1 - Multi-Ride Error

After prompting the user to find multi ride tickets from San Diego to Portland, they are taken to the multi-ride search page. On this page, the user can specify the ‘to’ and ‘from’ destinations as well as the date they want to start. After attempting to purchase a monthly multi-ride pass between SAN and PDX, the user is presented with an error but no suggestions for how to fix it. The user then tries to have their train arrive at a different station in San Portland. At this point, due to the lack of error feedback, they are using trial-and-error to find what will give them a monthly multi-pass.

Problem Area #2 - Seat Selection Error

When the users are directed to select their seat, they are led to this page. The users were confused about how to select a ticket. There’s no clear affordance that tells the user that they need to press the button next to the price to select the flexible ticket, and once they do, how that relates to the departure time they want. Additionally, they had trouble finding the “add to cart” button, as it disappears when you scroll through train options. The sorting tools are also not automatically displayed, so our users did not discover their existence. 

We asked the user to select a flexible fare. The users often clicked on the Flexible drop down at the top of the screen. While it provided them more information on each type of fare, it did not explain how to book it.  The different fare types, “Saver”, “Value”, “Flexible”, etc on the top of the modal feels like a filter, so all of our users were not expecting that the fare descriptions would pop up.

When the previous attempt to select a flexible fare did not work, they would click on the underlined seat type within each of the possible trains. This would lead the pop up displayed on the left instead of moving them forward in purchasing a ticket.

Component To Re-Design - Seat Selection
Goals
  • Make it easier to select routes/seats/fares

  • Understand the difference between fare types

  • Make a more straightforward way to select seats given their fare type

  • Clarify how to proceed to the next step

  • Make the different routes more apparent to the user

Prototype #1

This re-design clearly displays fare types, costs, and the train options available. Filtering is more discoverable, and it is obvious which types of seats are available, and which are not. Breadcrumbs clearly display user journey, making the user work load minimal.

Prototype #2

This prototype creates a visual hierarchy between fare type and seat selection. The prices and call to actions are more visible, making the process smoother for the user. Amenity and transfer information is displayed immediately and clearly. 

Conclusion

While we were unable to test our prototypes, we believe that we created designs that would enhance the ticket purchasing process through Amtrak's website. Travel is already a frustrating experience, so users should face less challenges when figuring out their transportation. 

These prototypes were completed with Nielsen's 10 general principles for interaction design in mind. Our goal was to make our designs as user friendly as possible. Given more time, we would likely test our designs, and create more iterations that combine the positive aspects of each prototype.