TLS Central
UX Research, UX Design, UI Design
A tool for American Express to communicate with their luxury hotel partners around the world.
TLS Central
TLS Central is a tool used by American Express to facilitate communication with its luxury hotel partners worldwide. I worked on this large-scale system for over two years, contributing to both the hotel-facing and Amex-facing interfaces.
This case study focuses on phase 1 of the redesign process.
Platform: Web
Team: Me + Additional UX Designer
Timeline: 8 weeks
Identifying Issues
Before embarking on the global redesign, we aimed to identify the key pain points our users were experiencing, allowing us to prioritize the areas that needed the most improvement for phase 1.
To achieve this, we conducted 10 user interviews with the primary contacts at our hotel partners—those who interact most frequently with Amex through TLS Central. These users rely on the tool to complete action plans, update property information, and report renovations or closures for their properties.
We followed up by creating an affinity map, which revealed a clear picture of the key frustrations our users encountered while using the tool.
Method:
Moderated Usability Testing on the current product, paired with User Interviews to gather qualitative data.
Audience:
Primary Contacts for hotels and resorts (lodging).
Goal:
Identify biggest pain points within TLS Central.
”I’m so scared that there is something I need to update that I missed. This program rules the market.”
-a powerful quote from a frustrated user
Our Primary User
Key insights from user research
Our users primarily raised concerns about performance metrics, as maintaining their status in the AmEx program is critical to them. They often feel disorganized navigating TLS Central, struggling to ensure they are up to date and compliant with program requirements.
Recurring Frustrations:
It should be easier to check card member survey results.
The platform is not easy to navigate.
There’s a need for more straightforward access to property scorecard results and reservations training.
A centralized, easily accessible location for managing all outstanding action items on the hotel's end.
my personal favorite - The UI looks ARCHAIC
TLS central users need a way to save time by easily checking their performance metrics and be able to navigate the site with ease so that they can get back to performing their job duties.
Check-In
At this stage, I received approval from my team at AmEx to proceed with the redesign of the site navigation and performance section.
These two areas sparked the most discussion among users and were the most contentious. I chose to focus the redesign efforts here to achieve a quick and meaningful impact.
Current State
Here’s what the dashboard looked like prior to the redesign: it lacked navigation and provided no means to rectify mistakes. Users had to navigate the platform on their own to find solutions—definitely not ideal!
Round 1
These aren’t exactly wireframes. You might wonder why that is. The answer is straightforward: American Express has a highly developed design system, making it easier to begin with a somewhat high-fidelity approach.
Add navigation to the page so that the user can navigate within TLS.
Allow users to enter and select their brand or property from anywhere on the site. This way they are always seeing information that is custom to them, instead of updating from a dropdown on each screen.
Allow users to access their list of actions from the navigation so that they know what needs to be done during their site visit.
Consolidate all of the details tabs to one page for a more seamless experience.
Let’s Test
We utilized Maze to test the new navigation and Performance Dashboard prototype. Users were assigned a set of tasks to complete, and we tracked metrics against the original version.
Tasks:
• Identify the number of missing partner training and go to the training page
• Download the most recent scorecard
• Check current outstanding action items
94%
Direct Success
31%
Less time on Task
18%
Less time per screen
Further improvements
It turns out that our users aren't particularly fond of pie charts and would prefer them to be replaced with something more easily understandable.
Users reported not being able to understand their card member survey results.
• Time on task decreased by 20% with the new prototype.
• 98% direct success rate with the new prototype, while 20% of all success was indirect with the original site.
• Confidence scale average was an average of 4.5/5 for the new prototype and 2/5 for the original site.
After phase 3 of the design process, we tested the performance section again with 5 additional users.
Revisiting
Designing a complex system like this involves careful time management and patience. It was rewarding to engage with a process-oriented project that allowed me to leverage my expertise in new ways, and of course to design something that will help people in their daily lives.
Final Thoughts
Thanks for reading!
@2024 Emma Hosn