Western Union is used by millions of customers with thousands of locations in over 200+ countries supporting at least 40+ languages. Each country comes with its filter requirements as well as the types of locations available. 

The goal was to redesign the Western Union agent locator to be responsive, account for increasingly complex filter requirements, get it closer to the current design system, and improve speed for users to find locations.

My Role: Full-stack product designer

Team: 1 Product Designer, 2 Contract Developer, 1 Product Manager

Duration: 8 Weeks

Software Tools: Sketch, Illustrator, Adobe XD, Usertesting.com

Defining the problem

Agent Locator was late to adopt responsive features and apply the design system because of technical constraints. Users were increasingly becoming frustrated with using the website since it wasn’t intuitive to find the right location. Western Union also wanted to provide additional ways to direct users towards digital solutions if the same transaction could be completed online.

My main problems to solve:

1. Improve ways for users to find the correct locations.

2. Design an experience that prioritizes quick scan-ability while still providing users with enough information.

The main constraints:

• Time — We had to work rapidly to build the product while an extra budget was available in the 2-month window.  

• Technology — We had to work with a 3rd party vendor who was not going to adopt our tech-stack.

Initial Search Page

Users were frustrated when landing on this page. The majority of the space is taken up by filter choices. Search is unnecessarily difficult because users have to input location information, country selection, and be knowledgeable of what an address is vs. agent filter. The first step to improving this page would be to remove unnecessary elements and simplify the input process.

Based on our observation of user behavior and site traffic the filters were rarely used at all. When asking first-time users about it, the vast majority of them said that they had no idea what any of the filters meant. For users who had more familiarity with Western Union retail locations, the filters were more understandable. To improve filter usage we needed to organize them as well as de-prioritize them.

Dense List Items

When comparing other search list items amongst our competitors, it was clear that each of our list items was a lot denser. While there could be arguments made that our list items were too large and did not provide enough items per view, the first step to improving them would be to standardize their font sizes, font type, and minimize the usage of different font colors.

In the early days of search engine optimization, text on the page served as a great tool for the marketing team to ensure that any search engine that crawls the website has the correct information listed out in search results. However, from a user perspective, these extraneous descriptions provided no real value to the user in their need to find locations.

Early Iterations

After spending time figuring out what types of solutions I would be pursuing, I drew out early sketches to explore solutions. Soon afterwards I created wireframes and tested them with early users to check for early interactions problems. Feedback was positive so I fixed a few details and started to the next stage of the designs.

Less Cluttered Search Page

It’s easier to understand the filters after grouping similar actions together. Ideally, country selection would be removed and Agent Name search results would be combined into one search. However, due to technical constraints, those areas were untouched in this iteration of change.

Organized Filters by function

To improve the utility of filters we organized them into functions. While this improved engagement, there was still some evidence based on card sorting sessions that there wasn’t a strong understanding of the filters. If it was within the project scope, it would be prudent of us to work on the nomenclature and word filters to the functions better towards what our users would understand. However, we could not make such changes because we needed to work with Retail & Marketing teams in future projects to create filter names that matched business language with filter labels that our customers could understand.

Font styling to improve List Items

The redesign sought to establish a visual hierarchy by using font weights and color instead of only font sizes for a cleaner, more modern look to establish clear and distinct information.  We could add value to each item by providing exacting timing for when a location would be closed to prevent users from needing to take another interaction to get the information they need. Future optimization could reduce the list item size but we would need to do so in future A/B testing to find the best balance for how much information each list item would hold.

Removed Extraneous SEO Text

There are now ways to provide that same information in the code for search engines to crawl without taking up valuable space that isn’t relevant for the user once they reach the page itself. Working with the development and marketing team we were able to get rid of extraneous SEO elements within the pages themselves without losing their important information necessary for search engine optimization.