During my time at Egencia, I got to work on dozens of features for a variety of products. Some were AB tests and some were whole-feature migrations from the legacy codebase to the new, modern codebase. Listed below are a handful of these features I designed for two of the teams I supported.
When a user has upcoming trips, we display up to 3 of them at the top of their homepage. Each of those trip cards has an action bar at the bottom which encourages travelers to attach other products to their trip (like another hotel night or a rental car). When we initially tested the placement of this action bar, it was a 2019 AB test conversion winner, however, the interaction flow wasn't totally clear and there was still some feedback indicating confusion.
In order to address this feedback and add some delightful interaction touches to this feature, we wanted to add contextual messaging as well as a clear way to exit this attach context.
If we indicate to all users when they have entered Attach Context and when they have exited Attach Context, then we will reduce page reloads by helping users understand where they are in the flow.
The timing of this test wasn't ideal as our traffic was significantly impacted due to COVID-19. The AB test was called inclusive, but the new design was rolled out. We saw feedback coming in from our various channels that the experience was clearer. One aspect that I'm really proud of is that I was able to challenge my engineering team to implement CSS animations. I felt strongly that the animations were critical to understanding the new design and they're one of the few instances of delight across the website.
Often when travelers book an airline ticket, they are not booking a refundable fare class. One option that airlines provide to customers is the ability to cancel that ticket and get an eCredit for that airline to apply to a later trip. We refer to these credits as Unused Tickets. On the Egencia Homepage, we display a module called Unused Ticket Tracker to inform Clients and Travel Consultants how many airline credits they have and for which airline.
After conducting interviews with Travel Consultants, we found that many of them did not realize that the UTT was clickable and contained all the relevant data they needed inside a popup. Agents instead would navigate to the legacy profile page of the traveler they were trying to book for in order to complete their tasks. Interviews with our Clients also revealed opportunities to alleviate confusion around the presentation of data. In the current design, the presentation of information did not inform the user what to do or how to use an unused ticket.
Separate client feedback indicated that companies want their arrangers and travelers to utilize these airline credits because its often a big money-saving lever they can pull, but right now, it's too confusing and not clear our user how they can self-serve. Indicating that many companies were losing money when these credits expired.
If we improve the design of the UTT module for all site visitors that have an airline credit available, then we expect to increase Travel Consultant productivity and increase client redemption of these credits.
Overall this has been a huge win for Agents (and inconclusive with all other user types). Unused Tickets are one of the most common tasks for agents and extremely time-consuming. They would have to navigate between multiple pages and programs to determine whether or not a credit could be used. With this new design, we have noticeably decreased time-on-task and have received lots of positive feedback from the agent team.
On the Client side, we added more educational content and surfaced more relevant information which will hopefully encourage our primary users to redeem their credits. Next iterations of this design will focus on getting users to redeem these credits instead of providing general awareness.
In an effort to streamline our search forms and provide a more personalized experience, we wanted to surface recent searches inside of the form inputs. We know that it takes multiple searches before making a single booking, and in business travel, it's highly likely that our travelers are running searches with similar criteria to make comparisons. I explored a variety of options - different combinations of content, different visual designs, different triggers, etc - but in the end, we ended up focusing on different priorities at the time. These designs never made it into testing. The below prototypes were built in Framer.
In collecting feedback from our main channels, we discovered that the inability to find receipts is the number one detractor on the Trips product (from NPS surveys). In addition, more than 330 comments in January 2019 were related to receipts, most of which stating that receipts are hard to find or access. Lastly, we recorded comments on Post-Trip Agent Contact surveys on travelers' inability to locate receipts post-trip.
If we provide a clear way to find the receipts of a recently completed trip, then travelers will more easily find their receipts, resulting in less client feedback about this particular problem.
Our goal with the design was to solve a major user pinpoint, however, this AB test was a major conversion loser. After digging into the data, it seemed like 1) our test was configured poorly and therefore the data was skewed, and 2) our technical ability to make this useful just isn't there yet. Receipts are a complicated aspect of managed travel. It's difficult to know when a receipt will be ready, whether or not the supplier or Egencia provides the receipt, and often users have to call a travel agent to resolve their issues anyway. We weren't able to create a strong enough decision tree that would point the user to a readily available receipt, and thus we weren't really solving a problem, just making it more apparent.
For Travel Arrangers and Travel Managers, their search form includes another input that regular Travelers do not have; a traveler picker. These users book travel on behalf of other people and therefore need to select whom they are booking before they can run a search. For a long time, the traveler picker was the primary source of negative feedback on the homepage. When a user clicked into this input, a dropdown list appeared in alpha-order, which suggested that the list was comprehensive, when it really wasn't (we just displayed the first 12 travelers). Further, many of the travelers that appeared in the list were irrelevant to the arranger (ie. 8 of 12 displayed travelers could be guest accounts), or the arranger was unable to find the person they needed because or a legal name vs. preferred name issue. All of these pain points lead to an effort to make the experience better.
If we provide a way for Arrangers to access a curated set of their travelers, they will be able to complete their tasks more effectively by finding the traveler they need to book for more quickly, resulting in less feedback about the traveler picker experience.
After release, we saw a lot of positive feedback from Arrangers. We made it much easier for them to find the people they most frequently select, as well as cleaned up the interaction. The AB test was inconclusive but since we saw an increase in step conversion, positive feedback, and engagement we rolled it out. This was a big step toward progressing to a more personalized experience on the homepage.
Travel Managers want to have access to all their travel data - hotel spend, air spend, out of policy metrics, ecological impact, YoY metrics, and much more. One feature that Egencia provided in our legacy product was the ability for these TMs to configure all possible data fields, customize the row and column output order, and cross-cut their data every which way. This legacy wizard took in all these inputs, then output an Excel file for download. When it was time to migrate this feature onto the new tech stack, the team wanted to significantly cut down the capabilities of the wizard. Primarily because the functionality we'd have to build essentially meant we would be rebuilding Excel or Google Sheets into our own product, which we did not want to do.
We chose "Data Hub" as the name for the new version of "Custom Report Builder." Our goal with Data Hub was to streamline the steps a Travel Manager had to take to get to their download, create a clearer design flow, and allow space for more features like saving and managing templates, managing past reports, and more. We also wanted to clearly indicate the user's progression in this flow, while allowing them to move between each step without losing their selections.
Below Left: Legacy version of Custom Report Builder. We removed quite a bit of the legacy features in order to retrain our users and better manage the migration.
Below Right: The new Data Hub feature set is pared down from legacy, but a clearer set of steps for the user to follow.
Meetings & Events is an obscure legacy product offering that allows clients to create an external event, set event-specific policies, and invite "guest" users to book for themselves on the platform. One feature the Homepage team wanted to migrate from legacy was Meetings Catcher. This feature would read a user's search criteria and check if it matched any existing meetings or events. Then it would surface some sort of indication before the user entered the search funnel and prompt whether or not they wanted to continue as part of the event. The goal for this feature migration was to support M&E clients and help their users stay within policy, since the policy would change if a search took place inside an Event context. This feature ended up not getting built, as team and company priorities shifted elsewhere.