While Aerie offers a great deal when customers buys a set number of underwear (referred to as an "undie bundle"), only about 50% of customers take advantage of the bundle deal. Furthermore, we learned from customers that some find it cumbersome to sort through and keep track of undies in their bag when trying to get the bundle deal online (ex: 10 undies for $35).
My Role: Project planning, wireframing, prototyping, user testing, feature prioritization
The Team: Product Management, User Experience, UI Design, Site Merchandising, Engineering
Timeline: 3 months
This project request started outside of the conventional AEO project process. The SVP of Digital asked me to create an exploratory design concept that made it easier for customers to get the undie bundle deal online.
I was told that less digital customers were getting the undie deal versus in-store customers because it was harder to shop for multiple undies online. I was not given any data or customer research to support this information, so before starting design concepting, I wanted to make sure this was an actual problem that needed solving.
I worked with data analytics and Aerie site merchandising to gather data to prove the assumption that less customers were get the undie deal online than in the stores. The data told us that undies (in all amounts) were purchased 48% more often in-store than online and for both in-store and online customers, only around 50% achieved the bundle deal.
So while we didn't know yet WHY customers were more likely to purchase undies in-store versus online, we did see an opportunity to try and increase the number of customers getting the undie bundle deal, because of those 50% of customers not getting the deal, they were more often buying three or more undies at a time than only buying only one or two.
We also learned through the data that if a customer was buying 10 undies (a common undie threshold for the bundle deal), of those customers 65% purchased more than one size in a single order. This statistic ended up coming in handy when we were designing a size-saving mechanism in our design.
After looking at the data, the UX research team sifted through customer comments and customer intervew footage to look for any evidence to support that customers found shopping for multiple undies online cumbersome.
Nothing significant came up from customer comments, but we did find two usertesting.com videos from the previous holiday season where customers walked through and described how getting an undie bundle as a gift took a lot of time and required a lot of repetitve actions. Even though we only found a few instances of this experience, I felt this was evidence enough that there was an opportunity to streamline this undie purchasing process for our digital customers. Ideally if we had more time, we would have talked to online undie customer's directly to better understand whether this was an issue.
After obtaining the customer research, I created a simple prototype user test to see whether we could solve the known undie customer's pains without causing new ones.
The experience allowed customers to add an undie from the category or product page to a side tray (known as "the bundler") that was accessible from any page on the site. The customer could simply tap/click to add or take away undies from the bundler and see their progress toward the bundle deal without leaving their shopping path.
Testing Research Questions
The test concept proved successful. Users intutively used the "+bundle" feature unprompted. The users that used and understood the feature said they liked it because it "created less back and forth" and allowed them to "see how close I am toward getting the deal." The 11% of users that didn't use the bundler proved that the bundler did not get in the way of their natural shopping experience. I reported the customer research and findings to the SVP of Digital and we discussed adding content to the experience to help customers understand that the bundler was not the bag.
After the initial research and testing was performed, a project team was created. We had a very short timeline to finalize the concept and develop it in time for the holiday season. I worked with the project manager to break the project into 3 iteative phases based on functional necessity and technical feasibility.
Phase 1: October (the MVP)
Phase 2: January
Phase 3: Timing not established
The first phase was the minimum viable product - the simpliest working form of the concept - so that we could get the project out the door for the holidays. Below are the live site designs that illustrate the experience.
When the first phase launched, I worked with a UX researcher to perform user testing on the live site experience. Based on the test findings, as well as some additional issues discovered during QA testing, I made the following design enhancements for the project's second phase:
As a part of our initial plan, we also implemented size preservation in the bundler and updated the design to accomodate two bundle deals running concurrently.
The undie bundler launched in October 2019. In January 2020, the feature boasted a 23% interaction rate, which well exceeded the business's expectations considering it had not been actively marketed to and was a new and unfamiliar paradigm for our customer.
We also reported the following KPIs for customers who interacted with the undie bundler (compared to non-bundler undie shoppers):
Buzz Feed mentioned the newly released undie bundler over the holiday in their article: "44 practical things worth buying on cyber weekend."
The last important feature I designed as part of the initial bundler concept was dynmaic messaging in the bundler that told the customer how many more undies they need to add to get the bundle deal. A main component of this feature was the ability to see prices for each undie in the bundler. When the promotion was achieved, the prices would update to the "bundle" sale price letting the customer know the promotion had been acheived (as seen below). This feature was slated to go live in the third phase, which has yet to be implemented.