Trunk Club Scales Personalized Shopping with a Stylist in Your Pocket
Increase in Stylist-Member Connections
Increase in Stylist Efficiency
User Engagement with New Feature
How the tech-savvy retailer increased stylist productivity 90% through mobile app optimization
The Trunk Club stylists need to provide top-notch, personal service efficiently in order for the company to scale its business.
“If members are just there to window shop, it really cuts into our stylists’ time,” says Justin Hughes, Director of Product at Trunk Club. “We have a lot of people sign up and immediately request a bunch of things via the [Trunk Club] app. That requesting process can be painful for the stylist because it generates a lot of noise, when a lot of the guys aren’t actually interested in receiving the product.”
So they wondered: As the company grows, how can they scale their customer-stylist interactions and make them as efficient as possible? And what are the mobile-specific behaviors that Trunk Club can incorporate into their data-driven portrait of a qualified customer? With these questions in mind, Justin and the Trunk Club engineering team used Optimizely to run two experiments on their mobile app experience.
Business Impact of Experiments:
- 43% increased likelihood that a member and stylist would connect and ship a trunk
- 90% increase in Trunk Club stylists' efficiency
- Safely introduced a “liking” user behavior that engaged 10% of active users
- Increased featured merchandise requests by 8.6%
Requesting a trunk: No credit card required (Top) versus requiring payment information (Bottom)
- The stylist and customer were 43% more likely to connect and ship a trunk of clothing, a key step in Trunk Club’s customer conversion funnel
- Due to higher quality of leads from the app, stylist productivity improved 90%
Requiring credit card information led to fewer overall conversions, but the users that requested a trunk were more qualified and converted to purchases at a higher rate. The experiment provided stylists with more productive leads. “We saw about half as many trunk requests come through—but then we were twice as likely to actually convert those trunk requests [to shipments],” says Hughes.
This was a tremendous win both for the stylists, who are focused on offering a high level of service to thousands of new members each day, and for the Trunk Club customer, who was more likely to understand that requesting a trunk was an action that required a purchase intent, and more likely to move further down the purchase funnel after requesting a trunk in the app.
The Liking Experiment
Just like with the credit card experiment, the Trunk Club mobile team wanted to find a way to allow users to engage with the app in a way that wouldn’t inundate stylists with requests—but that also wouldn’t reduce the number of qualified trunk requests they were receiving through the app.
They wanted to experiment with how users could indicate style preferences. Trunk Club app users at first communicated their style preferences only by requesting clothes. The team wanted a low-risk way introduce a new action that would fill the gap between browsing and requesting clothing that would both engage users and provide useful data to the stylists.
Using Optimizely, the team was able to validate the introduction of a new “like” button feature. They first tested the feature a limited amount of their app traffic to ensure that key metrics like trunk requests remained stable.
- Would a softer form of preference indication—likes—be used?
- Would like cannibalize purchase requests in the app?
Liking: Users add choices to their trunks versus users like/heart their style preferences.
- 10% of users in the experimental treatment engaged with the feature
- Merchandise requests increased by almost 9%
The new like button was an immediate success, and the Trunk Club team began to “double up” on the amount of traffic included in the “like” treatment of the app.
Members would “like” products 10-12 times in a sitting, shattering engagement expectations and offering a great window into the member’s wardrobe preferences and style. “We also found that people were more likely to submit a trunk request when we got them to like things in the app,” says Hughes.
By testing the introduction of the “like” feature, the team was able to ensure that clothing requests were not cannibalized by the new behavior, and proved that introducing liking was actually beneficial for both user engagement and trunk requests, the team’s top-line metric.
We’re trying to take a stressful, painful experience and make it as easy as possible, and Optimizely helps us do that.
Mike CruzVP of Engineering, Trunk Club