GOAL: INCREASE USER ENGAGEMENT (CONVERSION) ON BANANAREPUBLIC.COM
SOLUTION: SURFACE THE MOST RELEVANT PRODUCTS TO EACH USER THROUGH A PERSONALIZED EXPERIENCE.
Competitors such as Nordstrom and Amazon focus on creating a personalized user experience by offering users recommendations based on their purchase and browse history. GapLabs wanted to create a similar personalized experience that connected with the Banana Republic customer and improved conversion metrics.
BACKGROUND & RESEARCH:
The algorithm for generating product recommendations had already been built when my team was asked to design an experience to host these recommendations. We started with a competitive analysis to understand the places recommendations were surfaced on other retail sites.
Amazon's entire retail strategy seems to be around personalized recommendations. The home page is packed with every sort of recommendation - based on purchase history, browsing history, wish lists, etc! This was my Amazon home page today:
While this layout of endless rows of recommendations seems to work for Amazon, it wasn't going to work from a brand perspective for Banana Republic. Banana Republic wanted the opportunity to do storytelling on the home page- not just show products.
Nordstrom, also well known for recommendations, takes a different approach. After a branded "Let's Get Personal" call to action on the division page, they delve into a page specifically for recommendations. The experience surfaces to the user which actions on the site their recommendations are based on, and it is updated in real time.
Based on competitive analysis and discussion with the Banana Republic team on what types of experiences aligned with the brand strategy, we landed on a multi-tier approach.
We wanted to launch a 'Recommended for You' category page. Similarly to Nordstrom, we would confine recommendations to its own section for now until we honed the experience. After vetting results, we would expand into sprinkling recommendations into other areas of the site as appropriate.
Version 1 live on the site in January 2016:
What we learned: we stabilized the recommendations algorithm and how it plugged into the site, but the Recommendations Page was not receiving much traffic. We decided to provide product recommendations along her shopping journey in placements and moments that were more relevant to her.
Based on the learnings from the previous phase, next we wanted to include recommendations in the shopping bag, in-situ bag, and product page. Mock ups of some of these experiences are below.
What we learned: Showing recommendations interspersed in relevant parts of her shopping journey was very successful, especially the in-situ bag shown above. Context and timing are really important!
In the third phase, we wanted to add some additional capabilities to the product recommendations page to make it even more relevant to our user. Some of these capabilities we brainstormed included: offering the user the option to re-purchase something they've purchased before, proposing outfitting ideas to 'complete the look' of an item the user already owns from the site, or showing new colors available in a style the user previously bought.
Visioning for these ideas are below.
Based on these learnings from our early tests with product recommendations, we decided to partner with a 3rd party recommendations algorithm vendor to continue to strengthen our capabilities. We worked with them to A/B test variations on where and how to show recommendations. I continued to consult as the UX Designer on all personalization initiatives at Gap Inc.
ROLE & TEAM
Brand: Banana Republic
Team: UX Designer, Front End Developer, Product Manager, Marketing Lead, Web Producer, Big Data / Business Analytics Engineer
My Role: UX Designer
beginning November 2015