Overview
Purple was placing all of their mattress cross-sells on the mattress product page, resulting in a massive cognitive load from both a narrative and information architecture standpoint. Customers were overwhelmed with the sheer quantity of information they were being asked to parse.
Gathering Data
To validate our suspicion about the root issue on the PDPs, we used data from several inputs. We had conversion data supplied by our analytics team and insights we gained from viewing heat maps and dozens of user sessions via Quantum Metric. The insights seemed to back our hunch: we were presenting customers with too many options – i.e. the cognitive load was too high.
Competitive Audit
Next we did extensive competitive audits. We found that most of our competitors were using similar patterns, so we then looked to out-of-market examples for inspiration. Based on a couple of examples we found, I came up with the idea of using an interstitial page for the the add-ons. We believed this would be successful for the following reasons:
1.
Significantly reduced cognitive load
2.
Decisions points broken up, allowing customers more “breathing room” to focus on the large-ticket items.
New Architecture
They key feature of the new workflow was our proposal to introduce and interstitial screen where we would house the promotions. Doing so would halve the number of decisions users needed to make on the PDP.
Design Exploration
We looked at several layouts, interactions and transitions for what we were calling the “post ATC” user flow. These were reviewed by stakeholders internally to help narrow down the final direction.
Final Design And Testing
Now that the direction had been approved, I created clickable prototypes I could use to test. I ran unmoderated test via Usertesting.com. The feedback identified areas that needed further refinement.
One of those refinements were added affordances to reduce friction between the PDP and interstitial page.
Results
It’s critical to define what success looks like before launching any kind of A/B test. In this case it was simple - we were looking for a lift in attachment rates and, by extension, the average order value (AOV). The results exceeded our expectations.
We saw a 7% lift in AOV, 35% increase in bed frames purchased with a mattress, and 61% lift in bedding upgrades. Combined, this was estimated to generate an extra $2.5M in revenue per year.