*Forgettable, but in a good way. The process should be so effortless for the user that it becomes an afterthought. This ensures a seamless shopping experience—more like a small speed bump than a full-blown, lane-changing traffic jam.
After years of feature enhancements, security fixes, and third-party payment integrations on our white-label e-commerce platform, the checkout experience my team had inherited was a tangled mess—a giant band-aid ball. It was slow. It was confusing. And it was causing session drop-offs, even after customers had invested significant time and effort building their item-heavy carts.
We had to do something! But where do you start and how do you get stakeholder buy-in to make sweeping changes? You lean on the data and validate against industry standard UX resources (like the Baymard Institute). By taking an approach that has function over form and puts the user first, we were able to reduce checkout abandonment to 21%-31% (varied by retailer).
Here's the definition from HotJar:
The percentage of potential customers who start the checkout process but don’t complete it.
Various factors influence this rate, and because our white-label solution allows clients to customize forms, payment methods, and sign-in wall placements, we faced the unique challenge of analyzing dozens of data sources to identify bottlenecks.
First, let's talk about what we found.
The initial checkout workflow was designed over nearly 8 years ago and worked off a stepped structure where users were processed from one view to the next after completing each form. So first things first, we opened up Google Analytics. The results weren’t overly surprising but here is what we found and the assumptions that we made.
.png)
Being a white-label solution with varying client needs, our product allowed retailers the flexibility of guest orders. But with growing requirements around third-party services, most instances were not able to support this feature as shopper information was required for certain service integrations (e.g. DoorDash and Instacart delivery). In looking at the data, most retailers had opted for required sign-in at:
Unsurprisingly, placing a gate check at the first sign of e-commerce intent is a bad idea. Similarly, requiring it during checkout can deter users who may get frustrated at the idea of more steps. However, in some configurations, a signed-in session is necessary. The data suggested that positioning this step between shopping and checkout results in the lowest exit rates—but we needed to back this further.
Next, we rounded up our list of retailers for competitive analysis–which includes retailers from both grocery and non-grocery.
Shopping and checkout are distinct cognitive workflows. Placing user authentication elsewhere forced customers to navigate an unrelated task, disrupting their experience. The transition from shopping to checkout was the ideal point for this shift—right when users were already expecting a change in context.
The issue was starting to look clearer and at this point the team was working through new workflow diagrams to ensure that enforcing a change would not break any existing UX workflows.
While we had a strong foundation for why our changes worked and why they should be implemented, we always backed our assumptions with data. The Baymard Institute has a number of research articles that provided points for us to focus on.
Changes were implemented one client at a time while we monitored Cart Abandonment metrics. As we saw positive results, it became easier to convince retailers to adopt the recommended user authentication settings. Over three months, we tracked sessions and collected user feedback through surveys. Ultimately, we achieved a ~17% reduction in checkout drop-offs and a 3% decrease in exit rates at the authentication step. With industry checkout abandonment rates exceeding 70%—and anything below 40% considered ‘good’—I’d call that a win!