I led the end-to-end redesign of Huckberry’s checkout, from identifying the opportunity and conducting competitive research through design, experimentation, and launch. The work introduced guest checkout to reduce friction for first-time customers while preserving account creation as an important part of the customer experience.
A new checkout entry experience that gave first-time customers a clear guest path, recognized returning customers, and moved account creation beyond the purchase barrier.
more mobile orders, about 50 a day
Huge shout-out to Chris and the team for incredible comp research and new designs that simplified checkout for new customers, driving an incremental $4.5 million in revenue a year.
Account creation had long been treated as a retention lever at Huckberry. But drop-off at registration kept coming up with the data team, and a checkout review created room to ask: did signup need to happen before purchase at all?
Cart
Account creation
No guest path for first-time customers
Account credentials are required before checkout begins
Registration sits directly between the cart and purchase
Checkout
123Before proposing solutions, I audited checkout entry across 6 direct competitors, counting the steps between cart and payment and the information required before a customer could buy. I also reviewed 6 larger retailers to understand whether scale changed the pattern. It didn't.
| Brand | Guest checkout | Steps to buy | Fields before payment |
|---|---|---|---|
| Huckberry | Not offered | 5 | 11 |
| Patagonia | Offered | 2 | 5 |
| Arc’Teryx | Offered | 2 | 6 |
| Bespoke Post | Offered | 2 | 9 |
| Vuori | Offered | 1 | 8 |
| Todd Snyder | Offered | 1 | 9 |
| Buck Mason | Offered | 1 | 9 |
| Competitor average | 1.5 | 7.7 | |
Every brand reviewed allowed customers to purchase as guests. Requiring an account before purchase made Huckberry the exception, from direct competitors to much larger retailers.
Competitors averaged 1.5 steps and 7.7 fields before payment. Huckberry required 5 steps and 11 fields — more than 3× the steps and ~40% more fields to make a purchase.
Every checkout began with email. Recognition, authentication, and account creation happened after the customer entered the flow rather than acting as a gate in front of it.
Guest checkout wasn't a new idea. What had kept it off the table was retention. Account creation helped Huckberry turn first-time buyers into repeat customers, so simply removing registration wasn't enough. The experience needed to reduce friction without abandoning the business value behind it.
A way to complete their purchase without committing to an account first.
To preserve account capture because checkout signups supported repeat purchasing.
How might we reduce checkout abandonment while preserving the retention strategy the business depended on?
Let customers complete their purchase as guests, then move account creation into the post-purchase experience.
Cart
Guest checkout
No account required
Order confirmation
Optional account creation
Before moving into design, I worked with engineering to define what we could realistically change without introducing unnecessary risk to a revenue-critical flow.
from design to handoff
Work within the existing checkout architecture
Target high-impact moments within the existing flow
Minimize disruption to a revenue-critical flow
Rather than rebuilding checkout end to end, I focused the work around the largest barrier: entry into the flow, with supporting changes where they were needed across the broader experience. This gave us room to rethink how customers began checkout, support the new guest path through completion, and bring the experience back in line with the design system.
Even with an A/B test, changing a flow this close to revenue carried risk. Within our technical constraints, I explored two approaches that introduced guest checkout at different levels of prominence.
12
12Start balanced. Learn before pushing further.
We chose the balanced version for the initial test. It removed the account barrier while preserving a prominent path for returning customers, giving us a lower-risk way to validate the broader guest-checkout hypothesis. If the results held, the more aggressive guest-first concept was ready as the next experiment.
We launched the balanced experience against the existing checkout, initially exposing it to 10% of traffic, split evenly between the two, and ramping as the results held.
| Metric | Control | Variant | All traffic | Mobile |
|---|---|---|---|---|
| Orders placedPrimary | 0.37 | 0.39 | +3.72% | +6.56% |
| Product revenue | 94.52 | 96.09 | +1.66% | +4.48% |
| Average order value | $251.07 | $246.17 | −1.95% | −1.29% |
Decision metrics: balanced entry vs. control, sitewide and mobile
One metric moved in the other direction: average order value fell by roughly 2%. But the broader behavior supported the hypothesis. Removing the account wall brought more first-time customers through checkout, whose initial orders tended to be smaller. The increase in order volume more than offset the lower AOV, resulting in higher overall product revenue. Worth watching, not worth reversing.
The simple front-end decision to “continue as guest” created a broader systems problem behind the scenes. The final experience needed to account for new customers, recognized customers, authentication failures, post-purchase registration, and guest order management without fragmenting the checkout experience.
A lightweight email step minimizes perceived effort while allowing the system to identify returning customers automatically.
Existing customers are recognized and given a clear path to sign in without losing their progress.
Contextual messaging helps customers recover from authentication errors without losing momentum or trust.
Registration moves after purchase, when the customer has already received value and account creation no longer stands between them and the order.
A new order-lookup experience lets guest customers track and manage purchases without requiring an account.
For years the team had worked on optimizing registration. The larger opportunity was questioning whether registration belonged at the start of checkout at all.
The test also settled the question we'd left open. The balanced entry became the permanent experience, and when we later tested the guest-first version against it, it didn't perform as well. Keeping an equally visible path for returning customers mattered more than pushing everyone toward guest. Login-related support tickets trended down too, a quiet signal the smoother path was doing its job.
Problem: Navigation and the product hierarchy lived in separate admins, so nav categories drifted from the landing pages.
Impact: One workflow for merchandising, the full catalog in the nav, on every device. Now tracking nav clicks and conversion.
