How a Small Change Led to 81K Additional Orders

The Setup

A major DTC ecommerce brand had a login problem hiding in plain sight. Fewer than 2% of site visitors were logging into their accounts before checkout. The downstream effects were significant: reduced rebuy opportunities, complicated checkout flows for returning customers, and limited engagement with loyalty features like saved addresses and order history.

The login functionality existed — but behavioral data showed users weren’t engaging with it. The question was why.

The Insight

The login option was represented by a small person icon in the header utility nav — a common pattern, but one that relies on users recognizing the icon’s meaning and associating it with account access. On a page filled with competing visual elements (cart icon, wishlist, search, navigation), the icon blended into the background.

Users weren’t choosing not to log in. They couldn’t find where to do it. The problem wasn’t motivation — it was visibility.

The Approach

The hypothesis was intentionally simple: replace the person icon with explicit “Log In” text. No new features. No modal redesign. No changes to the authentication flow. Just make the existing action findable.

The test ran across all devices, measuring login clicks, account creation, Buy Again page traffic, and order placement.

The Result

Login clicks increased 36% with 99% statistical significance — on the first day. Orders reached significance within 2 days. Account creation trended up double digits, and traffic to the Buy Again page grew substantially as logged-in users could now access their purchase history and reorder.

The downstream impact: 81K additional orders and a 4.1% lift in overall order placement rate. A single text label change created a measurable revenue impact at scale.

The Takeaway

The highest-leverage experiments aren’t always the most complex. This test succeeded because it started with a behavioral observation — “users aren’t logging in” — and traced it back to a root cause that had nothing to do with the login experience itself. The authentication flow was fine. The problem was upstream: users couldn’t find the front door. Sometimes the best optimization isn’t building something new — it’s making what already exists visible.