Tailor AI
Case StudyA 13% Lift on the Homepage, No Dev Queue
ShopBack's US marketing team had a new hero in mind for their homepage. They specified it, Tailor built it in the browser, and they ran an A/B test against the original homepage design. Four days later they had a clear winner and shipped it to every visitor that same week.
The Challenge
ShopBack wanted a quick way to experiment with an animated, attention-grabbing homepage design, to find out whether it could improve CTA clicks while keeping their clear unique selling point: users earn Cashback when they shop, travel, and play games.
The Approach
ShopBack scoped the pilot tight on purpose: one page, click-based goals only. That kept the setup simple and let the US marketing team start testing quickly.
One line of JavaScript
The Tailor script went into the site tag, once. That let ShopBack's team easily create new homepage variants and A/B test them.
Marketing set the tests,
Tailor built them
ShopBack decided what to try. Simple swaps they made themselves using the Chrome extension. For ideas that are more sophisticated, they sent the copy and layout, and Tailor built it, usually back the same day or the next. Neither route needed ShopBack engineering time.
Winners become the new baseline
A winning version goes to 100% of visitors and becomes the page the next test runs against. Each experiment starts from the last one's win.
Scope note
- Where
- The shopback.com homepage, ShopBack's US site.
- Split
- Even, across everyone who landed there.
- Metric
- Share of visitors who clicked one of the 15 calls to action Tailor found on the page. Sign Up, Join For Free, Join Now, Add to Chrome, and their variations.
The Results
One experiment, four days, a clear winner that shipped to every visitor.
Rebuild the hero
A two-column hero with a new image, and an animated typewriter headline, tested against the original homepage design.
Two-Column Typewriter Hero
Jul 2 to Jul 6, 2026 · 4 days · 6,554 visitorsControl
Original page: 534 conversions on 3,249 visitors (16.4%)
Treatment
Rebuilt hero: 612 conversions on 3,305 visitors (18.5%)
The extra clicks spread across several buttons, which points to the layout working and not one button getting easier to find. ShopBack sent the winner to 100% of visitors on Jul 8, two days after the test ended. Nobody touched the codebase to do it.
Measured in Tailor
Swipe to view →
What We Learned
A tight scope got them testing sooner
One page, click goals only. A small footprint meant testing started within days rather than after a long evaluation.
A hero rebuild was enough
Reworking one section of the page ShopBack already had moved CTA clicks 13%. The rest of the homepage stayed as it was.
Each winner becomes the next starting line
Send a winning version to 100% of visitors and it becomes the page everyone sees, with no code change. The next test runs against the better homepage.
Same-day turnaround on variant requests
ShopBack sent the copy and layout they wanted and got a clickable preview back the same day or the next, every time. Iterating on a headline stopped being a scheduling problem.
Four days from idea to a proven winner
ShopBack ran this on the homepage they already had. Their team called the shot, Tailor built it, and the test settled it in four days. The winner went to every visitor two days later. The same loop runs on whatever page takes most of your paid traffic.
Test Setup
Run this play on your own homepage
See what Tailor would change on your highest-traffic page. Enter your domain for a live preview, no setup required.
Short on time? Book a demo and the Tailor team will set up the experiments, tailoring, and measurement for you.