Over €4.5 million in additional revenue from visitors who were already in the shop

Thomas Sabo is a well-known jewelry brand with its own online shop. Jewelry is an emotional purchase. But online, you lose what convinces people in a store: touching, trying on, getting advice.

The question was which doubts hold customers back in the shop. And which changes actually bring more orders. Instead of a relaunch, we set up a testing program. Every change had to prove itself in an A/B test.

We started with an analysis phase. A web analytics screening showed where users drop off. Several independent experts evaluated the shop for usability, communication, and consumer psychology.

In a MotivationLab, we accompanied jewelry buyers. First, we explored their motives and expectations for buying jewelry. Then they went through the shop and several competitors. The lab revealed clear hurdles. Important product information was hard to find. The checkout lacked a sense of security. And the brand's benefits went unnoticed. In a roadmap workshop, we prioritized the hypotheses together with Thomas Sabo.

Value signals: Show why here and not somewhere else. The product page put the piece of jewelry more clearly at the center. A bar with the brand's benefits accompanied customers all the way to checkout. Result: +4.98% orders and +7.76% revenue per user.

Zero Risk Bias: Remove the last bit of risk. People value a completely eliminated risk more than a merely reduced one. So in the cart, free returns moved to the front. Distracting exits stepped back, and a note right before submitting took away last doubts. Result: +4.19% completed orders. The variant went live for all customers immediately.

Salience: What isn't in view has no effect. We added more details about the quality of the materials, mostly below the fold. Result: no measurable effect. The learning: You have to persuade where the purchase decision happens.

Hedonic shopping: Orientation isn't everything. New filters were meant to make searching on the category page easier. They were used far more often. But they didn't bring more orders. Jewelry shoppers want to browse and get inspired. Efficiency can crowd that out. The variant was not rolled out.

Consistency: The same promises on every device. The mobile shop lacked the value promises of the desktop shop. So quality, free returns, and the matching packaging were added to the mobile product list and the mobile product page as well. The winners were rolled out to all devices.

Together, the winning variants brought 9.4% more orders. Through the end of 2019, that adds up to over €4.5 million in additional revenue. Relative to the cost of analysis and testing, that is an ROI of over 2,400%.

We calculate with net revenue and straight-line depreciation over 36 months per test. Testing took place on desktop. Because the winners were also rolled out on mobile, we apply the effect across all devices.

Conservative view based on our statistical ROI model, taking net revenue and depreciation into account.

Security sells. The winners removed doubts instead of showing more product.

Orientation alone isn't enough. For emotional products, a page has to invite browsing.

Tests without uplift are results too. They prevented bad investments and sharpened the next hypotheses.