How Rausch used Personizely to drive +106% revenue per visitor with A/B testing

Goals & challenges

As Rausch continued evolving its digital presence, the team focused on a key question: how much impact can product presentation have on buying behavior?

They had already built a strong foundation, high-quality products, recognizable branding, and a polished online experience. But on the product pages, there was still uncertainty around how much the wording and visual details influenced results.

Could AI-generated descriptions outperform human copy? Would different tones or formats affect how users convert or how much they spend?

To find out, Rausch launched a series of A/B tests with Personizely. The goal was to move past assumptions and gather clear data on what actually improves performance.

A/B test 1: AI product description experiment

Objective

To evaluate whether AI-generated product descriptions can outperform the original human-written copy in driving purchases, improving revenue per visitor, and increasing conversion efficiency on the product landing page for Rausch Herz Gold.

Test setup

• Original version: Human-written product description traditionally used by Rausch’s internal content team.

• Variation 1 (ROT): AI-generated product copy written with a focus on clarity and structure, but limited emotional appeal.

• Variation 2 (BLAU): AI-generated copy optimized for emotional resonance, sensory language, and premium positioning.

Insights

Variation 2 (BLAU) was the clear outperformer across all key metrics:

• Conversion rate jumped to 58.70%, a 78% improvement over the original version.

• Revenue per visitor more than doubled, reaching €43.38, indicating not just more conversions, but higher-value ones.

• Average order value also increased by nearly €10, suggesting that the emotionally optimized copy didn’t just convert, it also inspired bigger baskets.

The confidence level of 99.95% validates that these differences weren’t random. This was a true win driven by the content’s emotional resonance and clarity.

Takeaway

The test revealed a compelling truth: emotionally intelligent AI copywriting can beat both traditional human content and more neutral AI variants. By speaking to the senses and reinforcing premium value, Variation 2 (BLAU) didn’t just lift conversions, it transformed revenue efficiency across the board. For Rausch, this wasn’t just a better product description; it was a scalable insight into how storytelling, even in chocolate, drives digital growth.

Variation 2 (BLAU) achieved the highest conversion rate at 58.7%. It outperformed both the original and Variation 1 (ROT), showing that trust-focused, clear messaging influenced more customers to complete purchases.

Revenue per visitor increased across all versions, with Variation 2 leading at €43.38. This was more than double the original, confirming that structured and confident messaging can significantly boost transaction value.

Average order value was highest with Variation 2, reaching €73.91. The original and ROT performed similarly, but the emotionally refined version encouraged customers to spend more per purchase.

A/B test 2: AI-generated hover image on collection page

Objective

To test whether replacing the secondary image on product cards (visible on hover) with an AI-generated variation would influence user purchase behaviour and revenue metrics on the collection (category) page.

Test setup

• Original version: Product cards displayed a traditional lifestyle or pack-shot image when hovered over.
• Variation 1: Product cards displayed an AI-generated image designed to better emphasise texture, lighting, and premium cues on hover.
• Placement: The A/B test was deployed on Rausch’s Gebäck (biscuit) collection page.

Insights

This test yielded modest but consistent gains:

The conversion rate increased slightly, from 20.08% to 20.81%, suggesting the AI-generated imagery may have had a subtle impact on persuading hesitant users to click through or complete purchases.

Revenue per visitor improved marginally (+0.74%), resulting in an estimated €1,422.90 uplift annually, based on current traffic.

However, average order value slightly decreased, meaning users may have converted more frequently, but with smaller carts.

With a confidence level of only 23.51%, the uplift isn’t statistically reliable yet. But the consistent directional improvement suggests this could be a scalable idea with further refinement or paired with additional visual enhancements.

Takeaway

While not a breakout winner, the test demonstrated that even small UX adjustments, like AI-generated secondary images, can yield measurable improvements. The experiment revealed the potential of AI-powered visual optimization to subtly influence shopping behavior, especially on visual-heavy pages like product collections. Rausch now has a low-effort, scalable strategy to explore further: refining AI hover imagery to create even stronger visual momentum toward purchase.

Variation 1 with AI-generated hover images showed a small increase in conversion rate (+3.66%) compared to the original (20.81% vs. 20.08%), but the result was not statistically significant (confidence: 23.51%).

Revenue per visitor increased marginally from €13.66 to €13.76 in the variation. While not a major uplift, it signals that small visual changes on collection pages can influence shopping intent.

The average order value decreased from €68.03 in the original to €66.12 in the variation. This implies that while more users completed purchases, they spent slightly less per order.

A/B test 3: Hover video on product card

Objective

To evaluate whether replacing the static secondary image on hover with a product video (GIF-style auto-play) increases user engagement, conversion rates, and overall revenue metrics on Rausch’s collection pages.

Test setup

• Original version: Product card shows a secondary static image on hover.
• Variation (Mit Video): The hover image is replaced with a looping product video (GIF-style) showing the product in action.

Insights

• Hover videos outperformed static images in all key eCommerce metrics.
• Conversion rate rose to 9.33%, a 25.39% increase, showing stronger engagement and shopping intent.
• Revenue per visitor grew by over 52%, confirming that video interaction didn’t just attract clicks, it translated into real sales.
• Average Order Value increased significantly from €64.30 to €78.03, meaning video viewers were not only more likely to purchase, but they also spent more per order.
• The experiment delivered a noteworthy revenue uplift (+€705.88 across test reach), even with modest traffic.

Takeaway

Motion leads to emotion and action. The hover video provided a rich, immersive preview of the product experience, translating to higher-order values and more confident purchases. This test demonstrates how micro-UX enhancements, like contextual motion, can significantly boost performance at scale, without requiring a deep funnel redesign.

Rausch can now explore rolling out this enhancement across additional categories or testing product-specific video storytelling, maximizing return on assets already created for marketing or social media.

The hover video increased the conversion rate from 7.44% to 9.33%, a relative lift of 25.39%. While the confidence level was 59.97%, this indicates a promising direction for future visual enhancements.

Revenue per visitor rose from €4.79 in the original to €7.28 with the video, a 52.16% improvement. This suggests the video content helped turn more visits into higher-value transactions.

The average order value increased from €64.30 to €78.03 in the video variation, a gain of €13.73 (+21.4%). Shoppers exposed to video spent more per order on average.

Insights & recommendations

1. Emotionally intelligent copy drives performance

The most significant insight from Rausch’s testing journey came from their first experiment: emotion-rich AI-generated descriptions (BLAU) outperformed both traditional human copy and technically clear alternatives (ROT). By tapping into sensory language and emotional cues, Rausch saw a 78% lift in conversion rate and more than double the revenue per visitor. The takeaway? When the product is emotionally resonant, like premium chocolate, storytelling outperforms mere information.

2. Visual micro-UX changes can compound impact

While the AI hover image (A/B Test 2) showed only modest gains, it still revealed a useful lesson: small visual tweaks on collection pages can create directional lifts in conversion behavior. Even without statistical significance, the consistent upward movement across conversion rate, RPV, and purchases validates that subtle image refinement, particularly in a brand where aesthetics are core, shouldn’t be overlooked.

3. Motion boosts engagement and spending

In A/B Test 3, replacing a static hover image with a looping product video led to strong improvements across the board: conversion rate (+25.39%), revenue per visitor (+52.16%), and average order value (+21.4%). This confirms that contextual video, even in a small format like hover previews, can dramatically enhance user experience, conveying premium quality, texture, and brand mood in ways that static visuals cannot.

4. Average Order Value (AOV) is as important as conversions

In all three experiments, AOV proved to be a crucial revenue lever. Especially in the first and third tests, users exposed to richer content (emotional copy or hover video) spent significantly more per purchase. These insights suggest that optimizing for “purchase depth,” not just frequency, should be a strategic focus for premium brands.

5. Test assumptions, even the polished ones

Rausch already had a refined online store. Yet these experiments revealed that even highly polished UX can hide missed opportunities. Without testing, emotionally intelligent copywriting, immersive hover videos, or image improvements may never have surfaced as revenue drivers. The broader insight: when every detail is already strong, the next level of optimization lies in testing what seems “good enough.”

Results & impact

1. Data-driven content leads to measurable growth

Rausch's first A/B test clearly demonstrated that emotional, AI-generated product copy can outperform even well-crafted human descriptions. The emotionally tuned version (BLAU) lifted conversion rate by 78.43%, revenue per visitor by 105.82%, and average order value by nearly €10. These were not marginal shifts, they were game-changing leaps validated by statistical significance. This result turned a simple copy test into a proven strategy for storytelling-driven eCommerce growth.

2. Small visual changes can guide future UX refinement

The second A/B test, replacing hover images with AI-generated visuals, did not deliver a statistical win, but it consistently pointed upward in terms of performance: +3.66% in conversion rate, +0.74% in revenue per visitor. Though average order value declined slightly, the overall momentum supports continued visual experimentation, especially in product discovery contexts like category pages.

3. Motion content converts at a premium

A/B Test 3, introducing looping hover videos, outperformed static imagery across all key metrics:

• Conversion rate increased by 25.39%
• Revenue per visitor jumped by 52.16%
• Average order value climbed by €13.73 (+21.4%)

These results confirmed that video formats on hover drive deeper engagement, better product comprehension, and higher spending. Even with modest traffic, the variant delivered measurable revenue gains, proving that small UX enhancements can yield meaningful returns when applied with precision.

4. Clear benchmarks for scaling

These experiments established strong benchmarks for what works at Rausch:

• Emotion-rich content = higher AOV + CVR
• Motion = better engagement and larger baskets
• Subtle imagery updates = opportunity zones for growth

Armed with this data, Rausch now has validated paths to scale its digital merchandising, without needing to overhaul the entire shopping experience.

Ready to discover what truly works for your business?

Just like Rausch you can turn data into decisions with Personizely’s A/B testing tools - no guesswork, just real growth.

Start your 14-day free trial and see what optimizations move the needle for your store. Prefer a walkthrough first? Schedule a live demo with our team and we’ll show you exactly how it works.