top of page

Bosonozka Case Study: Driving Revenue Growth Through AI Product Discovery

  • Writer: Antónia Volčková
    Antónia Volčková
  • Jun 23
  • 2 min read

Updated: Jun 24


Bosonozka.cz is one of the largest Czech e-commerce retailers specializing in barefoot footwear. They sell online across multiple European markets and operate a brick-and-mortar store in Brno. Managing hundreds of products alongside strong seasonality places high demands on search quality and recommendation relevance.



The Challenge: Unlocking Revenue Lost to Ineffective Search and Product Discovery


The default search and recommendation features on Shoptet were leaving significant revenue on the table:

  • Keyword-based search: Standard search relied purely on keyword matching. Natural language queries, typos, or colloquial expressions returned poor or zero results. Every edge case required manual synonym configuration.

  • Irrelevant recommendations: Similar and complementary products lacked visual and contextual connection. Customers frequently missed out on products they would have otherwise purchased.

  • Overwhelming manual management: A growing catalog and seasonal product rotations meant the team spent hours hardcoding rules instead of focusing on growth strategy.


The Goal: Increase total revenue and Average Order Value (AOV) without driving up marketing costs.



The Solution: Intelligent Product Discovery Across the Customer Journey


Bosonozka deployed Raventic’s AI product discovery stack across three critical touchpoints of the customer journey from search to checkout:

  1. Semantic Search: Replaced rigid keyword matching with a deep understanding of shopper intent. The system automatically processes natural language, typos, and synonyms without requiring manual rules.

  2. Similar Products: Displayed visually and semantically relevant alternatives on product detail pages. This keeps customers engaged in the shopping funnel even if their initial click wasn't a perfect match.

  3. AI Cross-sell in Cart: Recommends complementary items at the moment of highest purchase intent. Suggestions are based on actual product relevance and real performance data, rather than generic "customers also bought" logic.


Seamless Onboarding: The entire stack was implemented via Google Tag Manager. No custom development, zero IT dependency, and absolutely no ongoing maintenance required.



The Test: Validating AI Performance Through A/B Testing


To measure the real-world business impact, Bosonozka launched a controlled A/B test during the high-volume Christmas season.

  • Duration: 19 days

  • Sample Size: ~44,800 users

  • Split: 50/50 Raventic vs. default Shoptet functionality

  • Tracking: GA4 + Raventic Analytics (event-based tracking)



The Results: Double-Digit Gains in Revenue and AOV


Overall Business Impact


+17.5%

higher revenue

+19.8%

higher Average Order Value (AOV) 


in the Raventic group compared to the control group


Performance by Feature

Feature

Revenue Share

Semantic Search

40% of revenue went through on-site search

Similar Products

41% of revenue driven via product recommendations

Cart Cross-sell

12% of revenue driven via in-cart recommendations



Why It Worked


  • Full-Funnel Optimization: Raventic optimizes search, product pages, and the shopping cart capturing value at every critical decision-making moment.

  • Zero Maintenance: No synonyms, no manual rules, and no seasonal reconfiguration. The AI interprets shopper intent entirely automatically.


What Bosonozka Says:


The A/B test results were clearly positive both revenue and average order value went up. The implementation via GTM was fast and required no effort from our dev team. For e-shops looking for a quick solution without custom development, Raventic is definitely a smart move.

Jaroslav Bednář, CTO Bosonozka.cz



Want Similar Results?

Let’s talk about how Raventic can help your e-shop grow with smarter product discovery.



bottom of page