← Case Studies

AI-powered eCommerce platform transforming recruitment advertising

How Divante rebuilt meinestadt.de commerce with AI-driven product matching, dynamic storefronts, and checkout optimisation for Germany's largest local portal.

3x
Conversion rate uplift
2M+
SKU catalogue
-68%
Zero-results pages

Background

meinestadt.de is one of Germany's largest local portals, connecting people with jobs, local services, and regional information. Its recruitment advertising commerce flows needed to support complex product matching, dynamic storefronts, and a catalogue at national scale.

The platform had to serve both high-intent buyers and internal teams managing a large, fast-changing set of advertising products.

The Challenge

Keyword-only discovery was limiting product matching across a large catalogue. Users could describe what they needed, but search often failed to connect intent with the right advertising product or package.

The commerce experience also needed to support dynamic product presentation and checkout optimisation without creating a rigid monolithic stack that would slow future iteration.

Our Solution

Divante rebuilt the commerce experience around semantic product discovery, AI-driven recommendations, and dynamic storefront logic. Search moved beyond exact keyword matching to understand buyer intent and surface relevant products from a large catalogue.

The architecture kept commerce services modular and API-led, allowing storefront, discovery, and checkout flows to evolve independently as the product model changed.

Impact & Results

Search-to-purchase conversion increased 3x after semantic search and recommendation improvements.

Zero-results pages decreased by 68%, reducing dead ends in the buying journey.

The team gained a more flexible commerce platform for future recruitment advertising products and campaign formats.

"

Semantic search and AI-driven product matching turned a complex catalogue into a clearer buying path for customers and internal teams.

MD
Commerce teammeinestadt.de