Personalized Commerce at Scale for a Leading Baltic Beauty Retailer

Empowering a Baltic beauty retailer to scale smarter by centralizing product feeds, personalizing search, and bringing full promo control into business hands.

Quick Info

Groovy-configured General Product Feed

powering exports to 3+ third-party platforms.

Real-time integrations

with Google Products, Feefo Reviews, Peerius, and more.

Custom Apache Solr logic

enabling manual prioritization of search results by business teams.

AI implementation

for dynamic product discovery.

Client

A regional health and beauty retailer serving Latvian and Lithuanian markets.

Need

Boosting the existing ecommerce platform, improving customer experience, and enabling precise control over promotions, search behavior, and product data exports.

Solution

To streamline ecommerce content management, product discovery, and data exchange across integrated systems, Expert Soft implemented a range of improvements aligned with both technical goals and business impact:

Improvements

Centralized General Product Feed

We developed a unified product export system — General Product Feed — capable of transmitting the information to numerous configurations. Instead of maintaining separate pipelines for each business unit, the feed was configured using Groovy scripts and allowed for flexible business unit-specific adjustments. This eliminated redundant service development and simplified platform maintenance.

Business-user-controlled content updates

To reduce dependency on development teams, we designed a user-friendly system allowing customers to add information to their working system on their business unit. This enabled business users to upload promotional content, product details, and localized banners independently. Once changes were submitted, the system generated structured files, which we parsed and processed automatically. This ensured that content updates could be applied quickly across markets while reducing release overhead and engineering bottlenecks.

AI Keyword Recommendation Engine for dynamic product discovery

We integrated an external AI-based keyword recommendation system to boost personalized navigation. For each product category, AI-generated keywords are presented as clickable tags. When clicked, they route users to dynamically generated pages showing relevant product selections, also powered by the AI service’s filtering logic. This enhanced user engagement, improved cross-category navigation, and made product discovery more contextual and tailored.

Together, these enhancements supported scalable operations across content, catalog, and campaign workflows, empowering both technical teams and non-technical stakeholders to manage and optimize ecommerce experiences more efficiently.

Challenges

Handling edge cases in gift-with-purchase promotions

One of the client’s promotional campaigns included a “Gift with Purchase” (GWP) mechanic, where customers would automatically receive a complimentary item once certain products or cart values were met. However, a technical complication emerged: the same gift items were also available for direct sale. If customers manually added these products to their cart before meeting the GWP criteria, the promotion engine misinterpreted the context and applied a 100% discount, treating the manually added item as a triggered gift.

This led to pricing inconsistencies and potential revenue loss. We conducted a detailed investigation into the third-party promotion engine’s evaluation flow and collaborated with both technical and business teams to refine the discount logic. The final solution ensured that gifts were only added once the cart met specific criteria and prevented manually added items from being misclassified. As a result, we preserved the intended customer experience and safeguarded the integrity of promotional pricing.

Search personalization via Solr customization

The default Apache Solr integration in SAP Commerce Cloud didn’t provide business users with the control they needed to prioritize specific products in search results, particularly during campaign periods when surfacing high-margin or promotional items was critical. To address this limitation, the Expert Soft team designed and implemented a customization in Solr that allows local merchandising teams to define manual product ranking rules directly.

Results

Empowered business teams with full content and promo control

Business users gained autonomy to manage product data, search relevance, and campaign content without relying on developers, reducing bottlenecks and speeding up go-to-market.

Streamlined platform maintenance and cross-market scalability

Centralizing product feeds and system logic simplified support across business units, making the platform easier to scale and adapt for localized needs.

Boosted customer experience through personalization and precision

AI-driven navigation and manual search ranking improved product discoverability, making shopping more intuitive and tailored to customer intent.

01 / 02

Technologies

SAP Commerce Cloud, Apache Solr, Groovy, AI Keyword Recommendation Engine