+13% to the average basket: AI solution for a leading Ukrainian retailer

The era when deep consumer behavior analytics was a privilege of Western giants is long gone. Today, the Ukrainian retail market is seeing a rapid increase in demand for tools that enable retailers to make decisions based not on intuition, but on accurate data. One of the brightest examples of this data-driven approach is the Num8erz.Customer Insights platform, which is actively gaining traction both locally and internationally.

What is Num8erz.Customer Insights, an AI-based platform

It is a predictive analytics solution powered by AI and machine learning designed to provide deep, comprehensive analysis of sales and customer behavior. Built specifically for retailers and CPG companies, it transforms large volumes of sales data into actionable insights, helping customers receive relevant offers and businesses significantly boost their revenue.

Consulting4Retail and Num8erz, both part of the Atriny Group, are responsible for implementing this solution and have numerous success stories in collaboration with leading Ukrainian retailers.

Real-World Application

Phase One — Implementing the “Frequently Bought Together” Analytical Model

In 2024, the project’s first phase was launched with one of Ukraine’s top retailers. It focused on identifying products that customers frequently purchase together.

The goal was to detect product pairs or groups with the highest co-purchase frequency and recommend the complementary product to customers who only buy one item from the pair.

Delivering such personalized recommendations enabled the retailer to expand cross-selling opportunities, increase customer basket size, boost the share of private label products, raise revenue, and improve customer engagement.

Key steps of this approach included:

1. Gathering and analyzing historical data and transactions to identify reliable association rules.
2. Identifying loyal customers who purchase only one item from a potential pair.
3. Generating the ideal product pairing for each customer based on what’s already in their basket.
4. Integrating with user interfaces and testing effectiveness via A/B testing.

Scaling recommendations to a broader customer base.

Pilot Project Results:

• Increase in visits to recommended product sets by 1.3 to 1.5 times.
• 13% increase in the average customer basket size among those who received personalized recommendations.

Next Phase — Expanding Personalized Marketing Capabilities

Currently, preparations are underway for the second phase of the project. It will involve deep customer segmentation based on behavior, basket structure, and shopping frequency. This will allow the creation of personalized offers that factor in:

• Product relevance
• Seasonality
• Stock availability
• Active promotions
• Private label prioritization

The model continuously trains on new data and adapts to changes in customer behavior.

This phase will also include A/B testing with performance tracking across several key metrics.

Final Thoughts

In modern retail, the use of tools like Num8erz.Customer Insights is essential for companies aiming to take a systematic, data-driven approach.

They not only help optimize assortment and promotional activities but also create a solid foundation for personalized customer engagement, which is especially crucial in a competitive market with rising consumer expectations.

To learn more about how this platform can help you anticipate and meet customer needs, contact the experts at Consulting4Retail.