How Lisek.App increased product availability up to 98% and accelerated inventory turnover with Occubee
- Client:
- Lisek.app
- Duration of cooperation:
- March 2023 – present
- Industry:
- FMCG
Lisek.App is a convenience store at your fingertips, right in your phone. Operating in nine cities across Poland, it provides a seamless shopping experience with express delivery. The service offers groceries, cosmetics, household chemicals, and products for children and pets. It operates within the quick commerce (q-commerce) model, where short delivery times are critical.
The brand is owned by Zorro Delivery Sp. z o.o., a rapidly growing startup positioned as a leader in the Polish e-grocery market. The company manages a network of “dark stores”—dedicated hubs where couriers collect products for direct delivery to customers.
Background and challenges
Lisek.App operates in the highly dynamic q-commerce environment, where customers expect immediate product availability. Determining which products to order, in what quantities, and when to supply dark stores is a significant challenge—especially when calculations are performed in Excel, which has limited capacity and processing power. With approximately 105,000 product-location combinations, manual methods became extremely inefficient.
Replenishment decisions must account for more than just forecasted demand; they must also consider current stock levels across all dark stores and warehouses, storage capacities, goods in transit, and various quantitative parameters.Furthermore, the system must meet specific supplier requirements regarding minimum order quantities (MOQ) and values.
Because the Lisek.App assortment includes ultrafresh categories—such as bread, fruit, vegetables, and meat—expiry dates are a vital factor. Food waste caused by expired shelf life presents a major financial burden.
The logistics model also requires agility. Lisek.App utilizes a cross-dock model: selected products are ordered to a central hub, sorted, and immediately dispatched to dark stores without long-term storage. Other categories, such as bread, are delivered directly to dark stores by local suppliers. Currently, the average inventory turnover is 22 days.
As the business scaled and the assortment expanded, the complexity of inventory management increased. Maintaining reliability and meeting high customer expectations necessitated deep operational automation.
Initially, we managed inventory manually using massive Excel files. This worked at a small scale, but as the company grew, it became inefficient. If we had continued with those methods, our team would have to be twice as large to handle the hundreds of supplier orders generated daily
-
– says Michał Mazur, Head of Supply Chain and Operations Excellence at Lisek.App.
Goals and directions
of change
The supply chain management team at Lisek.App sought a solution to optimize the replenishment of both warehouses and dark stores. This optimization aimed to:
- Increase product availability by reducing out-of-stocks.
- Reduce waste by eliminating overstocks in food categories.
- Decrease capital tied up in inventory.
- Increase inventory turnover.
- Automate sales forecasting and supplier order generation.
Achieving these goals required advanced AI algorithms and machine learning models. These tools enabled a comprehensive and detailed data analysis that incorporates new variables—such as holiday calendars and special events (e.g., shopping Sundays)—which were previously difficult to account for in manual calculations.
Implementation
of Occubee
The scope of the Occubee implementation at Lisek.App included:
- – Sales forecasting based on Machine Learning (ML) models.
- – Automated calculation of product requirements for dark stores and warehouses.
- – Automated generation of picking orders for local suppliers and warehouses.
- – Automated generation of purchase orders for warehouse suppliers.
During the implementation, Occubee’s data science team developed two ML models. Initially, these covered 65% of the assortment, while the remainder utilized statistical methods due to sparse sales data. Today, the entire assortment is forecasted using ML models.
All business processes managed by Occubee were parameterized to meet specific client requirements. When calculating demand, the system accounts for goods in transit and defined minimum/optimum/maximum stock levels. It also incorporates business strategies, such as the specific sequence for filling warehouses.
Based on this demand, the system generates picking orders and supplier orders. These results depend on pre-defined parameters—including logistics units and MOQs—and selected replenishment strategies. Occubee also factors in lead times and supplier holiday calendars to generate orders with sufficient lead time to ensure availability.
Lisek.App operates in a unique business model. Instead of traditional stores, it has a network of dark stores; instead of a standard central warehouse, it manages a complex structure of warehouses and a cross-dock platform. It was crucial to reflect these characteristics in Occubee. We achieved this through the flexibility of our platform and close collaboration between teams. Both organizations share a ‘startup DNA’—thinking and acting in a dynamic, agile, and smart way
-
– says Michał Koziara, Chief Operating Officer at Occubee S.A.
.
Benefits and impact
on the Client’s business
The implementation of Occubee has optimized replenishment across Lisek.App’s entire network. Consequently, product availability has increased, waste has been reduced, and capital previously tied up in excess stock has been freed for other investments.
While many factors contributed to Lisek.App’s average basket value doubling during this period, increased availability played a major role. Using precise sales forecasts, overall availability rose to 94%, reaching as high as 98% for key product groups.
Simultaneously, the capital engaged in inventory was reduced through accelerated turnover. The average Days Inventory Outstanding (DIO) was slashed from 40 days to 22 days, significantly improving the company’s financial efficiency.
In the FMCG industry, controlling waste is critical. Occubee provided better control over the “ultrafresh” category. Advanced forecasting reduced overstocks, thereby decreasing food waste due to expiration. For dry goods (e.g., pasta, grains, nuts, sugar), waste has been almost entirely eliminated.
The system has fundamentally changed the daily operations of the supply chain team. Manual order preparation, which used to take several hours a day, is now automated. The risk of errors inherent in large spreadsheets has been minimized. Despite the number of suppliers increasing nearly fourfold, the replenishment team size has remained the same. Automation has allowed these specialists to focus on strategic decisions and further process optimization.
Moving to a fully automated mode required trust in the system, which we built through intensive cooperation with the Occubee team. We noticed that after a few weeks, the system could order as well as, or better than, a human. Our intervention is now minimal, yet I have a greater sense of control. I know that if needed, we can adjust parameters to meet new requirements. With our rapidly growing scale, this provides immense comfort
-
– concludes Michał Mazur.
Key results at a glance
