Migros -a Top 50 Grocery Retailer Worldwide- reduces inventory days by 11% and increases availability by 1.7%

Migros Case Study - At a Glance

Migros is one of the world’s top-50 grocery retailers, with over 2000 stores and annual revenue of $4 billion. Migros serves its customers with a variety of formats, ranging from 50,000 sq-feet hypermarkets to 1,000 square-feet convenience stores. With a well-deserved reputation for superior customer service and a wide selection of products, including high-quality fresh produce and fresh meals, Migros hypermarkets and convenience stores are the go-to grocery shopping destination for millions of customers every day.

Migros appointed invent.ai as their trusted partner to help manage their business-critical inventory operations. Following A/B testing in 200 stores, Migros rolled out the invent.ai's AI-Powered solutions across their 11 warehouses and 2000 stores. These solutions provide Demand Forecasting, Store Replenishment, and DC Replenishment.

Why did Migros choose invent.ai?

Migros is an innovative retailer with a history of efficient inventory management. As part of their DNA, they are constantly looking to improve their operations and customer service.

Following an in-depth assessment of 10 solution providers, they chose invent.ai's AI-driven forecasting and replenishment solutions to further reduce inventory levels without compromising product availability across their large, complex SKUs.   

A profit-optimizing solution to conflicting KPIs

A fundamental challenge in retail is managing the trade-off between conflicting KPIs, such as inventory turns and in-stock ratios, across a wide range of products and stores and multiple distribution channels. The core of invent.ai's solutions is the financial portfolio optimization approach which upgrades inventory decisions from a judgment-driven, KPI-measured process to a financial optimization-driven and measured process. This change enables retailers to simultaneously deliver superior customer service while squeezing every last basis point of margin from the business.

The team at Invent Analytics tailored their solutions to meet Migros’ specific needs. After successful results in the A/B test, the systems rolled out for daily planning of Migros’ 2000 stores and 11 warehouse operations by taking the following information into account:

• Store capacities and visual minimum quantities
• Demand fluctuations with special days and promotions
• E-commerce order fulfilment from the stores
• Expiry dates for perishable items
• Seasonal products such as fruits and vegetables
• Salvage costs of daily fresh products and fresh meals
• Promotion effects on competing or complementary products
• Supplier constraints

How Migros benefits from sophisticated AI-powered demand forecasting

Each day, the system detects outliers and lost sales and automatically selects the most effective AI-based forecasts from the product-store-day level forecast models. That means it can predict the probability of sales for each unit of each item at each store.

Store Replenishment Optimization

Taking all the business rules, strategic priorities, and the supply chain and store constraints, the system then calculates the optimal inventory levels and places orders that maximize profitability.

Advanced DC Forecasting & Replenishment

The solution generates store replenishment orders from multiple warehouses at the product-store-day level for over 30,000 SKUs at 2000 stores. The algorithm also calculates the e-commerce order fulfilment probability of each store to ensure availability with omni-channel orders. Total omni-channel demand at the product-store level is aggregated and supply recommendations are generated for optimal safety stock and maximum profit. The system also simultaneously plans warehouse replenishment from around 5000 suppliers.

The Migros team sets the strategy, invent.ai system takes care of the rest

Implementing the invent.ai's store and warehouse replenishment solutions has saved the Migros inventory planning team a huge amount of time. They no longer need to calibrate the system parameters to deal with changing operational developments each day. The system itself determines the optimal re-order points and order quantities taking into account product profitability, strategic considerations, changing demand patterns, and supply chain constraints.