Wunderlich is the leading fruit and vegetable wholesaler in Tegernsee with a long-standing tradition. The now thriving business began in 1930 with a vision in a garage, similar to those in Silicon Valley, and today it is being run by the fourth generation, Maximilian and Anian Rampf. Every day, their team delivers fresh regional products, selected European goods, delicatessen and eggs to the gastronomy and hotel industry around the lake.
In addition to the quality and freshness of the food, Wunderlich emphasizes the importance of regional products. Shorter distances are not only more environmentally friendly but also enable faster delivery and improved taste quality.
As part of a new generation of gastronomy, the brothers see the relationship with their customers as a chance to connect beyond just delivering products. It's about mutual understanding and an exchange about our food culture. As members of this new wave of restaurateurs, Maximilian and Anian are also focusing on the technological advancement of their own business: How can we work as sustainably and resource-conscious as possible without compromising on quality? Bit by bit, they have modernized their daily operations, moving away from traditional methods of using pen and paper. Starting with digitizing order slips and using a digital version of the price list, they now employ state-of-the-art technology to optimize the ordering process.
This is exactly where the collaboration with Paretos comes into play. In the fast-paced business of delivering fresh fruit and vegetables, it is crucial to work efficiently and reduce operating costs by knowing what each customer will order and, consequently, what Wunderlich needs to have in stock.
We offer an AI-based, fully integrated and automated end-to-end solution that allows for the prediction of product demand several days in advance. Based on this, automatic recommendations for the optimal daily order quantity are generated. This allows the Wunderlich team to accurately plan the quantity of each product that needs to be sourced from the wholesale market to fulfill incoming customer orders. The time for order preparation and delivery of products to regionally distributed customers is significantly reduced as a result.
Challenge
Our customer delivers a variety of fresh products daily, which makes long-term storage impossible due to both Wunderlich and their customers' high demands for freshness and quality. Regular restocking from the Munich wholesale market is necessary to fulfill orders for the upcoming days.
The major challenges here are that they don't exactly know how much of each product they will sell in the coming days, and thus, it's hard for them to estimate how much of each product they should pre-order from the wholesale market to fulfill all orders on time and with high quality. Previously, the team at Wunderlich had to manually plan and prepare each individual product daily for the next day. This costs several hours of work every night.
Every day, hundreds of complex decisions need to be made to determine how much of each product to purchase, considering the expected demand in the next few days, current inventory, and actual orders for the next day. Insufficient inventory causes a large portion of operating costs, as missing products need to be replenished throughout the day in additional trips (around Lake Tegernsee).
Solution
We provide Wunderlich with our software solution, a demand forecast as well as an order recommendation for more than 100 fresh fruit and vegetable varieties, broken down for the upcoming seven days. The basis for this is our Replenishment Planning Use Case, which is successfully used by a multitude of retailers, e-commerce, CPG (Consumer Packaged Goods), and FMCG (Fast-Moving Consumer Goods) companies.
Every night, the forecasts per product for the next seven days are updated, as changes can occur based on weather and general demand fluctuations. In addition, our AI examines the past demand and order history of the products to provide specific recommendations for action. Using this forecast combined with the current inventory, Paretos creates optimized recommendations for the order quantity of each product that should be purchased at the wholesale market the next morning.
Given the different volatility in demand for each product, which is influenced by seasonality and other factors, our software provides action recommendations with a percentage of certainty for each product forecast. Until now, the customer had to rely on experience and gut feeling to estimate how many products will be sold in the coming days.
Our solution enables a data-driven approach to ensuring product availability, while simultaneously considering product freshness and other factors. This considerably reduces the effort associated with order preparation and simplifies daily work. Stockouts and costly ad-hoc deliveries, as well as guesswork and gut feeling, become things of the past.
"Because daily freshness isn't a coincidence." – Walter Rampf
Results
Wunderlich has been a customer of Paretos for almost two years and the results speak for themselves: Paretos' forecasts improve the accuracy of product demand by 30 percent compared to pen, paper, and gut feeling. At the same time, the error rate for top products is under 20 percent for the next seven days. What does this specifically mean for the Wunderlich team and their customers?
In the first step, increased customer satisfaction can be demonstrated, as the right amount of the desired products is delivered today (without later restocking). The "missing" restocking benefits both the environment and Wunderlich's staff: 8,000 kilometers on the road are saved per year and a total of 58 work days are saved through automatic predictions and less manual effort. At the same time, accurate AI-based forecasts from Paretos actively prevent food waste, as no fresh products need to be unnecessarily held in stock and, in the worst case, discarded.
The implementation of Decision Intelligence empowers Max and Anian to set important impulses for optimizing internal processes for tracking and ordering, laying the foundation for a more automated end-to-end process. The automated order recommendations for product demand bring Wunderlich one step closer to full automation of the ordering process.
