Arcor transforms sell-out into intelligent decisions and takes its alfajores to more points of sale

When sell-in is no longer enough to see real demand

‍

In the food and candy business, minor planning imbalances quickly translate into empty shelves, expired products and dissatisfied customers. Grupo Arcor experienced this challenge in one of its most sensitive categories: the Alfajores line, with 10 SKUs, different lifespans and high turnover. Demand was met from a plant and 3 distribution centers that supply more than 185 accounts in Argentina, so understanding what was happening at the point of sale was critical.

‍

The planning process was guided by sell-in and by consensus between commercial and supply chain areas. The signal that reached production and logistics was contaminated by commercial negotiations, pre-orders and operating practices of the different channels. The result was known: volatility in the plan, suboptimal use of production lines, distribution centers with limited visibility of future demand, and short-term corrective actions.

‍

At the ends of the chain, the impact was twofold: stock failures that resulted in lost sales and, at the same time, surpluses with a limited useful life that led to obsolescence or defensive promotions. Arcor needed to see beyond the customer's order and get closer to real consumer demand, integrating sell-out data, inventory and stock policies into a single planning logic.

‍

‍

From data noise to a clean consumer signal

‍

The first step was to redefine the demand and inventory planning process around the sell out. Together with Novix, Arcor initiated a proof of concept in this complex category, taking advantage of the fact that it had daily stock and sales information from a significant part of its own customers. On this basis, we worked on three fronts: data quality, portfolio segmentation and definition of inventory policies by type of customer.

‍

A VMI model was designed in which Arcor manages the inventory of its distributors, calculating target levels by SKU and counting according to consumption behavior, useful life and lead time. To achieve this, different sell-out sources were integrated, formats were harmonized, outliers were cleaned and models were built to forecast future consumption in tactical and operational horizons.

‍

The resulting policy translates the projected sell out into weekly replacement needs, ensuring a “healthy” stock at each node. When planning promotions or launches, the sales team enters expected sell-out adjustments, which the model transforms into supply requirements. Subjectivity ceases to dominate the process and is replaced by a signal of genuine, traceable and shared demand.

‍

‍

Intelligent VMI with Pyplan: from prediction to action

‍

On the Pyplan platform, demand planning, inventories and VMI in a single digital model. Generative AI and machine learning agents combine statistical models to adjust forecasts and stock policies, while the collaborative layer allows us to simulate scenarios and align S&OP with weekly replenishment in CDs and distributors.

‍

Pyplan's unified architecture connects tactical vision with operational execution, giving Arcor a single version of the truth for sales, supply chain and finance. In this way, decisions stop being reactive and become predictive and prescriptive: the system identifies bankruptcy risks, suggests replenishments, prioritizes SKUs and orients production based on real consumer demand.

‍

The impact is seen throughout the network: fewer bankruptcies and less obsolescence, greater turnover of SKUs among customers, better used production lines and a level of service that allows alfajores to arrive fresher and in the right quantity at each point of sale.

‍

‍

Do you want to take your demand and inventory planning to the next level?
Learn about our solutions or talk to an expert:
https://pyplan.com/es/entre-en-contacto/

‍