Grupo CRM increases its planning intelligence and transforms decisions into competitive advantage

When the complexity of retail requires faster and more accurate decisions

‍

The CRM Group operates one of the most dynamic retail ecosystems in Brazil, powered by iconic brands such as Kopenhagen, Chocolates Brasil Cacau and KopKoffee. Its network exceeds one thousand stores, with a strong national presence and a portfolio characterized by seasonal products, frequent launches and high sensitivity to promotions and campaigns. This demanding environment requires managing peaks in demand, anticipating consumer behavior and maintaining availability across the network — all without immobilizing excess capital — an extremely complex balance for any premium retail operation.

Before the project, the planning process faced challenges inherent to the business model. Forecasting relied heavily on manual analysis, with a strong focus on sell-in and limited sell-out visibility, making it difficult to integrate the reality of in-store consumption with supply and production decisions. The consensus meetings were based on multiple sources, which reduced agility and generated differences between areas on the interpretation of demand.

In addition, extreme seasonality —especially at Easter and Christmas—required inventory preconstructions and fine coordination between marketing, trade, sales, supply and production. The lack of a unified vision limited the ability to react to sudden changes, affecting the level of service and increasing the risk of bankruptcies or excesses.

In this context, CRM needed to transform its planning process, increase demand accuracy and connect its end-to-end operation to make faster, more collaborative and consumer-oriented decisions.

‍

‍

The search for a more intelligent, integrated and collaborative planning model

‍

The company defined a clear objective: to professionalize and digitize the process of Demand Planningg and S&OP, integrating everything from historical analysis to inter-area collaboration, consensus and operational preparation. It was essential to connect tactical planning (S&OP) with operational execution (S&OE), unifying metrics, criteria and visions so that sales, marketing, finance and supply worked on a single truth.

Key needs included:

• Incorporate external and transactional signals into the forecast: sell-out, campaigns, elasticity, events and seasonality.
• Model portfolios with more than 300 line and 150 seasonal SKUs, each with very different behaviors.
• Balance inventories, working capital and availability in critical seasons.
• Increase forecast accuracy and reduce operational plan preparation times.
• Implement a collaborative process with greater governance, visibility and alignment.

The company sought to move from a reactive approach—continuous corrections, isolated analyses, fragmented decisions—to a truly predictive process, supported by advanced analytics and a flexible platform capable of modeling business complexity.

‍

‍

The Pyplan revolution: from manual analysis to a predictive chain powered by AI

‍

Pyplan became the technological base for digitizing the CRM Group's end-to-end process. The solution integrated modules for data ingestion, exploratory analysis, statistical forecasting and machine learning, commercial collaboration, consensus and simulation, building a unified, flexible and scalable model, fully adapted to the particularities of the business.

The process began with the automatic incorporation of transactional and historical data. From there, analysts were able to explore series, identify events, segment portfolios and evaluate elasticity. In the forecasting phase, Pyplan enabled statistical and machine learning models for the construction of the base volume, substantially improving assertiveness thanks to the calibration by category, channel and type of sale.

‍

The Commercial, Marketing and Trade areas entered their settings directly into the platform, fueling a transparent and traceable collaborative flow. The consensus went on to include the sell-out and the opening of new stores, aligning the forecast with the actual retail operation. Finally, Pyplan integrated margin simulations and comparisons against expected revenues, raising the quality of the S&OP and allowing for comprehensive decisions on inventories, promotions and supply.

The AI agents incorporated into Pyplan — capable of analyzing large volumes of data, explaining demand drivers and detecting anomalies — made it possible to reduce the operational burden and accelerate the cycle. As a result, the team achieved significant improvements: +60% in forecast accuracy, - 50% in lost sales, - 16% in bankruptcies and a 75% reduction in the time needed to prepare plans, moving from a highly manual process to predictive, prescriptive and deeply collaborative planning.

Tactical and operational integration transformed the way in which CRM plans its end-to-end chain, connecting strategic decisions with daily execution and increasing the maturity of the process year after year.

‍

‍

Discover how Pyplan can transform your end-to-end planning.

Learn about our solutions or talk to an expert:
https://pyplan.com/es/entre-en-contacto/

‍