Let AI summarize this page for you
Over the past few years, replenishment and supply chain management systems have become one of the fastest-evolving areas of innovation in retail. Much of this progress has been driven by increasingly accurate demand forecasting models, enabling retailers to automate replenishment with a high degree of precision. Today, automated replenishment has become an industry standard, helping retailers optimize inventory levels, reduce stockouts, and improve on-shelf availability.
What do Retailers expect from the Next Generation of Supply Chain Management Systems?
Gianni CASSANO
Country Manager of Retano
“The primary purpose of any automated replenishment system is to generate reliable demand forecasts and accurate purchase orders. Above all else, retailers expect these systems to make the vast majority of replenishment decisions autonomously, calculate supplier order quantities correctly, and do so faster and more accurately than a human planner. That’s why advancing our forecasting models remains one of Retano’s top priorities.
Retano SCM cleanses and enriches operational data, continuously learns from historical sales patterns, and identifies complex demand signals that conventional forecasting methods often miss. Based on these insights, the system automatically generates replenishment orders, enabling retailers to achieve a high level of automation while significantly reducing stockouts, optimizing inventory, improving shelf availability, and ultimately increasing sales performance.
However, we’ve observed a common pattern among retailers that have already embraced automated replenishment. Once automation becomes stable and consistently delivers value, a new question naturally emerges: How can we get even more out of the system? This is where human expertise becomes essential — not because the system is making mistakes, but because real-world retail operations include factors that no mathematical model can fully capture. These include local store characteristics, regional events, assortment changes, and countless operational nuances that are only visible to the people managing stores on a daily basis.
For this reason, we designed Retano SCM not only as an automated replenishment engine, but also as a comprehensive decision-support platform. Our first step was introducing a contextual AI assistant that helps users understand the system’s recommendations. We then expanded the platform with tools that allow planners and buyers to intervene precisely where human expertise creates additional value — while preserving all the benefits of automated replenishment.”For this reason, we designed Retano SCM not only as an automated replenishment engine, but also as a comprehensive decision-support platform. Our first step was introducing a contextual AI assistant that helps users understand the system’s recommendations. We then expanded the platform with tools that allow planners and buyers to intervene precisely where human expertise creates additional value — while preserving all the benefits of automated replenishment.”
Can Demand Forecasts be adjusted? With Retano SCM, yes
In many replenishment platforms, planners never actually work with the demand forecast itself. Instead, they receive replenishment orders that have already been generated by the system and can do little more than approve or reject them, without influencing the underlying forecasting logic. The problem is that demand doesn’t always follow predictable patterns. Some stores experience sudden spikes or recurring fluctuations that forecasting algorithms often classify as statistical noise or outliers, reducing their impact on future predictions.
In reality, these “anomalies” frequently have perfectly valid business explanations. For example, stores located near stadiums or major event venues experience demand patterns driven by local events. On match days, categories such as bottled water, beer, and snacks often see dramatic sales spikes. Without proper planning, these products can quickly go out of stock.
To address this challenge, Retano SCM introduces a dedicated event management capability. Buyers can upload event calendars — such as sports schedules — and use them as planning inputs for affected stores. Based on this information, they can adjust demand forecasts by applying category-specific uplift factors for selected dates.
The AI assists throughout the process by analyzing historical sales data, identifying similar situations from the past, and recommending an appropriate uplift value together with a confidence score.
Local assortment management integrated with automatic replenishment
For most retailers, assortment decisions are split between head office and stores. The central office defines a high-level framework — for example, up to 100 products per category — without going into the details of each individual store, while ordering decisions are left to store teams. This creates a significant operational burden: in practice, out of 100 available items, a store orders about 30 and must manually set the remaining items to zero. Moreover, if a product is not included in the local framework, it cannot be introduced even when there is clear demand. Any change to the assortment requires centralized processes, approvals, and timelines that are often incompatible with the speed of the market.
Retano SCM enables stores to directly manage local assortment within the automatic replenishment system. Decisions are made by the person responsible for store operations.
Before taking action, the user has access to a complete operational view using widgets such as Category Health, Category × Store Heatmap, and gap analysis, which highlight in real time issues, opportunities, and network performance. The system flags both products that are candidates for inclusion and those that should be removed, highlighting underperforming categories, OOS risks, low rotation, and untapped opportunities.
For example, a new flavor of chips may perform very well in some stores within the same cluster but remain absent in others due to the centralized assortment framework. In such cases, the system highlights the opportunity, estimates the expected impact on sales and profitability, and flags potential cannibalization effects, while still leaving the final decision to the store manager.
All changes follow a controlled workflow, with full traceability and rollback capability. After implementation, the system monitors effectiveness over a 30- and 90-day horizon, comparing expected and actual results in terms of sales, profit, and OOS levels.
This new functionality does not replace the centralized category management model but complements it, enabling stores to manage assortment directly within the SCM system, supported by data and system recommendations.
It is even more important that the system operates stably within the real-world complexity of the retail network: with different store formats, local characteristics, events, and exceptions.
Automation is only the foundation. Value emerges when the system remains manageable and flexible precisely where standard calculation is no longer sufficient.
This is the direction in which Retano SCM is evolving — as a system that supports retail in managing replenishment within the everyday complexity of the network.concludes Gianni Cassano.
