Common RFP Questions: RELEX Forecasting & Replenishment (RELEX F&R)

RELEX Solutions has compiled this reference list of common RFP questions and answers concerning the RELEX Forecasting & Replenishment (RELEX F&R) solution. This page provides detailed information about the F&R software’s modules and capabilities.  

RELEX Forecasting and Replenishment (RELEX F&R) helps retailers, wholesalers, and distributors proactively plan and optimize operations, improving their end-to-end supply chain efficiency, maximizing stock availability and inventory turnover, and reducing waste. RELEX F&R delivers measurable value for more than 350 customers worldwide and provides unique configurability to adapt quickly to any business change. 

1. Q: How does the RELEX forecasting engine work? 

A: In the RELEX F&R solution, the RELEX Forecasting module uses advanced machine learning (ML) algorithms and intelligent data mining techniques to identify demand drivers and calculate the most accurate forecast possible. It also provides seasonality and trend insights, transparency, and automation for better demand planning, replenishment, and supply chain planning decisions. 

1. RELEX uses automation and advanced machine learning models. 

RELEX ML models use multiple data inputs, including: 

• Demand patterns, such as seasonality, trends, and weekday-related variations. 
• Business decisions, such as changes in promotions, pricing, range, display, and space. 
• External factors, such as holidays, weather, local events, and competitor activities. 

Combining demand models, variables, and parameters, RELEX Forecasting calculates forecasts on a selected level of granularity, typically the product-location-day-level. 

2. RELEX handles massive data sets and ensures data quality. 

RELEX’s proprietary in-memory database provides the computational power needed to forecast millions of SKUs on a granular, daily level. It supports calculations for all planning purposes and horizons, from the next day to the next year. 

RELEX F&R also performs automated data cleansing and model regularization to filter out noise and outliers for more reliable daily forecasts. It separates stockouts, outliers, and flagged, one-time events from recurring demand-influencing factors and their estimated impacts. This regularization prevents overfitting and model distortion. 

3. RELEX manages seasonality and step changes. 

RELEX Forecasting automatically identifies seasonality and trends by scanning data and detecting demand patterns on all relevant product hierarchy levels. It applies the appropriate forecasting model to each product-location in the active assortment.  

RELEX F&R also has automatic changepoint and outlier detection to identify significant changes in demand, such as sudden step changes. This capability allows the system to register the impact of demand-influencing data not always captured as explainable variables, such as:  

• Changes in sales presentation and facings.  
• Competitive store closures and openings.  

4. RELEX enables planner review and control. 

RELEX F&R users can interact with the demand forecast easily and apply their expertise by, for example, adjusting and mass-updating at any level of aggregation based on visual monitoring and alert views. The RELEX platform is not a “black box” system, so users have visibility into how demand forecasts are being calculated. This transparency builds user trust in the system and eases adoption. 

5. RELEX automates calculations and measures forecast accuracy for improved outcomes. 

To ensure reliable, quality forecasts, RELEX F&R provides a range of forecast accuracy metrics, including absolute and percentage errors, forecast bias, and others. These metrics allow both the system and users to monitor performance, identify anomalies, and assess the impact of manual adjustments. This high level of automation extends to new item forecasting and day-level demand distribution. 

Together, these RELEX F&R capabilities help retailers get the most out of their data and forecasts for optimized replenishment, increased efficiency and cost savings, and higher customer satisfaction.  

2. Q: What level of forecast aggregation does RELEX support in terms of products, frequency, and timing? Does RELEX F&R support intraday forecasting? 

A: RELEX Forecasting typically generates forecasts on a granular, product-store-day level. These forecasts can be aggregated at all relevant levels for different planning purposes. RELEX allows companies to manually or automatically define which products should be included in the forecast calculation. 

Although forecasts are calculated on a product-store-day level, the RELEX forecasting models are trained using larger pools of data to more accurately assess patterns and trends. For example, seasonal trends could be detected from a pool of products in similar stores. 

RELEX F&R also offers intraday forecasting for special purposes for which day-level forecasting is insufficient, such as: 

• Distribution center (DC) capacity management: Creating forecasts for DCs with capacity limits to ensure smooth goods flow when other measures are not applicable. 
• Workforce planning for store shelving: Using customer, POS, and delivery forecasts to project till shelving labor requirements for workforce optimization. 
• Ultra-fresh products with multiple daily deliveries: Creating demand forecasts for selected, in-store assortments of highly perishable items with deliveries throughout the day. 

3. Q: Does RELEX Forecasting & Replenishment support forecasting and planning across different time horizons? 

A: Yes, RELEX F&R supports optimized planning decisions across different time horizons in both forecasting and replenishment processes. 

Time horizon management in forecasting 

For instance, the forecast horizon is defined based on actual need and can be freely configured in the system. The system supports multiple time dimensions: hours, days, weeks, months, and years. For replenishment use, the horizon is typically set at six months during implementation. For long-term seasonal planning, users can set longer horizons (the longest horizon supported is usually 36 months). RELEX F&R can automatically change the forecast horizon based on lead-times or other planning horizons.  

Time horizon management in replenishment 

For automated ordering, RELEX Replenishment considers actual future order needs, not just the forecast. Actual future order needs, or “demand projections,” refer to how many units planners will need to order, as opposed to “forecasts” which refer to how many units customers are expected to buy. The horizons for these types of demand projections can be calculated long-term and can incorporate current inventories in downstream warehouses and stores.  

RELEX F&R also calculates projected order proposals, deliveries, and inventory balances. For products with limited shelf-life, RELEX calculates projected spoilage. 

Forecasting and projected order proposals are supported at practically any defined time horizon. The projections are derived from day-item-location level simulations, based on the current demand forecast, current inventory balance, open orders and their estimated delivery dates, and relevant replenishment parameters (such as order calendar, safety/presentation stock target, and order batches). 

RELEX’s approach to planning and forecasting makes it easier to predict when orders will need to be placed and how many units each location will need. This approach provides more accurate, actionable information for planning across time horizons.  

4. Q: What functions does RELEX F&R perform automatically, and how much power do users have to manually review and adjust plans? 

A: RELEX Forecasting & Replenishment provides the optimal balance of automation and human oversight, automating calculations, applying changes across appropriate products, hierarchies, and parameters, and measuring success to improve outcomes. 

Automation and user adjustments in RELEX Forecasting 

The RELEX Forecasting module is designed to be robust and self-correcting, minimizing the need for user modifications to model configurations. At the same time, RELEX F&R allows users to manage and edit forecasts at any level of aggregation or detail, with the aid of graphs and data tables. For instance, RELEX users can define custom time periods and then view and edit the forecast for that period at different hierarchy levels. They also can update forecasts by using absolute quantities or by applying logical operators (in other words, adding, subtracting, or multiplying values) relative to the current forecast. 

RELEX Forecasting also makes it easy to automatically apply changes at different aggregate levels. For instance, let’s say demand has increased for an item, and the planners adjust the overall forecast to reflect that increase. Forecasts that are mass updated at a higher level of aggregation are automatically disaggregated to the product-location group level. RELEX ensures the increase is split across stores proportionally by using the previous forecast’s distribution of units to calculate the percentage of demand each product-store should receive. Edits to monthly forecasts are similarly distributed to weekly forecasts. 

RELEX Forecasting & Replenishment supports the use of forecast accuracy and error metrics (such as MAD, WAPE, RMSE, and bias), which can be measured at any aggregation level for any time period. These metrics can populate exception management views and trigger alerts for users to review and act on when necessary. 

Automation and user adjustments in RELEX Replenishment 

The RELEX Replenishment module enables both automatic and manual adjustments to replenishment models and parameters. The RELEX system supports several different replenishment methods, including ROP, MRP, CO-MRP, cross-docking, and others. 

For most products, replenishment parameters are set and optimized automatically. RELEX’s automatic optimization most accurately calculates suitable parameters for a product-location, based on historical sales or forecast errors. Any exceptions can be handled separately with customer-specific rules or set manually.  

5. Q: How does RELEX F&R support planning across product lifecycles, including new product introductions and phase-out stages? 

A: RELEX F&R provides lifecycle management functionalities and parameters that manage product introductions and ramp-downs in demand forecasting, order projections, and replenishment. The system classifies products flexibly, maintaining data and optimizing plans according to product-specific attributes, such as lifecycle phase, introduction and termination dates, and reference products. It automatically forecasts phase-in and phase-out items, using group-based heuristics to help manage multiple, simultaneous product introductions. 

How RELEX F&R handles replacements and new products 

When a product is a replacement for an existing item, the system incorporates the sales history of the replaced product into the new product’s forecast.  

For new product introductions with no sales history, RELEX Forecasting automatically suggests appropriate reference products with relevant sales history that can be used to generate initial forecasts. Typically, the new product’s base demand is taken from the reference product, while other characteristics like seasonality and weekday patterns are mapped from product group aggregates. The automatic reference logic considers attributes such as product group and price range, but users can also configure criteria based on available master data. They can also maintain and mass-update data on different levels.  

When better demand information becomes available, RELEX users can manually override the reference forecast, but historically, the automatic process has proven to be far more accurate than manual forecasts or manual reference picking. 

How RELEX F&R handles phase-out items 

RELEX Replenishment provides product lifecycle management functionalities and parameters that effectively manage product ramp-down.  

RELEX Replenishment uses introduction and termination dates to set the replenishment time period. Items can also be temporarily added to an inactive assortment to determine whether they are replenished during this time period. 

RELEX Replenishment automatically reduces purchase orders and stock levels for products to be terminated. The system uses defined phase-out periods to reduce the stock to zero by the termination date. It supports end-of-season clearance by pushing out the remaining inventory of seasonal goods at the DC-level to stores based on seasonal sales or future forecasts, accounting for on-hand inventory and potential open orders. 

6. Q: How does RELEX Replenishment generate and manage order proposals? 

A: Within RELEX F&R, RELEX Replenishment generates daily order proposals based on accurate product-location level forecasts, using forecasting and order calculation algorithms. RELEX can also calculate order proposals more than once a day. 

Optimal order quantities are proposed based on calculated order need, accounting for:  

• Optimized safety stock levels and the current inventory situation. 
• Demand forecasts. 
• Open orders. 
• Lead times, batch sizes, and delivery schedules. 
• Replenishment parameters and business rules or limits.  
• Additional factors, such as spoilage, shelf presentation targets, space or capacity limitations, and supplier delivery restrictions. 

RELEX has multiple ordering models, which are automatically assigned to product-locations based on business rules. These ordering models include: 

• Forecast-based ordering. 
• Internal transfers. 
• Coordinated ordering. 
• Purchase-to-order. 
• Cross-docking. 

These ordering models and business rules allow retailers to flexibly manage and account for all ordering requirements. 

7. Q: Does RELEX Replenishment support automatic static and dynamic safety stock calculations and management?  

A: Yes, RELEX Replenishment automatically optimizes safety stock by managing availability targets with flexible rules that can be configured to match business needs. Many of RELEX’s customers set availability targets using a combination of ABC (sales volume-based) classifications and XYZ (sales frequency-based) classifications. 

RELEX F&R uses four safety stock management methodologies: 

1. Static safety stock. 
2. Dynamic safety stock. 
3. Day-level safety stock. 
4. Forecast-driven safety stock. 

Static safety stock methodology in RELEX Replenishment 

“Static safety stock” refers to a specific quantity or number of units of safety stock. The static approach works best for items with very stable demand. It automatically optimizes safety stock at the SKU-store level based on the cycle service level or fill rate target, accounting for delivery schedules, lead time, demand variability, forecast error, batch size, and minimum order quantities.  

Dynamic safety stock methodology in RELEX Replenishment 

“Dynamic safety stock” refers to the number of days, not units, of supply needed. It is often used for seasonal products or items with fluctuating demand patterns. The system can automatically increase safety stock at the beginning of the season and decrease it towards the end, accounting for factors like maximum lead time. 

Day-level safety stock methodology in RELEX Replenishment 

“Day-level safety stock” refers to safety stock parameters that can be dynamically changed and set for specific days of the week. This approach is useful for situations with significant day-to-day operational differences. For example, if a store is closed on Sundays, planners want to run the stock close to zero by the end of each Saturday, while ensuring there is enough ordered for Monday’s delivery to prepare for the upcoming week. Setting the safety stock parameter according to weekday helps planners control inventory levels. 

Forecast-driven safety stock methodology in RELEX Replenishment 

“Forecast-driven safety stock” refers to an adaptive order parameter that uses a safety stock optimization model based on forecasts instead of past sales. In this scenario, RELEX uses the forecast to estimate the true standard deviation of demand, or in other words, how significantly demand is likely to fluctuate. Adaptive safety stocks are calculated on a daily level so that the forecast, using wider demand signals, can alert planners to future demand increases and the need for more safety stock, even if previous sales history does not indicate a potential demand increase.  

This forecast-driven adaptability helps planners react to seasonal changes more quickly and accurately, ensuring better availability during seasonal peaks and lower excess stock at the end of the season. 

8. Q: How does RELEX Replenishment handle exceptions and workflows? What is the exception management process? 

A: RELEX Replenishment allows companies to tailor their exception management process to business-specific requirements or changing planning needs. 

Creating rules for automated processes in RELEX Replenishment 

RELEX’s proactive exception management allows users to easily create automated rules customized to specific metrics, parameters, and operational timelines. These rules effectively highlight potential issues, such as spoilage risk, projected stockouts, and purchase order delays, which might impact forecasting, replenishment, and allocations. The system can perform daily recalculations to provide the most current insights into critical concerns. 

Daily user workflows in RELEX Replenishment 

Typically, only a small fraction of the order proposals created by RELEX F&R needs to be reviewed. It is not unusual for teams to accept all proposals automatically and focus on performance monitoring by exception. Users can review and accept order proposals flexibly in the user interface.  

Exceptions usually include: 

• Valuable order lines that exceed a trigger threshold. 
• Order lines for items with ongoing or planned promotions. 
• Order lines for items with current or projected stock-outs. 
• Order lines for new or discontinued products. 

In typical use cases, order proposals are checked and approved in the RELEX system before they are sent to the customer’s back-end system and converted into actual purchase orders. Alternatively, order proposals can be sent to the customer’s back-end system after the calculation has finished, either for review or to be converted into purchase requisitions or purchase orders. 

9. Q: Can RELEX Replenishment create time-phased order plans? 

A: Yes, RELEX Replenishment can generate time-phased order plans (detailed schedules indicating when orders should be placed and in what quantities), based on forecasted demand, inventory levels, lead times, and other relevant factors. 

Constrained replenishment in RELEX F&R 

Constrained projections are an extension of normal projection logic and MRP order proposal calculation logic.  

In addition to calculating projected order proposals, RELEX Replenishment includes a constrained replenishment function that incorporates information about supply constraints into order proposals and projections. This information includes minimum constraints, such as warehouse force-outs, and maximum constraints, such as scarcity in warehouses and container constraints. RELEX uses these constraints to make projections and order proposals more reflective of supply realities as opposed to assuming infinite supply.  

Constrained replenishment uses “constraint groups” that match product-locations to the limits defined for that group. Product-locations can be connected to groups manually via the UI field “constraint groups” at the product-location level, or they can be automatically created by pre-configured workflows.  

Constraint groups can be assigned static or dynamic limits. Static limits are fixed amounts set for specific dates. For instance, a location may be able to receive 500 units on March 15^th. On the other hand, dynamic limits automatically adjust as orders are placed. For instance, if 100 units are ordered today, tomorrow’s available limit shrinks by 100. 

10. Q: Does RELEX F&R support multi-echelon inventory planning?  

A: Yes, RELEX Forecasting & Replenishment supports multi-echelon inventory planning for retailers with a multi-tier supply chain setup. Tiers include everything from regional distribution centers to e-commerce channels, customers, customer groups, and more. 

RELEX improves forecasting, replenishment, and allocation across tiers by using store-level data to drive DC-level replenishment decisions. The solution uses forecasts and replenishment parameters to help calculate projected store orders per product, store, and day, several months to a year in advance. These order projections consolidate data on stores’ current inventory, safety stocks, visual minimums, delivery schedules, and any planned inventory movements, like building promotional displays or shifting orders to reduce capacity bottlenecks. The system aggregates order projections across all stores supplied by a distribution center to form an accurate, customer-driven forecast for that DC. 

RELEX’s multi-echelon inventory planning improves supply chain transparency, capacity planning, supplier collaboration, and logistics like cross-docking, pick-to-zero, and shortage situations.  

11. Q: Does RELEX F&R use AI? 

A: Yes, RELEX F&R uses multiple kinds of AI, including specialized AI algorithms and generative AI. RELEX is also developing and launching AI agents that will work collaboratively with human planners, analyzing data, making recommendations, and taking autonomous actions under human supervision and control.  

How does Specialized AI work in RELEX F&R? 

RELEX F&R’s machine learning-based forecast models analyze enormous swaths of data including historical sales, weather patterns, and local events to generate precise SKU-level forecasts. These forecasts drive automated replenishment recommendations and help reduce waste, improve availability, and ensure efficient supply chain operations. 

What generative AI capabilities does RELEX F&R offer? 

RELEX F&R, and every module available on the RELEX platform, gives users access to Rebot, RELEX’s generative AI assistant.  

Rebot was the first retail and supply chain planning AI assistant on the market, debuting in 2023. Since then, it has been adopted by RELEX customers and employees around the world. Rebot provides on-demand support, using a vast, machine-readable knowledge base of documented supply chain expertise and solution information. Rebot enables companies to speed up onboarding, upskill users, troubleshoot issues quickly and easily, and get the most out of their RELEX solution. 

What AI agents are available with RELEX F&R? 

RELEX Diagnostics is one of the first modules on the RELEX platform to incorporate AI agents.  

Event before the advent of agentic AI, RELEX Diagnostics was helping retailers better understand their inventory performance, using RELEX F&R data such as order proposals and inventory levels to help determine the root causes of stockouts, spoilage, and excess inventory. These insights help users take proactive actions to minimize lost sales, increase margins, and cut waste tied to overstocking and spoilage.  

Now, Diagnostics also offers AI agents and gen AI capabilities that automatically analyze operational data, identify the root causes of stockouts, spoilage, or inventory imbalances, and recommend corrective action. Future agents will be able to act on those recommendations autonomously. 

12. Q: How do I submit an RFP to RELEX? 

A: You can contact RELEX or request a demo using this form: https://www.relexsolutions.com/demo-request/.