Chainalytics Helps Global Retailer Improve Seasonal Forecasting and Fill Rate

Chainalytics’ integrated demand and supply planning experts help a global retailer turn around falling service levels

One year after implementing a powerful new forecasting and replenishment technology, the retail business unit of a global multi-channel specialty equipment and merchandise company began to experience rapidly falling order fill rates from its global fulfillment centers to its retail locations. With months-long lead times for replenishment from vendors and the seasonal demand peak already building, on-shelf availability started to dip and backorders started climbing faster than product could be procured. Without intervention and a rapid course correction, the retailer was heading for a financial miss that would make business headlines for all the wrong reasons.

Lower order fulfillment rates can have a multitude of possible causes: operational process inefficiencies; warehouse congestion; workforce turnover and labor; incorrect product mix; supplier shortages; delayed replenishments; lack of inventory visibility, under-forecasting; and more. Yet, with a new technology recently implemented, the supply chain planning organization had quickly come under the microscope of executive management. With sales at risk, temperatures were rising faster than backorders themselves, resulting in finger pointing between sales, planning, operations, distribution, and the technology provider. Seeking a solution, the organization engaged Chainalytics’ Integrated Demand & Supply Planning team to rapidly establish a root cause framework and diagnostic approach, which could quickly help identify the cause of the problem, and then drive visibility and confidence in a mitigation plan that targeted the symptoms at their source.

Working shoulder to shoulder with the retailer’s demand and supply planning leaders, the Chainalytics team compiled a database of sales orders, replenishment receipts, archived forecasts and more, and immediately began analyzing the data vs. the list of possible root causes thru an agile analytics approach as part of the intervention effort. Simultaneously, the Chainalytics team began working with the technology provider to run alternative system configuration simulations and back-testing in search of tuning improvements which might provide improved performance irrespective of the root causes identified.

Despite the highly charged atmosphere and complex multi-party collaboration required, the team had determined the primary root cause of the service issues within a matter of weeks. While the planning technology was functioning well from a technical perspective, the original implementation had failed to fully comprehend and represent key aspects of the retailer’s global business: at the detailed item-level of demand behavior over the year, and with respect to dependent demand and replenishment requirements between global regions.

As is often done when a business has significant seasonality, groupings of items had been defined and assigned to varying seasonality profiles during the technology implementation. The optimal definition of groupings is itself a complex problem; the goal is to find the minimum number of groups to manage which collectively yield the best possible forecasting and planning result by capturing the key trends and behavior of demand over time.

During the initial implementation, an automated routine had been leveraged to define the seasonality groups using an analytics-based approach. However, the exercise lacked sufficient business review and human intelligence. The resulting groupings inadvertently obscured offsetting peaks and valleys, creating a smoothing effect on the forecasting models. The modified demand signal then resulted in over-investment in some items and significant shortages for others. Unfortunately, in the rush to get systems up and running as quickly as possible, taking the time to fully understand detailed business requirements and demand drivers is all too commonly left by the wayside.

The business had also recently moved replenishment planning into a centralized function for the globe and shifted international replenishment to come from the primary warehouse in the U.S., creating new interdependencies between direct and indirect regional demand which were not yet well understood at the time of the implementation. As a result of insufficient change management and education at the time the multi-echelon system was created, an incorrect perception grew presuming the regions outside of the U.S. were “stealing” the inventory which would be needed to support demand within the U.S – and was the primary cause of shortages within the U.S. However, in fact the real root cause was an insufficient representation of the new replenishment dynamics of a multi-echelon system and the propagation of demand up and downstream between the warehouses.

Given that the Director of Planning would now have to ask for significantly more capital than budgeted to then expedite procurement of the right inventory while simultaneously witnessing a spike in backorders, tanking on-shelf availability, and the company’s overinvestment in non-productive inventory, the team was then challenged to identify the optimal supply plan and item replenishment quantities which would maximize the reduction in backorders with the least additional working capital required. The Chainalytics team worked closely with the technology provider to simulate and optimize alternative replenishment plans in a non-production environment which mirrored the retailers’ go-forward configuration after the forecasting and replenishment refinements.

Within four more weeks, the team had formulated an action plan that had been reviewed, refined and blessed by all key stakeholders. The forecasting configuration refinements would be moved into production before the next monthly cycle, monitoring process and dashboards had been set up to track performance more tightly, and an optimized replenishment plan had been created to drive product availability and order fulfillment back to desired levels.

The plan was put into place, and although the peak season for the business had passed by the time service levels were completely restored, the business had effectively managed the crisis through a proactive, data-centric, expert-driven approach which leveraged tight collaboration between the planning organization, the technology provider, and the Chainalytics team.

To learn more about how Chainalytics can help with all your planning challenges, contact us at

In this article