One of the largest food distributors in the Middle East found it difficult to accurately forecast demand. This led to inaccurate orders to their suppliers and reduced profits because they couldn’t fulfill customer demand across their distribution network. Their forecasting, based in Excel, was lengthy and manual, which impacted the quality of their forecasts and forecasting cycle times. The manual forecasts were at a high level and did not incorporate SKU/customer combinations.

The food company’s challenges included:

  • the accuracy of demand forecasts
  • long lead times on order placement
  • limited functionality in Excel

The food distributor’s forecast accuracy suffered because orders were typically placed with their suppliers for a one-year period with only minor changes to the agreed-to delivery schedules, which were based on stock on hand. The forecast was also hindered because it was difficult to accommodate size and complexity differences between customers of different types, such as chains and independent retailers.

The distributor used promotions to increase sales and clear inventory that was near the expiration date, but their basic, less sophisticated forecasts made it difficult to predict the impact of product promotions. These challenges made it difficult to set optimal inventory levels, which increased the risk of excess stock, shortages of product, lower fill rates, lower service levels, and growing costs associated with products past their expiration dates.

The company approached CT Global with their problem and we suggested a solution built on SAS DDPO.

A SAS DDPO solution blows away Excel forecasts

CT GLobal’s DDPO solution used an automated method to select the best statistical model to arrive at the most accurate forecast. Our SAS solution allowed the company to determine the impact of product promotions by SKU on forecast demand and allowed users to create “what if’ demand scenarios. It also included a calendar that changed every month based on religious holidays and the lunar calendar.

Our DDPO solution also featured:

  • A champion model selected from 64 possible models
  • Forecasts for each SKU/customer combination
  • New product forecasting
  • Product Phase in and Phase out forecasts
  • Ability to update forecasts with comments
  • Consensus planning using a collaborative method that integrates statistical baselines with business judgment to finalize the forecast
  • Workflow documentation and approvals for changes in the forecast
  • Inventory optimization based on the highly accurate consensus forecast
  • Optimized supply based on service level, warehouse capacity, supplier capacity, transportation costs, and more
  • Optimization models including constraints such as required lead times, costs, and targeted service levels
  • Ability to view inventory level/out of stock position based on forecasted demand, orders placed, expiration dates, and expected receipt dates
  • Roll up from orders placed to fill containers, pallet loads and MOQ’s

After implementation, the food company’s forecast accuracy improved with low MAPE scores. The SAS DDPO forecasts helped them reduce costs through improved stock management and stock expiration; manual efforts were reduced; forecasting and consensus planning were faster. The company achieved higher profits based on lower inventory costs, optimal inventory levels, and the attainment of target fill rates and service levels.