Attribute Based Planning for the Consumer Electronics Industry
The sheer number of options available in consumer electronics products can be overwhelming as many in the United States likely experience every holiday and especially during the Thanksgiving holiday as they compared Black Friday and Cyber Monday deals. Televisions are a prime example as the technology has come a long way in the last 10 years and manufacturers jockey for differentiation. LED versus OLED; flat screens versus curved; edge lighting versus rear; 4K, Ultra High Definition, High Dynamic Range, 3D, the list goes on. All these options made me think about how difficult it must be to forecast demand for ever-changing consumer electronic product lines and the potential advantages of using attribute-based planning to manage the SKU-proliferation this industry faces.
Attribute-based modeling uses demand profiles in place of direct historical sales data. Demand profiles are built from historical sell-in and sell-through data reflecting a variety of demand and seasonal attributes. A particular models forecast is patterned after the profile that best characterizes the product. As actual demand information is captured, the profile is validated or alternate profiles identified to dynamically adjust the product’s forecast.
Attribute-based modeling consists of three unique steps:
- Create Demand Profiles: Creation of demand profiles is based on mathematical concepts known as Chi-squared analysis. The demand planner selects products to be included based on attributes such as color, type, region, etc. Multiple attributes can be applied simultaneously.
- Assign Demand Profiles: New products can now be assigned to Demand Profiles. It’s vital that planners be able to set criteria for how a new product’s attributes must match the attributes of a demand profile.
- Assess Accuracy of Demand Profile Based on Demand Signals: In a perfect world, demand profiles, once set, would never need adjusting—new products would sell just like other products with similar attributes. In the real world, POS and other demand signal feedback should be used to assess how well the current demand profile fits versus other demand profiles that have similar attributes. Relative-Error-Index (REI) calculations let planners see which demand profile has the most accurate fit based on current demand trends. The original demand profile may be swapped out for the profile that has the lowest REI attributes.
Supply chain organizations routinely rank forecast accuracy as a major obstacle in meeting their goals. Accurate forecasts are the foundation for profitable business growth. Creating accurate forecasts requires comprehensive modeling capabilities plus the flexibility and ease-of-use to shift methods as life cycles progress and market conditions change.
- White Paper: Eight Methods That Improve Forecasting Accuracy
- White Paper: Winning the Inventory Battle in Consumer Electronics