Forecasting: Define, Enable, Measure
Based on the preceding research and discussion with companies, it is evident that working effectively in uncertain times requires a robust operating model that integrates key processes and systems on the demand-side of the business. These are typically siloed processes that address sales forecasting, marketing campaign management, channel management, etc.
Figure 5: Shift from passive/reactive to active/predictive demand management
There is a need to shift from passive/reactive demand management to demand management that is more active/predictive:
- Sense the demand of customers early and correctly
- Influence the demand to favorably align it to capability
- Budget for variability in demand during fulfillment responsiveness
- Focus on innovation to realize first mover advantage during the short life cycle of the product.
The following operational steps could possibly form a starting point to evolve functional silos and fragmented processes to active/predictive demand management:
Streamline Information Gathering and Analysis
- Streamline the master data of products, customers and suppliers
- Establish the mapping of unique product definitions to that of customers and suppliers
- Establish a correct data hierarchy to ensure the uniform roll up of forecast information across customer segments, product part numbers and geographies
- Determine the approach on governance of data to be followed uniformly across the company. This includes streamlining sources for data imports, methods of data collection, data cleansing and data finalization guidelines
- Define methods and rules to utilize informal, qualitative information to supplement quantitative information
- Design and implement information exchange mechanism based on above need
- Evaluate options of reward/penalty mechanisms with customers based on accuracy and comprehensiveness of forecast data
Formalize Forecasting Processes
- Identify the forecasting parameters, such as forecast granularity, forecast frequency, forecast horizon, etc., for different product families and customer segments that should be followed uniformly by different departments of the organizations.
- Define processes for:
— centralization of input data obtained from different sources
— periodic refresh and leveraging of customer segmentation during analysis of forecast inputs
- Establish guidelines that promote multi-tier visibility of forecast information. Identify the recipients of the information and information sharing process
- Define processes and methods to analyze the accuracy of previous forecast based on current data.
- Institutionalize specific set of metrics across the organization to report forecast performance
- Review previous forecasts to validate the forecast rules and assumptions
- Establish the parameters for reward against forecast performance
Leverage Forecasting Tools
- Conduct maturity analysis of the business processes and requirements to ensure the correct forecasting tool is selected
- Assess data readiness in accepting the new tool
- Build a robust information mechanism to automate data capture
- Formalize a process of acceptance of the new forecasting tool by the users through structured change management programs. Support user transition through adequate training. Institutionalize appropriate rewards for transition