The current state of forecasting methods is based on rear-view mirror-history approaches. Most of the market has these types of solutions. But companies know that if they could have more predictive and real-time data they can improve their forecast accuracy or even make strategic adjustment to the business to avoid downsides. Methods are evolving to address better forecasting.
The easiest way to illustrate this evolution is by way of a driving analogy: Consider the three points of visibility - the rear-view mirror; your high-beam lights, and your low-beam lights - to help navigate your business:
The rear-view mirror"”popular methods used in forecasting, like time-series analysis, rely upon sales history. This is like driving by looking at the rear-view mirror. And in a world where the past is less an indicator of future behavior, this is likely to drive up forecast error. But nevertheless, such methods (like attribute forecasting, time series, event and seasonality, attach rate forecasting) constitute an important first step of the forecasting process to establish the "baseline" view.
Aligning with management targets is the high-beam view"”Management goals and targets as well as inputs from across the organization - finance, marketing and sales -are used to make adjustments. The methods applicable in this step (like consensus forecasting, revenue sensitivity analysis, price forecasting) are often part of the broader sales and operations planning process to help reconcile the different views and set the quarterly targets that the supply chain must execute towards.
Adapting to near-term demand volatility is the low-beam view"”This is a major area of opportunity for companies to leverage social technologies. This type of "fine-tuning" is another key differentiator of B2B forecasting because there is potentially a large amount of account level intelligence that can be captured directly from customers, channel partners such as resellers and distributors, or by sales representatives. For example, backlog methods, current actual inventory and its "state" (on allocation, nearing expiration or end of life) help analyze the order pipeline for short-term trends - changes in the demand pattern.
Combining these views with sensitivity analysis and much richer data sources from collaborative clouds, near-term forecasting techniques are gaining an audience in the corporation. As users succeed at rear-view and high-beam dialogue, they will seek finer-tuned solutions and processes. Getting great at planning is a continuous process. So new solutions in demand management will continue to be purchased to complement or replace older techniques.
New sources of information provide great promise. 2013 will see the search for solutions that can make sense of these. However, challenges remain - the data is not in a pristine format that can easily be analyzed. Therefore, our focus on mobile, social and big data solutions needs to be paired with the development of methods from the technology provider before users can gain full value of these new ideas.
Keywords: demand volatility, forecast accuracy, aligning forecasts and business strategy, sourcing and procurement needs, demand data, value chain IT, supply chain management IT, supply chain management
Enjoy curated articles directly to your inbox.