Predictive intelligence helps CFOs forecast demand: A new form of software can help finance executives get a window on the actions and decisions that can improve demand--and the bottom line.

• Author: Mahaffy, Dale
• Position: Forecasting

Combine the state of the economy and the ever-increasing operational role the CFO plays in today's corporations, and it is crystal clear that accurately predicting demand and profitably creating demand are crucial to success. CFOs stick their necks out each quarter to give investors and boards of directors not just the facts about the company's past or its current performance, but projections on where the company is headed.

The planning process is inherently collaborative. To date, the CFO has had to predict demand by relying on individual spreadsheets -- highly summarized projections and plans from various managers representing individual business units, geo graphical regions and lines of business, and based on disparate information resources within the firm. The aggregated spreadsheets become the annual operating plan and are updated regularly.

Because of the highly aggregated nature of the information, however, most of the details get lost in a sea of averages -- resulting in tremendous volatility and uncertainty as the plans are executed on a granular level -- ultimately causing gaps in the planning process. One example is that hundreds of millions of dollars of lost opportunity occur because out-ofstocks at mass merchandisers run at 14 percent or more on promoted items. In today's climate, an organization cannot afford to lose revenue, and potentially customers, by running out of promotional products at the point of purchase.

In an attempt to better control the demand planning process, most organizations develop in parallel a bottom-up plan based on weekly sales forecasts, taking inputs from sales and field representatives. As a result, companies spend valuable time and resources combining projected revenue from merchandisers and actual sales data to update the operational plan. Yet, they lack the ability to provide real-time insight on all of the demand information by product, customer region and store -- insight that facilitates strategic planning and reveals day-to-day adjustments of plans not captured in pure statistical models.

In addition, the desktop spreadsheets currently used in finance and planning departments are usually not even using the best data available -- the data in the data warehouse -- to make projections. The data warehouse holds the standards for the product, geography, time hierarchy, rules and workflow, as well as definitions for measures and calculations like activity-based costing, economic order quantity...