Stop Me before I Forecast Again!

From the perspective of the FP&A group trying to support the forecast, there are three major mechanical faults with the process:

- There are too many line items to be forecasted efficiently and effectively
- The mechanics are too simplistic, too naïve (i.e. a single-driver, Cost-per-Person model)
- There is no validation that even this one, single driver is the best or right one to use

To briefly expand on each of these items:

- You are probably budgeting and forecasting about 20 line items per cost/profit center, times 100 cost centers per division, times however many business units you have – THOUSANDS of line items to forecast, each and every quarter/month.

- You base 90% of your forecasted line items on one driver: HEADCOUNT. While using just one driver might be accurate enough for a handful of line items, most are probably best modeled with a combination of weighted drivers. Travel might be a function of headcount, function (i.e. sales versus R&D), and level within department (manager versus individual contributor). Sales might be a very complicated function of GDP, marketing spend, customer satisfaction, # of stores and channel partners, and # of sales reps with more than 9 months in the account/territory (i.e. accounting for training time and sales cycle ramp-up). Cost of Goods might be a weighted average of revenue, headcount, inflation and fuel/commodity costs.

- Even if you agree to a multi-driver approach, how would you know which ones are the right ones? Which ones are significantly correlated and have a validated cause-and-effect relationship? And what’s the right weighting: 60/40, 40/60, or 40/40 with a 20% fixed component?

Further, it’s tough enough managing a single headcount-driven model in a spreadsheet, let alone trying to incorporate multiple weighted drivers, test and validate them, and then apply that on a monthly basis to thousands of line items.

The answer is to stop forecasting individual line items directly. Instead, focus on forecasting your key business drivers.

Once you've done the analysis, you’ll probably find that you have about a dozen key business drivers that are significantly correlated with your revenue, expense and staffing numbers. These might be:

- GDP, Unemployment, Interest rates, currency rates, CPI, demographics
- Third-party industry forecasts, channel partners, trained sales reps
- Headcount (type/function/level), fuel/commodity costs
- Capital spend, R&D spend, marketing spend, sales rep training spend
- # of stores/branches, customer sat scores/churn, employee turnover

You might initially test three dozen possible variables before trimming the list down to the dozen or so statistically significant drivers that will become the focus of your forecasting efforts. Analytical tools can evaluate all of these candidates, eliminate the ones that have no bearing on your results, assign the right variables to the right line items, apply the appropriate time lags, and determine the parameters/ relative weightings for each driver in each situation. Automatically. If the MBA’s on your staff can remember the difference between a dependent and an independent variable from the one statistics course they took, the analytical tools can do the rest, leaving your staff with the time to focus on the more important task of accurately forecasting the DRIVERS that the rest of the financial forecast can be built on.

Furthermore, once your staff starts talking the language of “drivers” rather than “P&L line items”, they put themselves in a position to work with line business management in a value-add, decision support manner. The language of lines items is all after-the-fact and reactive; the language of drivers is proactive, it means they can direct discussions about the impacts to the business and what effective financial and operational levers management has available in its tool kit to impact results before-the-fact.

Have you been wondering how to turn your FP&A staff from being primarily transaction oriented to being primarily decision support focused? Been wondering how you can tap into that value-add potential that you know exists within your highly trained and educated staff? Business analytics and analytically-based forecasting might be just what you’ve been looking for.