Companies still struggle with earnings and sales projections, but determined action can yield big improvements in the forecasting process.

For CFOs, the consequences of inaccurate forecasting go far beyond a plunging stock price and unhappy shareholders. A finance chief may be able to get away with one way-off-base earnings projection, but two is pushing it. "Very often you see that after, let's say, a second profit warning, companies fire the CFO because credibility is totally lost at that point," says Fritz Roemer, who leads The Hackett Group's enterprise performance management executive advisory program.

Yet for all the pressure on finance leaders to get the guidance right, companies still do a dismal job of forecasting quarterly earnings. Hackett defines an accurate forecast as one that falls within 5 percent of actual results. Only one in three companies hits that mark, according to the global strategic advisory firm's new Book of Numbers research, which includes an analysis of results from more than 70 large U.S. and European organizations.
And businesses are not much better at forecasting sales; less than half of the companies Hackett studied managed to produce an accurate sales forecast for the upcoming quarter.

Is ditching the quarterly earnings forecast in favor of "long-term guidance" a realistic solution? "Maybe that's one possibility, to say 'I won't do that anymore,' " notes Roemer. "On the other hand, the pressure coming from the Street is high, and [analysts] will still ask the questions: 'You just informed us about performance in the last quarter and maybe the performance was good -- that's good news, but I also want to know what will happen in the next quarter.' And, of course, even companies that don't provide quarterly public guidance are still on the hook for internal forecasting.

Hackett recommends the following action steps for companies that want to build a world-class forecasting capability:

• Align the forecasting process with business risk and market dynamics. Traditional approaches often fail to take into account a key factor in determining forecasting frequency: the organization's risk profile. "Around 50 percent of companies are what I call 'accounting driven' -- they take an approach that's strongly influenced by an accounting mindset rather than business needs," Roemer reports. "It makes a huge difference if you are operating in a highly volatile, highly risky market or a relatively stable market, so the forecasting process should reflect that."

Businesses that operate in a high-risk business environment should forecast more frequently, but that doesn't always happen. "We find a misalignment in many, many companies," says Roemer.

CFOs also should adjust their forecasting horizon as market conditions change. Fourteen percent of the organizations in Hackett's sample described themselves as operating in high risk/high volatility markets, a seven-fold increase from just three years ago. The firm expects that to increase by nearly 50 percent over the next two years.

• Consider implementing a rolling forecast. One way to match forecasts more closely to market dynamics is by moving to a rolling forecast. "The accountant's view is the fiscal year, and this is why, historically, companies had a year-end forecast," says Roemer. "But the process should not be accounting driven; it should be business driven. And the business doesn't end at the end of the fiscal year. If you talk to someone in sales or production, for example, they don't have the December 31 wall which people usually have in accounting."

Currently around 40 percent of large companies in Europe use a rolling forecast. The technique is less popular in the United States, where it has been adopted by around one-third of large organizations. And that percentage hasn't changed much since 2004, according to Hackett.

• Set accuracy targets for forecasting. Most organizations measure the accuracy of their forecasts against actual results, but surprisingly few -- just one in five of Hackett's sample -- set accuracy targets. Doing so can help companies identify sources of forecast bias and change the behavior of forecasters. "Maybe a person is very optimistic, or very pessimistic," says Roemer. "Or maybe what's behind the forecast bias is the company's philosophy or culture. What we often see is the effects of the culture of 'underpromise and overdeliver,' for example."

But whatever the source of forecast bias, Roemer adds, "if you make it transparent and you manage it, you can more or less eliminate it." And that should go a long way toward keeping the CEO, the shareholders, and the Street happy.