Next-level technologies like artificial intelligence and machine learning, coupled with access to big data, have given the chief financial officer superpowers. The problem: many CFOs are not using them.
Real-time data analytics allow finance teams to gain deeper knowledge of operations, risks, sources of efficiencies, and potential new business models, to name a few benefits. But in terms of forecasting sales revenue, perhaps the most important number in financial planning and analysis (FP&A), finance chiefs aren’t always leveraging the data and technology they have.
Traditionally, sales revenue forecasting has been a highly manual process. Explains Philip Peck, vice president of advisory services and finance transformation at Peloton Consulting Group: “People gather, compile, and manipulate data often within an array of fragmented working Excel spreadsheets and workbooks. Data comes from many disconnected source systems.”
With more and more data available, revenue forecasting this way becomes unwieldy and time-consuming. “The power of the analytics platforms, the power of the datasets, and the tools that we now have to crunch that information have the potential to produce a revenue forecast that is dramatically better than not too long ago,” says Peck.
Indeed, 74% of companies surveyed by Aberdeen Research in 2018 produced more accurate forecasts as a result of using sales planning analytics. Survey respondents using planning analytics also had a higher percentage of sales reps meeting quota and a higher percentage of opportunities won.
A significant number of organizations have developed unsophisticated forecasting habits. The Duke University/CFO Global Business Outlook Survey in the fourth quarter of 2019 found that for almost half (48%) of global respondents, future planning relies heavily on recent historical performance. Only one in five used internal models to forecast sales. And roughly twice as many companies took a bottom-up approach (i.e., a sum of divisional forecasts) rather than a sales outlook that originated from top management.
The problem with relying solely on a bottom-up approach, explains Peck, is that while the CFO has to sign off on the enterprise-wide numbers, he or she doesn’t necessarily have good visibility into all of the forecast’s underlying assumptions.
“Starting from the lowest level of granularity, the forecast could go through multiple iterations; there could be elements of conservatism or perhaps optimism throughout all the different cycles; and the CFO may have to put a lot of judgment into what may not have been a well-integrated, end-to-end process,” Peck says.
A top-down approach can give management a broader picture of revenue potential and help it identify sales patterns. Of course, there are downsides to topdown: a bottom-up approach enlists the participation of employees and managers in the process. That can instill a greater sense of the importance of meeting targets and force wiser spending decisions.
Why is sales forecasting so important these days? When sales forecasts miss their mark, the consequences are far-reaching. The December Duke/CFO Outlook survey found that when global organizations bungle sales forecasts, about 40% adjust their hiring plans. More than a third of finance executives responding (36%) said they revise spending on inventory and advertising, and 25% said they alter investment plans or production schedules. (See “Hard Target,” below.)
“We see some significant implications when sales forecasts aren’t met,” says John Graham, a finance professor at Duke’s Fuqua School of Business. “This says to me that as companies continue to adopt advanced analytics and other leading-edge technologies, we’re likely to see finance play a bigger role than what the current data suggests.”
For the finance chief, it will mean working more closely with the sales organization. Today, says Peloton’s Peck, companies are elevating the role of the CFO and the finance team in support of revenue forecasting. The era of hype and aspiration for what new technologies and big data can achieve is ending; it’s now the era of putting in place legitimate execution plans.
“As these plans pan out,” he adds, “we’ll see far more collaboration in a constructive and aligned way between the CFO, the finance team, and sales and marketing.”
In some organizations, sales forecasts are so crucial that the finance chief will take steps to increase ownership of the process.
At Tente Casters North America, a division of a German manufacturing conglomerate, CFO Pierce Kohls took over responsibility for the company’s customer relationship management (CRM) system. That meant, among other things, ensuring the CRM was properly populated and overseeing the analysis of the data. The move was prompted by the need to tighten the company’s long-run sales forecasts and better capitalize on customer insights.
Tente’s one-to-three-month sales forecast was fairly accurate, as is true for most companies that operate on a made-to-order sales process, explains Kohls. However, “once you’re looking at the six-month range and further out, that’s where the challenges are; that’s where sales forecasting becomes even more important; and that’s where the true value of understanding the data comes into play.”
Kohls’ rationale for taking charge of the CRM was, first, his FP&A background, and, second, the changing nature of Tente’s business.
“I think finance professionals are more data-driven as a rule, and that was kind of my argument as to why I felt I should take over the CRM,” Kohls says. “I recognized the value of data and what it brought to the table.” By that, Kohls says, he means not just recognizing that the company had large amounts of data but being able to identify key insights and translating those into actions.
For Tente, those insights were critical — like for many other manufacturing companies, forecasting errors have broad implications. “If you overestimate you end up with a lot of inventory sitting around, and that’s a big problem. Sitting on inventory means tying up a lot of capital that you could be investing in other areas of the business.” If you underestimate sales, on the other hand, “you run out of stock and run the risk of your customers going to your competitors.”
Taking over the CRM also made sense to Kohls because the nature of Tente’s business was changing, requiring more involvement from the CFO in the sales process. The company has gone from being just a commodity provider to more of a solution-based seller. That means more questions from sales managers that have to be deferred to the CFO. High-value custom sales opportunities, in addition, mean capital investment Kohls says.
Kohls’ heavy involvement meant a shift in the dynamic between the sales team and finance. But the overall result was well received, Kohls says. “The sales team actually liked my increased involvement in the sales process because they could get answers a lot quicker, potentially shortening the sales cycle.”
However, Kohls emphasizes, collaboration was crucial. He took the trouble to marry the insights he gained from his business intelligence tool (which is connected to the CRM and Tente’s enterprise resource planning system) with those he collected from sales in the field. “It’s not just about the CFO having access to the data and making forecasting decisions in an ivory tower,” he says.
The best approaches to forecasting combine the qualitative and quantitative, he says. “I get information from my salespeople in their territories and combine it with the insights I generate through [Microsoft] Power BI.” Kohls is also a big supporter of data democratization, which basically means providing wide access to the data.
“I want us all looking at the same data, the same metrics. I want us all analyzing the same things,” he says. “As a CFO, you have to actually show the sales team so they can trust the data.”
Analytics isn’t a silver bullet, of course — the right inputs are crucial. Improving the analytics and the data behind the sales forecast has been driving the agenda for Mark Schoolcraft, CFO at Midwest Industrial Supply, a privately held provider of de-icing, erosion, dust control, and soil stabilization services. Its customers include mining, construction, iron and steel, and mass transit providers.
With revenues originating in a number of diverse markets, Midwest Industrial has come to rely on predictive analytics to inform its business decisions. One reason: performance often depends to a large degree on factors beyond the company’s control, like weather, commodity prices, raw materials demand, and tariffs, all of which impact industrial production. It’s essential that Midwest has the capability to factor those variables into its forecasts.
Midwest Industrial has a fairly robust forecasting platform for all its business units due to the potential downside risk associated with forecasting errors.
“In a company like ours, with high growth and big capital outlays for heavy equipment, it’s important to get a good read on where we’re growing and what the growth trajectory is,” Schoolcraft says. The organization needs to keep a very close watch on whether it has the resources to meet the demand. “We really need an early warning system, because 10% or 20% plus-or-minus could have a major impact on the decisions we need to make,” he says.
Midwest Industrial uses different forecasting methods for each business unit, combining historical data and linear regression models with predictive analytics. “The complexity comes in with the impact of independent variables, impacts that perhaps history won’t tell us—like how future commodity prices will translate into sales,” he says.
Midwest uses Microsoft’s Power BI also, which allows it to draw from various databases. The company made the recent decision to weave data from AccuWeather into its sales, dispatch, and ERP systems. It did so because weather events hit Midwest Industrial hard last year. “It never rains in southern California, right? Well, guess what? Last year they had torrential rains,” remembers Schoolcraft. “We learned a lot from that and decided to add the weather forecast component into the planning system.”
While bad weather doesn’t really cancel projects, it moves them around and can have an impact on quarter-to-quarter numbers, Schoolcraft explains. “Before, the company would keep its fingers crossed that it was going to be sunny and bright.”
“Fingers crossed” is not the best way to do sales forecasting. While luck may still play a role in business performance, access to better data and tools should give companies a fighting chance of producing sales projections that lead to greater capital efficiency and informed plans for growth.