Financial instantiation has evolved from manual data collection and aggregation to real-time cloud collaboration with advanced statistical forecasting and planning tools.
For decades, organizations worldwide have asked planners to gather, format, and aggregate demand forecasts manually. This required them to call on regional managers, product-line managers, department heads, and others to turn in their numbers. Once that is done (if it actually gets done), the laborious and error-prone task of decentralized forecasting would proceed. Planners collect budget and forecast information from low-level managers, compile it in spreadsheets, and then try to make sense of it all each month, quarter, or year.
Using Excel in Planning from the Past
This semi computer-assisted process was a marked improvement from the paper age, but it is by modern standards an antiquated exercise in 1990s business technology: a mishmash of desktop spreadsheets, emails, documents, and version updates. In hindsight, we can appreciate that the manual data-collection process made sense in the Excel dominated 80s and 90s, and even into the web-based computing environments that emerged in the early 2000s. Budget administrators could send and receive multi-tab Excel docs, each tab representing one area of revenue, expense, CapEx, and so on.
Then, myriad Excel formulas could summarize the data and apply business rules. The problem was (and continues to be) that the basic inputs all had to be manually typed by people, often without any statistical basis. The outputs, therefore, are inaccurate. These non-statistical forecasting processes carried forth into many of the early web-based BI tools, which to some degree were an online refurbishment of earlier manual forecast processes.
To be fair, the ability to send and receive completed budget templates did empower planners to some degree, but the lingering manual processes surrounding these correspondences perpetuated the error, inaccuracy, and process inefficiencies of old. It’s unfortunate that to this day, what should be a bygone manual-forecast paradigm is still very much alive. Indeed, even though data collection-based forecasting can’t deliver high-level accuracy or efficiency, it remains in place at many large organizations.
Following are just a few reasons why this is so wrong:
- New employees: Budget planners who are new to the organization have no basis for understanding the particular ins and outs of the forecasting and budgeting process.
- Lack of historical data: Spreadsheet budget templates have inherent “real-estate” issues in terms of columns and horizontal scrolling. As such, many organizations summarize the history to a single column of prior year data, which ignores patterns such as seasonality, or one-time effects from discontinued items.
- The human aspect: This is not a scientific argument. Rather, I offer my own experience as a budget planner who has worked in many different forecasting and planning environments. We are all a bit optimistic when it comes to planning the next year’s budget. We see the future through rose-colored glasses. We tend to increase the sales numbers, hire more people, and spend more money on travel and marketing than is necessary.
We’re still not providing our budget planners with the tools they need to excel at (or even perform) their job.
What is the real cost of missing the forecast? A publicly traded company that misses the forecast gets punished summarily by Wall Street. Recently, Target shares tumbled 15 percent after reporting worse-than-expected fourth quarter same-store sales. This is but one of many examples.
Modern organizations recognize the criticality of accurate forecasting and are changing their cultures and technologies to promote advanced-analytic forecasting. With today’s cloud computing and grid computing environments, organizations can run hundreds of simulations on the fly to predict outcome probabilities and optimize plans.
This includes Key Performance Indicators (KPIs) such as capital optimization and workforce optimization. In both cases, Monte Carlo simulation is a highly effective tool for the office of finance to simulate uncertainty models. Advanced companies are leveraging statistically based forecasting while empowering budget planners to review, verify, and correct forecasts when needed. At the end of the day, changing the forecast culture from data collection to statistically based forecasting will drive higher forecast accuracy and reward the organization as a whole.