It's Never Happened Before: The Art of Forecasting the Possible

There's a popular — if somewhat vulgar — aphorism that we will paraphrase here: "Stuff happens."

This is profoundly true when it comes to forecasting, with Tim Berry, founder and CEO of Palo Alto Software Inc, going so far as to say that "business plans are always wrong." This is because no matter how sophisticated your forecast may be or how rich your historical data was, an enterprise is more than likely to encounter an event — adverse or otherwise — that wasn't part of the plan. And as often as not, this is due to the event being something that, according at all your data, has never happened before or at least never had this kind of meaningful impact.

"The absence of evidence is NOT the evidence of absence."

Planning the Unplannable

"How do you expect me to forecast for an event that's never happened before?" you may very well ask. In some respects, the lack of historical precedent may suggest that it is impossible or simply a waste of time to attempt to predict what has never happened. Why plan for the "unplannable"? 

This points to a fundamental fallacy inherent in many business forecasts: Mistaking the absence of evidence as the evidence of absence. This is something both astronomer Carl Sagan and former U.S. Secretary of Defense Donald Rumsfeld can agree on. Just because something has never happened before doesn't mean it will never happen. All we can glean from an event that lacks historical precedent is that it simply hasn't happened yet.

This doesn't excuse the forecaster from suggesting it was "impossible" to foresee. If we are strictly bound to follow historical trends, then we can easily miss the impact of some of the most meaningful developments driving business trends: technology, political movements, social and environmental changes. Shutting your eyes against the possibilities won't make your enterprise any safer or more secure.

Visualizing the Possibilities

Solid forecasting requires that the forecaster have an imagination for the possibilities. The key is to look at the world around you and speculate on how existing trends could evolve or how single events might occur and  disrupt your business. A personal example is saving up to pay medical bills related to a major illness. Just because you are feeling hale and healthy now doesn't mean you'll never get sick. If and when – and it's far safer to assume that it's just "when" rather than "if" – that happens, having a built-up reserve and plan in place will lessen the  disruption to your financial affairs.  Similarly, organizations usually provide for bad debts as a percentage of revenue in order to avoid the volatility of recognizing each bad debt as and when it occurs.

These scenarios aren't really of the 'End-of-the-World' caliber, but are still experiences that have a meaningful impact on any enterprise. Rising oil prices, terrorist attacks, global warming — beyond the specific threat to personal health and happiness, these have long-lasting and potentially tricky to pin down effects on business prospects. If your forecast hasn't visualized these types of events as possibilities, the impact could be devastating.

Forecasting is Both Art and Science

A key component of forecasting is science: the ability to look at the past, at trends, incorporate known or planned future circumstances, and develop a forecast based on that data; this of course is the science. The art however is to envision the spectrum of possible events and consider how an enterprise might prepare for and respond to them. This requires a shift in culture to encourage, perhaps mandate, the open discussion of ideas and possibilities, and a shift in modeling approaches to ensure that 'out there' scenarios are also reflected in the range of possible outcomes. While it is clear that specific events may be difficult, if not virtually impossible, to predict, one can artfully foresee the broader possibilities and plan for them.

While spreadsheets generally require users to save different versions thus introducing the likelihood of formulas and data differing across these versions, today's multi-dimensional tools allow simultaneous comparison across scenarios, as well as to the baseline, prior periods, etc.  This is the type of rich information management needs in order to assess alternatives, take action, and thus forecast for the unknowns.