The Association for Financial Professionals (AFP) held their FINnext conference over the course of a few rainy San Francisco days in March. The weather forecast had indicated precipitation, so it was kind timing to be indoors learning about forecasts of all sorts and their impact on our professional and personal lives.
During the California Gold Rush, the forty-niners ventured to San Francisco looking to strike gold. This group of conference-going financial planning and analysis (FP&A) professionals were “panning” content for the kind of gold nuggets they could bring back to their respective offices. I gleaned a few of those forecasting gold nuggets to apply to my work at Anaplan with my clients.
Superforecasting—the analytical approach
Author, and now, advisor to the prime minister of Canada, Dan Gardner offered the audience much golden wisdom. Gardner is a journalist who has spent time examining methods for forecasting the future. His New York Times bestseller, Superforecasting: The Art and Science of Prediction (co-authored with Philip Tetlock), applies research, statistical analysis, and human psychology to understand how, and why, we fallible carbon-based lifeforms make the determinations about the future that we do, and with those determinations, the decisions that follow.
In a room of FP&A pros, the dominant assumption that many of us share is that the more mathematically involved our planning and forecasting models are, the more accurate they will be. It’s all about the algorithm these days, right? Gardner’s research would say “not exactly.” Specialist knowledge, extreme intelligence, and arcane math do not win the forecasting accuracy prize at the end of the planning cycle. The drivers of winning forecasts include intellectual curiosity and a need for cognition.
Superforecasters are slow, cautious, and analytical, and they ask many “why?” questions. Accurate forecasters are actively open-minded; they go out to get as many perspectives as possible—actively draw them in and synthesize them into their thinking. They update their forecast frequently, in small amounts. The readily say, “I was wrong.” They try, fail, analyze, adjust, and try again. This process speaks to intellectual humility.
We all gain from approaching forecasting with this growth mindset that Gardner spoke of because we all wish to get progressively better. Things in finance, or in life, are not 0 or 100 percent. He suggests that we think in terms of probabilities every day. And that we take Fermi’s estimate approach, “Keep breaking it down.” Break down the components of your model to smaller and smaller pieces to better understand the problem and discover correct drivers.
Forecast often and collaboratively
When Gardner left the keynote stage, I immediately began thinking about how his lecture mapped to what I see in the FP&A marketplace of ideas and practices. We are all seeking those key business drivers, but how do we find them? Collaboration is key. Communicating across your organization and with your colleagues helps you uncover insights into the business operations. Falling into “group think” can cause your forecast to fail, whereas giving weight to a variety of perspectives helps you to accurately assess probabilities. At Anaplan, we call this practice connected planning.
Companies across all industries are moving away from annual budgets to periodic and rolling forecasts. This shift aligns with Gardner’s take in that the more often we forecast, the better we get at it. Another means to better forecasting is to evaluate forecast accuracy ex post facto. Monitoring and examining your forecast error to identify bias and acceptable levels of variation strengthens your forecasting muscles.
“There’s gold in them thar hills,” so go forth and build those forecasting muscles, look into one of Gardner’s books, and find your forecasting nuggets.
Interested in implementing rolling forecasts in your organization? Register for our upcoming webinar with IMA on how to make the move to rolling forecasts.