How forecasting has evolved to boost strategic planning

Business forecasting gives key decision-makers a compass to navigate the variables inherent in an uncertain business environment, and better understand the levels of risk associated with each decision. While it has been practiced in some form for over a century, recent technological advances — most notably in AI — have unlocked a new era of data, accuracy and insight.  

For consumers, the convenience of digital commerce and the draw of social media have led many to rely on the internet for commercial and social activities. This generates trillions of data points that can be tracked and measured — a window into real-world state of consumer behaviors, preferences and intentions. 

Business forecasting transforms this data into actionable insights that inform strategic planning and ensure brands stay responsive to fast-changing consumer behaviors and market conditions, and remain relevant.  

The recent explosion of Large Language Models (LLMs) like ChatGPT in the public sphere have catapulted AI, and its role in society and business, into the collective imagination. Yet, many have mistaken this sudden popularity for a sudden beginning.  

For example, teams at GfK have been working on transformer-based deep learning AI architectures or models — similar to what powers ChatGPT — for years to deliver cutting-edge global forecasting tools for Tech & Durables manufacturers and retailers. These models simulate neural networks, learning context and meaning by tracking relationships in sequential data — in the case of gfknewron Predict, to predict future market trends and reduce business uncertainty based on the world’s most comprehensive Tech & Durables sales tracking data. Notably, however, the forecasting models used by gfknewron Predict don’t suffer from nonsensical or inaccurate outputs like their LLM counterparts.

The key difference between manual statistical analysis and neural net-based forecasting is the depth and breadth of their analysis. Manual analysis limits calculations to a few variables at once, while neural nets have the power to analyze thousands of correlations (including higher order correlations) simultaneously. This makes them much more efficient and accurate when modeling complex systems.

The more sophisticated your organization’s forecasting system, the more successful it will be in lifting out the signal from the noise of consumer and market data points to create (and then leverage) a clearer, more accurate vision of the future.  

To be useful, strategic planning requires an accurate understanding of the business and its external environment. This ensures strengths and weaknesses are accounted for, and that opportunities and threats are navigated accordingly. As forecasting technology has developed, the value it offers strategic planning has grown exponentially. 

Think of a forecast as a map of unfamiliar territory. As detail is added to the map, its value and utility changes. At its most basic level, the map helps with orientation: helping a business recognize which direction it's moving in, using its previous location as a reference point. With more detail, the map can help the navigator avoid hazardous areas: warning businesses of potential disruptions and suggesting an optimal route. With even more detail, it can show navigators what’s worth pursuing: revealing opportunities for expansion into new markets, the development of new products, and ways to maximize profitability.  

Today’s most powerful forecasting solutions function like the best maps: they provide quality data and insights that help companies navigate the uncertain business environment with confidence backed by data. As such, they’re invaluable to strategic planning — preventing the mishaps inherent in taking the proverbial wrong turn.  

Forecasting has evolved to this point through advances in three key areas: Big Data, computational power, and scientific research.