Forecasting in business, at least by conventional means, is fraught with complexity at the best of times. When the health of the wider business environment is itself uncertain, forecasting becomes even more challenging. But a robust forecast is at the heart of effective business analysis – itself the foundation of good decision-making. That does not change, regardless of the circumstances in which the business may find itself.
Much has been written about the various methods of compiling a usable forecast. But the truth is that most large organisations do not have the time, effort and resources to analyse the whys and wherefores of business forecasting, and how its various foibles may or may not apply to them. Particularly in times of worry or stress, business and finance leaders just want a forecast that they can rely on. The traditional bottom-up sales forecast, which typically relies on the judgement and insight of sales people themselves, is not consistent enough to form the basis for critical actions.
Unfortunately, there is no instant solution to this challenge. But advances in predictive analytics mean that it is possible to create a reliable automated solution within a few months. An algorithm based on historical conversion rates, which can account for cyclical trends, seasonality, growth rates and other relevant factors, can provide a firm foundation for a reliable automated forecasting tool. Of course, to be accurate, such an algorithm will need to be unique to the organisation using it, and back-tested against actual results, in order to refine it over a period of months.
However, providing it is managed on the correct platform, with an intelligent ETL interface between the analytics engine and the pipeline data, this method can provide a reliable short-term forecast for organisations with a well-populated, fast-moving pipeline. The anonymised chart below shows a fifteen-month back-testing sequence for three-month forecasts from one of our recent client engagements. It’s easy to see how consistent the forecast has been during this period.
With an automated forecast like this in place, management can act on it with complete confidence. What’s more, the time saved through automation can be used by the finance and sales team in devising and implementing more effective sales strategies. With a platform that is able to segment the forecast according to the data structures and dimensions used within the business (divisions, geographies, products etc.) then the forecast itself can be a powerful addition to this process.
While improving forecasting within an organisation is not an effective emergency reaction to difficult times, organisations with a good forecasting process in place are almost always more resilient than those relying on conventional means. Although it is not the work of a moment to create a reliable automated forecasting process, it is certainly faster, and more likely to succeed than most corporate change projects. And it is certainly never too late to begin.