Demand forecasting in the retail industry

As the major UK retailers begin to report their results for the festive season, it is clear that early indications of poor footfall and depressed margins were correct in many cases. Unseasonable weather has been blamed in some instances, and it seems some firms may have been caught out by the profile of spending over Black Friday and later, which appeared to be weighted more than expected towards online purchases. The whole episode provides a vivid lesson on the limitations of traditional FP&A techniques, and the need for more powerful tools to cope with the vagaries of the modern consumer economy.

There are clearly many factors at work here, but two stand out. In both cases, they highlight a failure to integrate key pieces of non-financial information into the forecast. In the case of Black Friday, it’s clear that demand forecasting failed to take account of the fact that many bargain-hungry consumers would have been put off by events of the previous year, when scenes of disorder at stores made headline news. The unseasonable weather is more a matter of probability – or a failure to properly acknowledge the possibility that the weather could be warmer than the seasonal average.

In both cases, it is obvious that better integration of non-financial data with financial forecasting could have enabled more effective decisions around planning and stock purchases. The profusion of discounting even between Black Friday and Christmas eve, and the unusually deep discounts on offer, all point to ill-informed stock purchasing decisions that would have had a serious impact on the margins of even very disciplined retailers.

While the financial and reputational costs of running out of stock cannot be underestimated, it’s clear that this festive season saw many retailers tip the balance too far in the other direction. It points to a forecasting process that is focused too much on the financial model at the expense of operational factors, is based too firmly on averages, and fails to account for the full impact of any potential deviation from those averages.

Conventional accounting methods and forecasting software cannot easily integrate financial and non-financial information in a manner that is accurate and consistent. But organisations that develop this ability are better able to respond effectively to unfavourable external factors.