Forecasting: A Fantasy or Fortune

A topic that has the power to entice and elude, bedazzle and bewilder, in the same breath is what captivates the cerebrum in mankind. It is no wonder then the subject of Forecasting has enthralled the academia and business world with equal vigour. The volume of scholastic work, white papers, presentations etc. on this topic stands as a testimony to the importance attached to Forecasting. However given today’s customers’ quixotic choices, is Forecasting slowly fading away from the realm of business into a Fantasy or does it serve as a right platform from where Fortunes can be made.

A common understanding or a definition of forecasting so to say states, that a particular outcome will occur in an event with a relative certainty or sometimes broadly also summarized as prediction of future. It is definitely not without merit that this understanding or definition has been propagated through ages, as the term Forecast if split as Fore and Cast, the former implies that which is situated in the front or first in terms of space or time, while the latter in this context can be considered as steering something forcefully in a particular direction.

The above interpretation of forecasting makes it clear as to why in the realms of planning it is one of the primordial areas that is looked into as it sets the agenda and in most cases decides the outcome. The field of material planning follows suit by according forecasting a high pedestal as the entire supply chain can come to a grinding halt if the material forecast goes haywire. The questions which then arise are, with so much at stake should all materials be forecasted? should each of the customer options be forecasted? what will be the cost of forecasting? should advance planning systems be deployed or is it suffice to do it with just pen and paper? etc.

If ever there could be a one word answer to the above questions it would be ‘Depends’. Yes the field of forecasting provides some powerful tools, but yet one has to be cautious before trying to forecast every single material. A detailed study needs to be carried out on the materials, to understand their classification, their BOM structures, their consumption / demand patterns, etc. This in-depth study results in identifying those materials which are relevant for forecasting, but make note it does not imply it has to be forecasted as further assessment is to be done. As any expert material planner would confess, randomness in consumption or demand is the Achilles heel of forecasting.

The modern day enterprise software such as SAP comes to the rescue of planners by providing a plethora of solutions to address the complexities. This area of planning is well served by both the Production Planning & Control (PP) and the Advance Planner & Optimizer (APO) modules. The choice of whether to go for PP or APO or both can be made subsequent to a detailed study. However prudence must be exercised in making the selection, as no matter how important forecasting is, that alone cannot be the single factor in making the choice of adopting PP or APO or both.

Some of the parameters which should form part of the study prior to making a solution choice are which forecast models are required for planning, which error levels are to be considered, what level of forecast accuracy is expected, whether product life-cycle needs to be considered as a part of forecasting, what reporting and analysis tools are required by planner post forecast run, etc. The forecasting process itself is the same where-in the first step is where the historical values are adjusted, in the second step the historical values are initialized an ex-post forecast is created followed by the actual forecasting based on the model and the final step being adjustment of the results.

In addition to the forecast models such as Trend, Seasonal, Seasonal Trend, APO provides additional models such as Croston, Linear Regression, and Seasonal Linear Regression. It can also be used to do Causal and Composite forecasting and supports additional forecast errors such as MSE, APE etc. In PP the forecasting can be done at the material level where-in the forecast can be done based on consumption data or in flexible planning, where a few additional key figures can be included in the forecast. In APO, the Demand Planning (DP) provides for a more flexible platform for performing forecasting.

Given today’s rapidly declining product lifecycles, complex manufacturing chains and transient customer loyalty, we are possibly at an inflection point in the history of Forecasting, where getting it right could mean survival or being part of the history books. It is precisely these high stakes which puts a great onus on selecting the right solution platform along with a competent person to navigate which would go a long way to ensure that Forecasting serves as one of the powerful tools to unlock a Fortune for the company and does not merely remain as a Fantasy.

Thank you for reading through the blog, do leave behind your views, comments, critiques.