Supply Chain Management

Companies go through the exercise of demand planning in order to obtain a best estimate of the future sales. These estimates are then used by various functional managers to plan for sales, promotions, inventory, capacity and the finances. These plans drive critical decisions like resource allocation and personal/departmental/organizational targets. To achieve these targets, the management uses incentives (such as bonus, increments and promotions) as carrots. The incentives attached to these targets make the people take the targets seriously and passionately. In fact, I am sure people are more driven by the incentives attached to the targets than the targets themselves. To cut to the chase, there are enough vested interests involved to bias ones judgment for their own benefit. Not just that, the vested interests also drive people to aggressively influence others decisions to favor themselves- just like the proverbial salesman who coaxes the Eskimo into buying a refrigerator.

Tying this back to forecasting, manual judgmental inputs on future business assumptions are needed to adjust the statistical forecast- for a simple reason that it is difficult to represent these assumptions in a statistical model. Who gives these judgmental inputs? You got it- it’s the very same functional managers who are highly incentivized to meet their targets. Bingo, there lies office politics. When such political forces are attempting to bias the outcome of the process, the forecast quality is obviously compromised, pulling down with it effectiveness of resource allocation causing potential impact to corporate vitals.

However, the gurus of forecasting attribute two reasons for the forecast bias- intentional bias and unintentional bias. Intentional bias is caused due to the political factors discussed earlier, while unintentional bias is caused due to information gaps (such as not sharing the information of business assumptions with the group) or process gap (such as not having a well defined process for sharing the information used to arrive at a single consensus forecast that drives the supply chain).

In order to systematically minimize the intentional or unintentional bias from the forecast, companies implement what is known as sales and operations planning (S&OP) process. I have been part of one such implementation for a large CPG company. I will talk about the S&OP best practices I have come across in my next blog. Meanwhile, I would love to hear from you on your experience in dealing with forecast bias.