The Infosys global supply chain management blog enables leaner supply chains through process and IT related interventions. Discuss the latest trends and solutions across the supply chain management landscape.

« Convergence of IT & OT - an intelligent Asset Management | Main | Will best of breed EAM packages be taken over by ERPs? »

Is "Higher Forecast Accuracy" the silver bullet?

To answer this, lets take a step back and try to answer a more fundamental question - Do we need forecasts? Our immediate response to this will be YES. And for most of us, the response will be based on the following key challenges most of the Supply chain professionals face across multiple dimensions:
1. The drive towards Globalization has resulted in the focus to not only look at the developing markets for cheap supply, but also to tap these developing markets to drive future growth
2. Increasing lead times and lead time variability with most of the manufacturing bases of suppliers outsourced or offshore
3. Increasingly demanding customers with information at finger tips [thanks to internet] and lower brand loyalty
4. Intense competitive activity driving lower prices and reduced scope for differentiation
5. Increased pace of product innovation - rapid new product introductions combined with rapidly reducing product life cycles
Both the Product Supply Chain and Information chains are getting longer, making the task of the Supply chain professional challenging… and the need for Forecasting and Demand Planning more and more necessary, to ensure a steady flow of products to the right place at the right time in the right form.

Now that we agree that we need to forecast, lets take a quick look at what the Forecast entails…

In a book ‘Dance with Chance: Making luck work for you’ (published by Oneworld Publications in 2009 and written by Spyros Makridakis, Robin Hogarth and Anil Gaba) I recently read, the authors state the process of Forecasting whether carried out by people or models, is to identify some pattern or relationship amongst the relevant variables and then extrapolate from these the future. Identifying patterns intuitively is what humans do constantly and well. Models do systematically what humans do intuitively. The data available however is a combination of both these patterns and the noise [or the element of randomness].
The goal to achieve higher Forecast Accuracy gets challenging taking into consideration the following key principles:
1. The future is never the same as the past and hence a straight extrapolation will not always be helpful
2. The available historical data contains the underlying pattern inter-mixed with the noise [random element]
These principles imply that the goodness of fit of the models to historical data has little correlation to the potential of the accuracy of the forecast generated by these models. Also complex statistical forecasting models run the risk of over-fitting to the historical data and mistake patterns for noise.

Having briefly looked at some of the key aspects of Forecasting, lets now explore the concept of uncertainty, before we proceed to address the question with which we started…

As mentioned earlier, the available data is composed of both the underlying pattern or relationship and the noise or the element of randomness… It is this element of randomness that drives the measure of uncertainty or unpredictability in the data.

Well, we know what uncertainty is, but how do we quantify it?

One of the most commonly used measures of dispersion in the data is Standard Deviation, which is a statistical measure of dispersion of the various values of a dataset from its mean or Central Tendency, assuming that the data is normally distributed. Since this measure considers the squares of deviations, the deviations below and above the mean do not cancel each other out and gives more weightage to the larger deviations compared to the smaller deviations. The higher the Standard Deviation for a dataset, the more dispersed are the individual data points and thus more challenging the task of generating forecasts with higher accuracy.

Now that we have a way to quantify uncertainty, how do we consider this uncertainty?

I would again like to refer to the approach the book ‘Dance with Chance’ mentioned earlier in the blog, wherein the authors propose the following 3-A framework to deal with Uncertainty:

  1. Accept that you are operating in an uncertain world and thus identify the range of possibilities
  2. Assess the level of uncertainty using the available data and any additional inputs
  3. Augment the range of uncertainty estimated in the earlier step
    Applying this in the context of Demand Planning, we must not only generate statistical forecasts, but also assess the uncertainty involved in the demand patterns and further augment this uncertainty to create a range of forecast.
    One of the ways in which this can be achieved - The system generated statistical forecasts can then be further enhanced with any information of the external events that the Demand Planners are aware of along with the probabilities of these materializing. Further to this, we can apply a factor [which can be derived empirically or based on guidance from management] on the baseline statistical forecast to factor for the uncertainty involved.
    While this concept is very simple to understand and logical to follow, this approach is not followed commonly…

So to address the main question, my argument is as follows -
No… Higher Forecast Accuracy alone cannot be a Silver Bullet. The blog from my colleague covers in great detail various measures of Forecast Accuracy and suggests their applicability in different scenarios. These need to be further supplemented by considering many other aspects to address the challenges mentioned earlier.
One of these and also the central theme of this blog, is the acceptance and understanding of Uncertainty involved and dealing with the same by employing some of the levers mentioned earlier.

I would like to end by sharing the following quote from Donald Rumsfield - “There are known knowns. These are things we know that we know. There are known unknowns. That is to say, there are things that we now know we don’t know. But there are also unknown unknowns. These are things we do not know we don’t know.”
It is by accounting and planning for the later two, which can help us be better prepared to deal with uncertainty and thus help make our Supply Chains more flexible and responsive…

In this blog, we have briefly looked at some ideas for accepting and dealing with Uncertainty… Further to this, identifying and understanding the factors that contribute to this uncertainty has the potential to provide more leverage, as we could work towards factors that are within our control to limit this uncertainty itself, which would be the topic of one of my subsequent blogs.

Post a comment

(If you haven't left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Thanks for waiting.)

Please key in the two words you see in the box to validate your identity as an authentic user and reduce spam.

Subscribe to this blog's feed

Follow us on

Blogger Profiles

Infosys on Twitter