Still believe business forecasting is not for you?

Business planning & forecasting as a subject and function makes for an interesting discussion on its importance. With availability of data and technology the function is now making its presence felt in the increasingly competitive business world. However, this is not without its set of resistances specially from the traditional thought process. In this write up we share some of the myths that we most commonly encounter and our thoughts on same.


Myth 1 – Demand Planner needs no education on planning and forecasting

This is an observed myth rather than an admitted one from the companies. We observe people from varied background ranging from operations management graduates from top B-schools to a an admin executive taking over the role of ‘demand planner’. The background of the demand planning team gives the first impression about the seriousness of the management about the process. Irrespective of the background, the fact remains that no B school teaches demand planning beyond a chapter in operations or supply chain book and its just not enough. This requires a specialised training from the practitioners, both in terms of quantitative techniques and process best practices. The benefits accruing from this education investment helps both the professional as well as the organisation many fold

Myth 2 – ERP system already has a planning module
All ERP systems big or small are super shops for automating business process. They can and mostly have digital module for almost every business process. Hence it only makes business sense for an upsell opportunity to add on a simple planning module which may just mean replicating a spreadsheet based process. However, it does not necessarily mean the smartest business decision for the buyers. ERP systems are traditionally transaction-processing systems that record transactions digitally. Forecasting requires a specialized analytical firepower, which is not the forte of the ERP systems no matter how big the name may be. Also, the ERP systems fail to offer a strong collaborative capability that is extremely important for a planning exercise or prove to be very expensive with that feature. The tragedy of the story is that it is not even the case of penny wise pound foolish as most ERP companies charge much higher for the planning module than a far superior stand alone forecasting system. With better options available now, we hope the finance teams make a better note of this cash drain.

Myth 3 – All in One makes it the best
A very IT thought process that favors implementing one system that has all capabilities. We hear this often that ‘our IT team has recommended that we go in for system that offers all in one solution”. If this were a music system, it surely would have made sense! Even from IT perspective it makes sense to have one system to maintain than multiple. Unfortunately this is neither an electronic device nor the IT problem that we are solving. Today’s business issues are complex and require specialize solutions. Two decades of IT revolution has ensured that any decision related to a software is made by the IT team rather than the business team. Organisations must realise that this is no more the case as in today’s world everything is run by a software and hence its imperative that functional teams, who are knowledgeable about the need, decide on the IT solution rather than IT team deciding on the functional solution.

Myth 4 – Our business is different and too complex for forecasting
 One of the most common defense against setting up a good forecasting process and it makes our jaws drop. Really? So you do business in real time? No planning at all? Perhaps this would have been possible in the era of monopolies but not in the time of dog eat dog world. Well it may be tough to do forecasts but that’s the nature of the beast. By living in denial you are only making yourself uncompetitive.

Myth 5 – We have a lot of factors affecting our business, just time series won’t work
In the world of forecasting the regression method adds the x factor. So much so that a lot of people directly jump to the thought of regression when they hear the term forecasting.  Perhaps it sounds more sophisticated than the times series and makes them feel important who is running a complex business. The truth is that the lot of factors affect all the businesses and all businesses are complex to run. Forecasting is not aiming put running the business in auto mode. It attempts to flag, acknowledge and mitigate the risk.

As for regression, most people don’t realise but it is a highly unreliable method. We need to first forecast X (independent variable) to be able to forecast Y (dependent variable) and hence there is added uncertainty component. Demand forecasting is an operational exercise done up to the SKU level so that material planning can be done accordingly whereas you really cannot find correlation of an external factor for each of the SKU. Promotions and channel programs are business events and again do not require regression techniques to manage

Myth 6 – If we are doing forecasting then we must get the last mile accuracy from day one
A lot of people get over excited about the forecasting process. The thought process seems to be that if we are forecasting then why not forecast everything and let the algorithm take the decision that the teams will just need to implement. That could mean not only the demand forecast but even at what hour the delivery truck should deliver or even more bizarre data point. Data science and technology certainly makes a lot of things possible today. However, every possibility requires some basic steps and preparation. A journey of thousand miles begins with a single step and we must acknowledge that there is a difference between a forecasting tool and a crystal ball. We must understand our capabilities as well as the need and take one step at a time.

Myth 7 – We know this already, we do not need forecasts
We hear this often from medium size growing companies where the senior management is still in complete control of the business and can pretty much guess the trends and estimate the numbers with reasonable accuracy. It is quite funny at times, that they feel proud to be able to guess the numbers with similar accuracy as the forecasting system. In contrast they should appreciate the fact that an algorithm can replicate the knowledge that otherwise comes with years of experience to be able to guess. The question should be whether you really want to be guessing the numbers yourself and become a constraint for growth?. Remember that if you don’t then your competition will.

Myth 8 – Sales team cannot override every number, they don’t have time
 We agree with this completely. However, the fact remains that every number should not be overridden anyways. If that is the case then the purpose of forecasting gets defeated. The baseline forecasts should be as strong as possible and ideally the sales team should be overriding on exception basis where they bring in additional business information. However, since the sales representatives are used to filling up spreadsheets in the name of forecast it is their natural reaction. One of the big benefits of deploying a statistical forecasting process should be to reduce the efforts of the field sales team on spreadsheets, guessing the numbers and instead add value to it where needed.

It is myths like these that keep companies from advancing. The world has now shrunk to nothing more than a coporate battlefield. In this battle, the best weaponry would be effective business forecasting and be one step ahead in the game. For that to happen we need to polish our armour by using cutting edge, specialised forecasting solutions. After all a good defence is the best offence. And there is no greater defence than a firm enjoying the lean inventory and remarkable customer service that comes from reliable forecasting.

Comments

  1. Steven Miller says:

    Just a brief, but reinforcing comment on Myth 2 and Myth 8. First, about Myth 2, this is an excellent point….large ERP systems tend to include ‘planning’ modules because they desire to be an ‘end-to-end’ business functionality competitor with its client businesses. However, the problem tends to be that ERP systems tend to be good at transactional information functionality (orders, shipments et. al.), but lack the sophistication required by complex planning systems. Twice I have been involved in an ‘end-to-end’ ERP implementations. Both have been in multi-billion dollar (sales) organizations. I have been part of the planning implementation team and have found the functionality so generalized and so weak that we needed to re-write the functionality. We did it in such a way that we were upward compatible; however, it added project cost and time. More on Myth 8 in another post.

  2. Steven Miller says:

    Writing relative to Myth 8, with the 21st century’s demand planning modeling engines, like Planamind, there is absolutely no reason for anyone in the organization to review every SKU forecast! That may seem a strange statement coming from a person involved in the demand planning business function for 30+ years, but it is now truer than ever. With businesses more complicated than ever and product diversity (in kind, customer-specific and/or location), even greater than ever, the number of SKUs that a demand planner needs to manage is beyond the capability of individual attention. Attention should be paid only to those sets of SKUs for which specialize operational information exists (like a promotion from marketing). Also, the capability to produce exception reports based on extraordinary high relative error can easily be created and tailored to the organization’s unique requirements and differentiators. It is the egregious offenders of these error metrics that require unique individualize attention. The forecast of the remaining SKUs will prove sufficient to run a smooth product supply. To bust Myth 8, it can be said, ‘work smarter not necessarily harder’.

  3. Karthick says:

    Nice eye opener for young demand planner like me 😃

Leave a Comment

Your email address will not be published. Required fields are marked *