Expert talk: Do we really need the sales team for a quality forecast?

mSE Solutions

Rolf Winterhoff / Director of Operations at mSE Solutions GmbH, Lübeck

Fred W. Schellert / Global Director Business Development at mSE Solutions GmbH, Munich

Fred: „Let’s face it, when a demand planner asks for a monthly forecast review, most sales people’s natural instinct is …escape, get outta here! It’s just one of those things a sales manager always has a hard time doing – committing to forecast numbers. I recently had a discussion with a senior executive from a medical supplier who is in charge of the demand planning department. He has gotten to the point that he actually recommended excluding the sales team from the operative forecasting process. Rolf, what do you think about that?“

Rolf: It’s a complex matter, and before making any comments we need to understand the executive’s reasons for his recommendation: (A) Does he see the sales department’s contribution as less significant to the outcome? If so, we would need to look at whether there are tangible reasons for that, maybe for instance due to the nature of the business or products. (B) Or does he only find the current input as given from sales to not be helpful? Then the cause might be found in the setup of the process and the sales team’s interpretation of their role in the process.

Fred: The sales team’s contribution not significant to the demand planning outcome? Are you serious? How can that be? Isn’t the sales department always the part of the organization that is closest to the customer? Don’t you need the expertise of a sales manager to evaluate the actual and future situation on the customer side as well as quality of customer forecasts?

Rolf: That the expertise of a sales manager is important is beyond question. The sales team has direct customer contact and closest perspective about what’s coming up, but perhaps this point of view has a limited horizon of time. For products with short lead times due to very short or flexible sourcing/ manufacturing processes, that is very valuable. The time horizon, however, should be at least a bit longer than the customer order lead time; otherwise, what I can learn from the sales department is already in my order books. The question is, how much and how detailed does the customer involve our sales team into their future roadmaps. If we are not talking about “off the shelf” products, the customer might need a design-in, with certain technical requirements and specifications where the respective expertise is with product-/ technical marketing rather than sales.

Fred: With this in mind, what would be the requirements regarding data as solid basis for a forecasting system if the sales manager’s role is not the most important factor? Don’t you need a lot of reliable historical data?

Rolf: As to be expected, there’s no „one size fits all“ answer here. To form a solid database, we need to see how the available information covers the forecasting horizon. If your order reach and your order lead time are long enough so that supply matches demand then you are lucky. If your sales team has a time horizon that reaches beyond that – even better. From a certain point on you have to look into a crystal ball so to speak. Sure – historical data might help you in seasonal or any cyclic business, but even then you have to extrapolate since the demand behavior might be cyclic but the products will change.

There’s no escaping from (the need for?) increased structural flexibility. With increasingly shorter life cycles, late and volatile customer demands and a high level of digitalization, an optimized and agile supply chain is becoming more and more important for retaining a competitive advantage. Imagine the following: ae customer has the technical capabilities to inform a supplier about his specific demands at the very last minute and – naturally – expects the supplier to meet the demands. The challenge is to offer a maximum level of service and still be profitable. So the challenge is to make all demands that come in possible. Historical data can offer general insights, but it isn’t the answer to the problem.

Technically it comes down to a space with different scenarios regarding product mix and volumes. Based on historical sales data, customer information etc. we can assign probabilities to each scenario. We have to set up our supply chain end to end in a way that maximizes our chance of covering most of this probability distribution. Software developers have already taken these insights into account. There are quite a few products on the market that support this approach.

Fred: As an example, let’s take a look at standard medical devices that are used globally, perhaps with regionally deviating specifications.

Rolf: Well, if there are regionally deviating specs, then there is at least some limit to product flexibility – in the best case at the very end of assembly. Then the challenge is how fast that final assembly can be completed and the time-to-customer. “The faster the better” offers your sales department a competitive advantageIf you have set that up, then the regional sales figures are not that important.

But a device might be a standard, and yet have a high value, tying up a lot of capital. So you might want to only source or manufacture up to a stage where little or an acceptable amount of capital is tied up, and only continue to process from there when you receive an order.

Fred: Where do we get the necessary data from and who can do the analysis? What kind of special tools would you recommend?

Rolf: Today we need to have a results-driven approach. What decisions do we need to take, by when happens, what happens if, and so on. Based on those answers, we compile the necessary data. That is because data availability is not an issue: I don’t need to check whichI have, but have to figure out which data is significant. Integrated systems that provide scenario analyses are available as well and they can be shared with all players along the supply chain so that everybody can contribute at their respective stage. So it comes down to the availability of skilled planners – perhaps in different time zones – defining the speed.

Fred: To stay in the medical device segment, what about customer specific devices? Do we have to set up a different approach for this type of product? Customization very often could result in long lead times, and if the demand changes substantially over a short period of time it really is a challenge for the supply chain organization.

Rolf: Well, we already addressed parts which are standard up to a certain stage and then undergo a customization at the end. In other cases, it depends upon what sourcing/ manufacturing stage the customization sets in, how much capital has been tied up by then and how long time-to-customer will take from there. In cases where there’s a window of time for which you do not yet have the orders on hand but the sales team can give you good estimates about upcoming orders, then their input into the forecast is very helpful. In cases where the customer specificity goes beyond customization a design-to-order – or at least a design-in – is necessary. The timelines are then usually beyond the sales team’s horizon of time anyway.

Fred: Overall that really demonstrates how individual every supply chain is in terms of product specifics and customer structures. A good approach for one case is not adaptable to another case. These kind of complex matters always require a case by case analysis. However – thank you very much, Rolf