Last week, we had very interesting conversations with supply chain practitioners across two different industries viz. FMCG/Retail and Consumer Durables. There were some standard supply chain efficiency metrics that applied to the former industry group and which was clearly not so important for the latter. That got me to do some literature survey and dig deeper and what I found was quite fascinating. Hence sharing for the larger audience for whom supply chain network design has been important yet difficult to understand the underperformance.
Professor Marshall L. Fisher of Wharton Business School (and founder of his own SCM Planning startup 4R Systems) in a 1997 HBR article “What is the Right Supply Chain for your product?” states that never before has so much technology and brainpower been used to improve supply chain performance with PoS scanners, EDI, flexible manufacturing, automated warehousing, optimization algorithms etc. and yet the performance of many supply chains has never been worse. This probably holds true today as well and not just because of the pandemic driven disruption in the supply chain over the past few months. Professor Fisher goes on to say that poor industry coordination amongst end to end supply chain partners resulted in a waste of $30 B annually in the US Foods Industry and excess supply of products and shortage of others resulting in markdowns and stockouts in the same retail store. His view: lack of a framework of assessing the nature of demand and devise an effective supply chain strategy to satisfy that demand.
Products fall into two broad categories by demand patterns – functional and innovative. Each require a different design of their supply chain network with focus on different drivers viz. responsiveness versus cost. So is reducing the lead time to supply more important or is it increasing efficiency in distribution? Once we have the understanding of what kind of network design we need, it is quite easy for supply chain experts to build the right solution.
But here is the catch: we often mis-classify the type of product and therefore the supply chain metrics we need to improve. For instance:
Functional Products are basically essentials that satisfy basic needs, do not change over time with stable predictable demand and long product lifecycles. There will be product line extensions but these product families rarely face demand fluctuations beyond the occasional “hoarding” under trying times or price promotion driven sales push which is essentially pushing demand down in the following few months as underlying consumption trends do not change.
Innovative Products are ones in which innovation are brought into functional product categories or new types of products where new innovative features makes them highly profitable but with unpredictable demand. The typical lifecycle is short as imitators from competing players will erode the competitive advantage that innovators enjoy and profit margins will erode over time. The short lifecycle and greater product variety makes demand even more unpredictable across product skews.
With the high profit margins and unpredictable demand, the design and performance of the supply chain of innovative products is dramatically different from functional products. However, this is often not taken into account and sometimes, the new supply chain team is hired from “efficient” functional product organizations and introduce metrics they understand best – efficiency driven rather than responsive. Hence, cost of inventory and transport logistics is more important than service level (chance of stockout) and effective lead time management. Therein lies the challenge. The market mediation function of the supply chain is malfunctioning when the design and metrics are flawed. An example of such an error is the selection of suppliers based on low cost & efficiency and not their speed and flexibility.
Sport Obermayer is an “Innovative Products” manufacturer whose new products introduced per year is 95% of total products and therefore demand forecasts can be off by 200%. The primary purpose of the supply chain is to respond quickly to demand variations, minimize stockouts as well as markdowns (due to excess stock), invest aggressively to reduce lead times and choose vendors with speed, flexibility and quality. In terms of product, the differentiation should be postponed to as late a stage of the production or assembly process as possible. Hence, while a car dealer can potentially offer a million variants to the customer in terms of paint color, seat color & trims, alloy wheels etc. the final assembly of the mechanically assembled vehicle undergoes last mile customization on a made to order model which may take a week at most to customise and deliver the product. This “mass customization” process allows higher product margin and the minimized lead time to delivery keeps the customer hooked. Win win.
The advice to organizations is to classify each product skew into functional and innovative even though they may be pretending to be the opposite. For instance, having twenty plus variants of toothpaste with new brand promotions and flavors are essentially functional products in the garb of innovative products while color TVs and mobiles seeking efficient supply chain for newly introduced OLED/QLED TVs substituting LED-LCD variants is clearly innovative products where demand is difficult to predict. One of the solutions proposed and which has been very effectively utilized by Sport Obermeyer’s range of fashion apparel for sports is to accept uncertainty in demand as nature of the innovative product and do the following:
- Analyse past trends and find leading indicators of demand (weather, season, etc.)
- Cutting lead times and increasing the flexibility of the supply chain
- Create buffers of excess inventory or excess capacity
In simple words, utilize the lean period production capacity to produce goods with predictable demand and keep the capacity and inventory stocks to be able to produce the unpredictable product skews by postponing customization to a late stage as discussed earlier. This is called “accurate response” and one of the ways of calling out the unpredictable skews was to calculate the variability of forecasts amongst demand planners of the company working independently – where variance was low in the forecast, the demand was presumed to be predictable and vice versa, apart from historical data and expert judgment on the same. The system postponed decisions related to unpredictable skews until there is some early market signals such as early season sales and has two components: takes into account missed sales opportunities and factors in cost of stockouts and markdowns into the planning process. The lack of mismatch (or dramatic reduction thereof) results in ability to reduce price which is locked in for a demand mismatch that factors cost increase due to this. This model, of course, presumes incremental production of right skews is possible even though common inventory is procured as newsvendor model and cannot be second-guessed.
Net net, the system resulted in 99% product availability and retailers ranked Sport Obermeyer numero uno in terms of service. Current manufacturers and retailers of innovative products like Color TVs and Mobile phones dealing with long supply lead times and unpredictable demand for newly launched skews can learn to do much better in hindsight.