Bullwhip - do we understand the real problem?
"Traditional" Bullwhip
Bullwhip is usually understood to describe how companies within a supply chain tend to amplify variations in demand from customers such that downstream changes get increased as they're passed upstream leading to cycles of over and under production, excesses and deficits of inventory, periodic service issues and unnecessarily high levels of capacity.
It was first written about by Jay Forrester in his book "Industrial Dynamics" and Jack Burbidge in his article "The New Age of Production" - both in 1961. It has since become a subject of considerable research and a staple topic in SCM education, including use of the 'beer game' as a means to demonstrate how pervasive it is, even under the most benign conditions.
There are numerous reasons for demand amplification / bullwhip. The 'beer game' for instance demonstrates behavioural issues such as how, after a demand surge, supply orders get duplicated as players repeatedly re-order for backorders, forget that adequate orders have already been placed and feel pressurised to order extra "just to be safe". Operational issues are also revealed such as exaggerated increases in safety stocks and demand forecasts both of which get replenished in addition to the required replacement stock. And as these demands are passed up the supply chain further exaggerations occur due to the direct effect of, and demand latency caused by, large batch, infrequent cycle and multi-phase ordering. Other causes include promotions and discounting that can falsely provoke demand and, in times of shortage, rationing and gaming can similarly stimulate artificial growth.
Understanding of bullwhip has led to a variety of recommended counter measures. These include information sharing in which providing visibility of end customer demand to upstream companies can lead to more informed decision making and realistic responses including, but not requiring, VMI / CPFR; more frequent ordering & replenishment with smaller batches through milk-runs and SMED, order smoothing with use of control theory (a rich area of research in academia) and 'every day low pricing' instead of price promotions. One of the most potent initiatives in the fight against bullwhip is the implementation of pull replenishment across the supply chain because it decouples replenishment from forecast inaccuracy which, as we've seen, is often a prime cause of demand amplification.
There is, however, a far more damaging form of bullwhip that is nothing to do with inter-company trading, is experienced by nearly all ex-stock manufacturers and is entirely self generated - though most have no idea its happening and the costs that its increasing.
Internal Bullwhip
Internal bullwhip is that caused by interruptions to forecast-driven master production schedules that are brought about by Planners having to re-schedule to avoid imminent backorders brought about by under-forecasts. c80% of item level forecasts are >40% wrong so, despite the existence of planning time fences and safety stock, expedites of this nature are inevitable and frequent (and often prompted by MRP exception messages). They may prevent a particular backorder from occurring but in doing so they are using unplanned capacity and delaying the movement of other materials on shared work centres. This delay then cascades up the routing affecting all other affected items, causing inventory congestion and increasing their lead-times. The latter then also pose a service risk (because ERP/MRP/APS assumes lead-times are known and fixed) so yet more expedites and schedule interruptions are required. The result of this vicious circle is on-going wasted capacity (requiring over-time cost to recover), ever lengthening lead-times and, through Little's Law (inventory = lead-time days x daily output) excess inventory - not to mention continuous service problems and suppliers constantly being asked to bring forward delivery dates.
The way to eliminate internal bullwhip is to stop interrupting schedules with expedites and, given that high forecast accuracy is a pipe-dream, the only way to do that without causing backorders is to de-couple material movements from the inaccurate forecasts. So the alternative is the inexpensive option of implementing enterprise-wide pull replenishment.
Enterprise-wide Pull uses calculated and forecast-calibrated de-coupled re-order point/cycle mechanisms within the supply chain (eg. raw materials, break-bulk positions, finished goods locations) that enable materials to start moving in line with real demand (ie. material flow) and, thereby, autonomously ensuring that enough materials are positioned in the right places at the right time to meet actual demand. Counter-intuitively, the deliberate positioning of inventories within the supply chain, and demand/consumption-driven replenishment, allows lead-times to drop considerably because schedule interruptions cease - materials therefore move more quickly and inventory diminishes (usually by c40%), far less capacity is lost and planned service levels are achieved. Nowadays all this can be achieved using software that operates through your ERP and the greater the volatility of your demand patterns the greater is its impact (because with volatile demand the forecasts are that much more inaccurate). Enterprise-wide Pull doesn't, of course, eliminate event management but these softwares also support the advanced stock builds required to manage significant and exceptional demand increases, major promotions TV advertising, NPL, capacity constrained seasonality, factory holiday closures and the like as well as forecast driven S&OP/IBP which is still essential to ensure adequate capacity is maintained.
More information on software supported Enterprise-wide Pull (case-studies, courses, software providers, academic research, conferences etc) can be found at www.demanddriveninstitute.com and below is a selection of reading material on 'traditional' and internal bullwhip.
Once you've minimised your internal bullwhip, you might even want to consider extending Enterprise-wide Pull up and down the supply chain using its 'collaborative' version to tame 'traditional' bullwhip -see The 5 T's of highly effective supply chains
领英推荐
Supply Chain Executive at Retired Life
2 个月The Bullwhip Effect – Cheat Sheet. https://www.supplychaintoday.com/the-bullwhip-effect-cheat-sheet/
Logre niveles de OTIF de 95%+ y reduzca sus inventarios entre un 20% y un 50%. Desde 1997, hemos ayudado a los líderes de cadena de suministros a obtener estos resultados comprobados.
3 个月Simon, interesting as usual! Thanks for sharing. I would like to add that the real cause of the bullwhip effect, as described by Forrester, are the delays in the arrival of materials and information. As you well know, delays are a devastating phenomenon in any system. And a decoupled system minimizes those delays. This is one of the most powerful advantages of DDMRP that few people are aware or.
Agree David. The importance of integrated Education and collaborative execution across the end-to-end supply chain within a model of synergistic governance that maintains organizational alignment and ensures all are consistently working within a single source of truth is the basis for making a difference and reducing bullwhip.
Author, Consultant & Visionary. Accomplished Supply chain Professional, International Speaker
3 个月Bullwhip is self inflicted and it's not easy to fix, but it IS fixable. Education plays a huge roll in the fix, but ultimately it's a series of individual things that all need to happen at the same time. It starts with leadership, cohesive leadership up and down the supply chain. Everyone needs to be working with the same information, using one system of record and siloes being demolished. That requires leadership to step up, education, good truthful masterdata, adoption and accountability. Until all these 5 are working in tandem you will stay being whipped. It's one of the hardest supply chain things to do but, if done correctly it will be the best team exercise you'll ever go on.
Helping Companies Realize the Full Potential of SAP's Advance Planning (PPDS) with Add-Ons and Experienced Consulting. Author of ADDV Planner (addvplanner.com), a certified DDMRP Add-On for PPDS.
3 个月So much knowledge from such a simple game. We used to play it with beer bottle caps. That makes it even more visible. Sometimes the stock accumulation was so high that we ran out of caps (and we drank many beers to get them) Internal bullwhip is an interesting element of this fractal game (the more you look at the details, the more complexity it shows). Thanks for your contribution. I would like to chip in with a psychological point of view. As an IT consultant I often see that when the planner doesn't understand that the variability that is ravaging its efforts is an intrinsic factor in any process, it blames the system for the results. With an "I know best" approach, system suggestions are overwritten and manual charges are fixed, tying the system's hands to do its job. It's a vicious cycle as missing the system opportunities for improvement gives an even less optimal result that further reduces the trust in the system. There are no easy solutions. Planner education is a good approach. A planner who doesn't believe in magic bullets, who knows that variability is part of the game and that simplicity beats accuracy is, in my opinion, a better planner.