Playing the Game
I recently had the pleasure of participating in the Beer Game hosted by DePaul University in Chicago.
I have played — in fact, even taught — the game, but admittedly it has been a while. Even for so-called seasoned supply chain planners, this was a very educational experience. The attendees were from leading consumer products companies who sell to major retailers (Walmart, Target, CVS, Walgreens and major grocers). We were led by none other than Andre Martin, author of Flowcasting the Retail Supply Chain, as well as by DePaul supply chain leader Kevin Smith and some super smart supply chainers from JDA.
To play the game, two teams were pitted against each other to see who could best keep costs down and service levels high. Within the teams some tried to ply their supply chain expertise to attain these goals. Interestingly, both teams wound up with high excess inventory in the chain. One team had shortages at retail (very bad) and lower supply chain costs. But of course, without product at retail sites, you lose sales! As in the first time I played this, frustration over the lack of control in the chain (no matter how intelligent or experienced the player) was apparent in these game players, too.
Since Flowcasting first was introduced into the lexicon of retail supply chain, a lot has changed. Cloud supply chain technology and good B2B communications allow us to share data across the supply chain, which was a real challenge before. Yet it appears that in most of the world wholesalers and manufacturers still create their own forecast and pay scant attention to any forecasts they get from retailers. Admittedly, very many retailers don’t send forecasts and don’t provide decent data to their suppliers. The idea of using just the retail forecast isn’t really new, but it is one manufacturers haven’t paid attention to before. Manufacturers spend millions of dollars each year on super smart software to second guess the retailer. But is it really working? The record is mixed. And today, more retailers are getting their data act together to provide more transparency to their suppliers. Globalization and Omni-channel in particular are driving a rethinking of how suppliers’ chains are managed.
In the workshop, Andre Martin shared his experiences in implementing Flowcasting across multi-stage chains. Then the group played the game again using the new software which provides visibility to the multiple data points that planners need from POS, inventory levels, lead times, and so on. One principal of Flowcasting is don’t forecast what the system can calculate. The tool can look at the input data and create more accurate plans for the multi-tier players in the chain.
Simply, the new Flowcasting software sits on top of other forecasting modules or is incorporated into S&OP. Flowcasting allows for visibility as well as live simulation, if the parties choose to, between multiple parties. I envision Flowcasting being used in web conference calls between trading partners to evaluate not just the weekly or monthly plans, but promotional planning and new product introductions as well as for re-assessing and changing policies to be more compatible with the way in which the partners run their processes.
Collaboration and Clarity
Flowcasting, as Kevin Smith told me, is about a change in behavior, since although you will improve your forecast, it will never be perfect. The key is visibility into the dynamics between the customer, your firm, and your supplier data and being able to balance that dynamic rather than attempting to manage each entity’s discrete activities.
One lesson I can share, too, from repeated practice (my own supply chain project experiences) as well as the many research projects we have done is: more frequent current data cycles, and more granular data, and reduced information cycle time in forecasting from tier to tier makes a significant difference in customer service, inventory levels, and of course, forecast accuracy for the suppliers.
Often, a lot of attention is paid to the weekly buckets and being able to meet weekly demand. But our experience demonstrated that once companies start evaluating demand data, they can make other changes such as revaluating safety stock levels, re-negotiating lead times, and so on. Over time, they can deal with more fundamental issues such as lot/batch sizing, capacity, and re-optimizing their outbound networks. With Omni-channel now a key mandate from brands and retailers, having more frequent granular data with location/node intelligence is essential to ensure accurate available-to-promise without fattening up on inventory.
Some lessons learned from our exercise with Andre Martin:
- More frequent re-planning is being used to compensate for forecast inaccuracy at the retailer level. The challenge, Martin pointed out, is in connecting the multiple layers. Thus, participation in collaborative exercises is essential not just for regular planning, but also for use cases such as promotions, VMI/safety stock agreements, and new product introductions.
- Flowcasting provides each organization with a window into the overall supply chain dynamics. Businesses are able to see their business rules and policies and determine if they really are appropriate.
- Flowcasting reduces bias. We make many assumptions about other organizations’ data and behavior. However, are these assumptions based on reality? Often, we do not understand the requirements a partner has or the constraints under which they operate. Understanding these factors will allow us to change policies and terms to make the relationships work better.
_ _ _ _ _ _ _ _ _ _ _ _
To view other articles from this issue of the brief, click here.