- Image by Getty Images via Daylife
I just read Justin Fogarty’s (of Supply Excellence) article “Frugal Fatigue” or Consumer Confidence? and I can’t help but sympathize for the S&OP managers trying to make heads or tails of the demand signals they are seeing. Adding “Frugal Fatigue” as an additional consideration makes interpretation even more difficult and reinforces the need for a responsive supply chain. Forecasting in any environment is extremely difficult, but in this environment it almost seems futile.
Forecast accuracy will reveal itself in about one month in the form of either empty warehouses and unsatisfied demand or full warehouses and clearance sales. What do you think?
![Reblog this post [with Zemanta]](http://img.zemanta.com/reblog_e.png?x-id=560f58f9-022a-4998-994d-cbc33218a352)
TrackBack
• Digg This
• Add to del.icio.us
Tags: Demand planning, Forecasting, Sales & operations planning (S&OP), Supply chain planning
Posted in Demand management, Inventory management, Sales & operations planning (S&OP)
You can leave a response, or trackback from your own site.
Here are some of the things we are doing:
1) Abandon statistical models that need lots of history to work. Yes, that means you are sacrificing all that automatic seasonality-finding that Winter’s gives you. But, you have to recognize that you cannot assume the past year or so is any guide to the future.
2) Construct analysis tools that watch for the coming turn: e.g. six month run rate vs. three month run rate. You may not be able to predict it but you may see that you’re in it sooner.
3) Compare your current accuracy to your past. Likely it is worse; model what that means with respect to safety stock. Prepare a very good presentation to deliver the bad news to the management team – “With all the other things you have to worry about, you also have to think hard about reformulating your balance between service and safety stock.” Present in a way that proposes answers for them to pick from. The good news: your competitors have the same problem.
4) Hunt deals. Expend more effort than ever trying to understand and predict exceptional demand. Engage your sales force more aggressively. Take them to lunch if you have to. Convince them that they are a critical element more than ever before. If they don’t help you help them everyone suffers. If you do this and it works, this may be one of the unanticipated beneficial by-products of the recession.
5) Monitor forecast consumption more carefully. This is labor intensive so focus on the critical few. Be more suspicious about apparent outliers (Half way through the month and no sales? Sold the month’s forecast by the 8th?) Chase them down. You may not be able to predict the exceptions but you may be able to reduce your response time to them.
6) Operationally, increase the drumbeat. PO’s that you used to place monthly; place them every two weeks. Po’s that you used to place every two weeks; place every week. Make smaller more frequent bets. Yes, it drives cost but it increases flexibility and flexibility is at a premium nowadays.
These are great strategies. I think it is difficult to give up the notion that history is a good indictor of future results. It is especially difficult if you don’t have a plan in place to react quickly…it sounds like you have some well established processes that are serving you well. Thanks for sharing these.
It would be interesting to see if anyone here continues to find success using more traditional statistical models.