Snap! That’s the bullwhip effect.
When a business orders inventory to cover an expected increase in demand, and the increase tapers off faster than expected, they are left with extra inventory. Their distributor has a worse problem; the distributor sees a sudden demand spike followed by orders that drop to NOTHING while the inventory bubble is sold off. The manufacturer sees a stronger demand spike and a longer bubble with no orders, etc.
This amplification of small variations in demand is know as the bullwhip effect, and the effect on upstream manufacturers and distributors can be devastating. It can and does result in businesses being destroyed. I wrote a paper on the bullwhip effect for my Operations Management class and I got to thinking about it again when I read this headline: Motorcycle Sales Sink More than 40%.
The bullwhip effect has several contributing causes. Probably the biggest problem is that most businesspeople are not statisticians. They tend to model changes in demand as linear. So if demand went up by 100 units last month, and up by 100 units this month, then it’s probably going to go up again by 100 units next month once known seasonabl factors are substracted out. Distributors and manufacturers often make similarly un-damped estimates based on the amplified demand that they see in their orders, and worse, their data tends to lag the actual effect. So they are forecasting up and up at the same time the demand spike has started tailing off.
That communications lag is also a key contributor. Businesses that estimate future demand based on monthly or quarterly data may not see the key indicators that the demand landscape is changing, particularly if they have to plan manufacturing and distribution months in advance. A decision that seemed fantastic two quarters ago can be disastrous when it becomes apparent that the demand for a product is falling back to pre-spike levels.
A third contributor is self-interest. As a retailer, I want to make sure I have enough units in stock to meet demand. If somebody comes in the front door and asks for product X, and I don’t have any X left this month, then that’s a lost sale and lost profit. And I’m paying my sales staff to stand around doing nothing.
Now let’s say that demand for X this month is 100 units more than expected, so I send 100 customers packing. Planning for next month, I order 200 more X, on the assumption that demand will rise linearly. When the inventory arrives, to my shock, they’ve only given me 120 units of X. The distributor wasn’t ready for the spike, and sends word to the manufacturer that demand is outstripping supply.
Next month, if the demand for the product hasn’t tailed off, what will I, the retailer, do in my own self interest? I will over-order. Planning for 300 more units of demand next month, I may order 400 or 500 units of X, on the assumption I won’t receive all of them. If I don’t receive my whole order, I’m likely to continue to inflate my orders.
Now go upstream to the manufacturer. When they get a quarterly report from the distributor indicating that the distributor is out of units and the demand is thousands and thousands of units more than the distributor has been getting, the manufacturer is likely to spin up production. New machines, new lines, new workers, maybe even entirely new factories.
Now when the demand tails off — or worse, starts trending to pre-spike levels — the entire infrastructure has been upgraded to handle vastly more demand. Because everyone was acting in their own self-interest (“I need to have enough X to cover next period’s orders”), the manufacturer and distributors are now in serious trouble. Retailers will simply sit on their existing inventory until the bubble has passed, but the manufacturer now has all new production lines and the distributor is sitting on warehouses full of product that they can’t move.
So how do you solve the bullwhip effect? By far the most effective means is to use good forecasting methods combined with accurate communication between the retailer and all levels of the value chain. If the manufacturer has a clear window into real consumer demand, they can spin up the right resources, rather than plan for an apparent spike that may never materialize. Same for the distributor. And if retailers are given accurate information about the amount of product that is available and the amount they can expect when they order, they won’t be inclined to inflate their orders.
Companies like Wal-mart, which have close relationships with manufacturers, control the entire distribution system and have extremely accurate information on customer demand at the point-of-sale are in an enviable position to avoid the bullwhip effect.