H053

How does a 10% rise in sales turn into a factory crisis?

Shoppers buy a little more. The store orders extra, just in case. The distributor sees a jump and orders more still. By the time it reaches the factory, a ripple has become a wave.

Shoppers
+10%
a small, lasting rise in what people buy
Each step back
adds a cushion
and reacts to the orders it sees, not to the shoppers
The factory
boom, then bust
big swings, then idle lines and piled-up stock

One small change, four very different stories

Illustration: weekly orders at each step after shoppers start buying 10% more

50 100 150 200 wk 0 wk 4 wk 8 wk 12 wk 16 wk 20 wk 24 Shoppers buy Store orders Distributor orders Factory makes shoppers buy 10% more units/wk

Each link in a chain sees only the orders from the link below it, so small changes grow as they travel. A store that sees sales rise orders a little extra to be safe. The distributor sees that bigger order, not the shoppers, and adds its own cushion. The factory sees the biggest jump of all, ramps up, and then gets almost no orders while everyone works through the extra stock. Nobody did anything foolish; each one reacted sensibly to the signal in front of it.

This pattern, called the bullwhip effect, shows up whenever decisions pass through layers: supply chains, hiring, housing construction, even rumors. The cure is usually to let every layer see the original signal, not just the one next to it.

Try it somewhere else

During the spring of 2020, shoppers bought somewhat more toilet paper than usual for a few weeks.

Why did store shelves sit empty for weeks, and why were warehouses overflowing later?

Where the name came from

Procter & Gamble noticed it with Pampers. Babies use diapers at a steady rate, so store sales barely changed, yet the orders P&G received from distributors swung up and down sharply. Researchers Hau Lee, V. Padmanabhan, and Seungjin Whang described the pattern in 1997 and named it after the way a small flick of the wrist becomes a big snap at the tip of a whip. MIT's "Beer Game," played in business schools since the 1960s, recreates it with four players and almost always produces it.

About the chart

The lines are an illustration of the typical pattern, not data from a real company. Each step reacts a week after the one below it and orders a cushion on top of what it sees.

Related exhibits

Sources: Hau L. Lee, V. Padmanabhan, and Seungjin Whang, "The Bullwhip Effect in Supply Chains," Sloan Management Review, Spring 1997 (Pampers example). MIT Sloan School of Management, the Beer Game (Jay Forrester, 1960s). Chart is an illustration.