The milkshake that competed with bananas and bagels

The story has been told in classrooms and conference talks for twenty years, mostly in the version Clayton Christensen gave it. It appears in “Marketing Malpractice”, the 2005 Harvard Business Review article he wrote with Scott Cook and Taddy Hall, and again in his 2016 book Competing Against Luck, where he credits the research to Bob Moesta and a colleague. A fast-food chain wanted to sell more milkshakes. It profiled its milkshake buyers, asked them whether they wanted the product thicker, cheaper or more chocolatey, and adjusted it to the answers. Sales did not move.

The researchers then changed the question. They spent time in a restaurant watching when milkshakes were bought and what was going on in the buyer’s life at that moment. Many were sold early in the morning, to people alone, on their way to work, who took the shake to the car. Asked why, these customers described a long and boring drive, a free hand, and the need to hold off hunger until later in the morning. They had considered a banana, a bagel or a doughnut, and each fell short: gone too fast, too dry, too messy. A thick shake through a thin straw lasted the whole commute. The chain is not named in Christensen’s telling, and the details shift slightly between versions, so the story is best read as a teaching case.

The rivals were in the breakfast aisle

The first lesson is about competition. Inside the chain, the milkshake’s competitors were other milkshakes and other desserts. For the commuter, it was competing against everything that could make a dull drive bearable and get them to lunch. That set had nothing to do with the menu category.

The method calls this reading jobs to be done: customers hire a product to make some progress in a particular circumstance, and they will fire it without sentiment when something else does the job better. A company that defines its market by product category measures itself against the wrong rivals and improves the wrong things. More chocolate makes a better dessert. A commuter needs a slower one.

Watching beats asking

The second lesson is about method. The chain’s first research asked people about the product, and people answered about the product: sweeter, cheaper, thicker. None of them would have said “something to occupy me on the highway”, because that is not how anyone thinks about a milkshake. The job only appeared when the researchers looked at circumstance (time of day, being alone, the car) rather than at demographics or preferences.

That is a useful discipline for any team writing a value proposition. The premise of an offer, who it is for, is much sharper when it names a situation. “Commuters alone in the car with a long drive ahead” tells a product team what to change. “Adults aged 25 to 45 who like sweet drinks” tells it almost nothing.

One customer, two jobs

In some of his tellings Christensen added a twist that is often left out. The same people bought milkshakes in the afternoon for a different reason: as a treat for a child, where a parent wanted to say yes to something small. That job called for a different shake, thinner and quicker to finish, so the family was not waiting while a child fought with a straw. The same cup of ice cream does two different jobs before and after lunch, and a single product brief cannot serve both. Averaging the two would have produced a shake that served neither.

Handle the story with care

Because the anecdote is so tidy, it gets quoted as evidence, sometimes with percentages and a named chain attached. Neither comes from a published study, and practitioners such as Tony Ulwick have argued that the story frames the job loosely. We use it as Christensen did: an illustration that makes an idea easy to remember. The proof of a job always has to come from your own customers, observed in their own circumstances.

If your team still defines its competitors by the shelf it sits on, train your team to observe the jobs your customers are actually hiring you for.