Jobs to Be Done, Explained With the McDonald's Milkshake Story

Short answer: jobs-to-be-done asks what job a customer is hiring your product to do, not what your product is. The clearest illustration of the gap between those two questions is a McDonald's milkshake that a flavour overhaul couldn't sell — until someone asked buyers why they bought it instead of whether they liked it, and found two completely different jobs hiding inside one product.

What "jobs to be done" actually asks

The question is not what is my product? It is what is the job you're hiring this product, service or tool to do?

Buy nails and a hammer and you are not buying nails and a hammer. You are buying the ability to build a treehouse. The nails are incidental. The treehouse is the job.

This sounds like word games until you try to improve a product with it. Ask "what is my product?" and you get answers about features. Ask "what job is it hired for?" and you get answers about the customer's day — where they are, what else they considered, and what would count as the job being done badly.

McDonald's spent millions on a milkshake nobody wanted

McDonald's invested millions of dollars in a new milkshake recipe. New flavours, all the studies done, all the research saying people seem to love this flavour the most. Then a large campaign to push it.

They launched it. They spent all the money on the campaign.

Sales didn't budge at all.

So they brought in a researcher who specialised in jobs to be done, and his approach was not another taste test. He stood outside the stores all day, and every time somebody walked out holding a milkshake he asked them one question:

Excuse me — why did you buy that milkshake?

Then he listened to the answers.

The same product, two completely different jobs

The answers split cleanly in two, and the split was by time of day.

In the morning, people were buying a milkshake as breakfast. They were commuting. They could stop in quickly, it wasn't messy, they could drink it one-handed while driving, it took a while to get through, and it was filling enough to hold them until lunch.

The competition was not other milkshakes. It was everything else you can eat in a car:

• A Snickers bar is fine, but it's gone in two bites.
• A banana leaves you holding a peel.
• An apple is messy and awkward to eat while driving.
• A burger — good luck.

The milkshake won that comparison. It was the best available solution for a long, boring, one-handed commute.

In the evening, the same product was being bought by a completely different person for a completely different reason. It was parents. The kid had asked for a milkshake, it was after dinner, a promise had been made — so here, have your milkshake and stop complaining. And now the parents wanted to go home.

Same drink. Same store. Two jobs that have almost nothing in common.

The fix was consistency, not flavour

Once you see the two jobs, the product change is obvious — and it has nothing to do with taste.

Morning: keep the fresh flavours, but make it thicker. A thick milkshake takes longer to get through a straw, which is a defect if you're in a hurry and a feature if you have thirty minutes of traffic ahead of you.

Evening: make it thinner and less fluffy, so the kid finishes faster and the parents can leave.

With those two changes, milkshake sales started increasing — because somebody had finally worked out what job people were hiring the milkshake to do, rather than which flavour they preferred in a survey.

The flavour research was not wrong. It was answering a question that was not the one holding sales back.

What to do on Monday

The method transfers directly, and none of it requires a budget:

1. Go and ask people who already bought. Not prospects, not a survey panel — people leaving with the thing in their hand. The researcher's entire technique was standing outside the door and asking one question.
2. Ask why they bought, not whether they liked it. "Did you enjoy it" gets you a rating. "Why did you buy it" gets you the job.
3. Segment the answers by situation, not by demographic. The split that mattered was morning versus evening, not age or income. Look for the moment, the constraint and what else they nearly did instead.
4. Write down what your product is really competing with. For the morning milkshake it was a banana and a Snickers. For yours it might be a spreadsheet, an intern, or doing nothing at all.
5. Change the product to do the job better — even if the change is unglamorous. Thicker. Thinner. Not a new flavour.

If two groups are hiring your product for two different jobs, you do not have one product with a positioning problem. You have two jobs, and you can serve both — but only once you have named them.

Related reading:

• They Understood What I'd Built. They Still Didn't Buy. — the same gap, one layer downstream: understanding the product isn't understanding the use case.
• Three Sentences That Make Your Product Make Sense — turning what you learn about the job into a message people act on.

Building something and trying to work out who it's really for? Ordana matches SaaS founders with collaborators who already sell to the same customers — joinordana.com.