The first one-button mouse was a roll-on deodorant and a soap tray.

A few days after Jay Elliot agreed to join Apple, Steve Jobs called him at about nine o'clock on a Friday night. "Let's take a ride tomorrow," Steve said. "Meet me at the office at 10 a.m."

That was about all the explanation Jay got.

On Saturday morning they got into Steve's silver Mercedes and drove to Xerox PARC in Palo Alto. Steve had already been there with a group of Apple engineers and was going back for another look. Jay spent the drive guessing where they were headed. Maybe Steve wanted to show him a car. Maybe an office building. None of his guesses was close. In It's the Product, Stupid, he admits it plainly: "Up and through that whole period of time, I had absolutely no idea what the hell we were doing."

Three buttons at the end of a printer

PARC walked them through the world Xerox knew: copiers, printers, big equipment. At the end of a large printing system sat a workstation with a pointing device. Jay remembers it as "big and awkward and had three buttons, but the movement on the screen was what mattered: you could point, choose, and move information without typing a command for every action."

Jay is careful about the credit here, and I want to be too. "The lesson is sometimes told as though Apple invented everything it saw there. It did not, and I want to be fair." Xerox's researchers had built something remarkable. What changed, in Jay's words, "was the question Steve brought into the room."

The finger

Walking back to the car, Steve held up his index finger and said, "This is going to change the world."

He wasn't talking about the device. As Jay puts it, "What excited him was not another piece of equipment. It was the finger." A person could point at information and move through it without learning the machine's language first.

When Steve got excited, Jay says, his voice went up octaves. "He sounded like a kid who had just been given a new bicycle." But the excitement didn't stay excitement for long. It turned into a demand on the product.

The pharmacy run

They were back in Cupertino around one in the afternoon. The mouse they'd seen had three buttons. Steve wanted one.

So he called in one of the engineers and sent him to the pharmacy for a roll-on deodorant stick and a soap tray. The instruction, as Jay tells it, was "we want to build a mouse that only has one button, so start with that."

"The guy went and bought them," Jay writes, "and that was the first prototype."

The deodorant gave them a ball that rolls. The soap tray gave them a shape to rest a hand on. Put together, they gave the team something better than a description. In the book's words, "The ordinary objects gave the engineers something more useful than another speech about ease of use."

Jay doesn't oversell what the model proved. "The model did not prove that a one-button mouse could be manufactured, but it made the choice visible enough to argue with." And he doesn't apologize for how it looked: "The ball from a deodorant bottle did not look like the future, and it did not need to. It did its job anyway."

Why one button was the harder choice

Cutting three buttons to one sounds like less work. It was more. Jay knew one button "would force more difficult work inside the machine." Somebody had to decide what that single click meant in every situation.

That was the point. "The engineers still had to solve the complicated problems," he writes. "We simply refused to send them home with the user."

Once the model was on the table, the argument changed. It stopped being a general engineering discussion and turned into three specific questions, the ones Jay lists in the book: "What must the software decide? Which action should be obvious? What would you lose?"

Make the thing, then look at it

This is the habit I'd steal from this story. The book puts it in seven words: "Make the thing, then look at it."

Most teams argue about products in documents and slide decks. A roadmap can promise "simple onboarding" for a year and nobody can argue with it, because there's nothing to hold. A soap tray you can argue with.

So ask yourself: what's your roll-on deodorant?

1. Pick the one decision your team keeps talking around. The confusing first screen. The pricing page. The setup call.
2. Build the cheapest version someone can hold or click. The book is specific about what counts: "Cardboard, a drawing, a draft, or a screen recording counts."
3. Put it on the table and ask the three questions the soap tray forced. What must the product decide? Which action should be obvious? What would you lose?

Jay has a test for whether your big idea is doing any work: "If the vision cannot change a physical decision this week, it is still only language."

A deodorant ball and a soap tray passed that test. Yours can pass it this week.