Why the best product, on its own, isn't enough to win

More than 4,400 mousetraps have been patented in the United States. According to research published by American Heritage in 1996, fewer than two dozen ever made their inventors a cent. The wooden board with a spring, the one you can still find at the hardware store, was first patented by William C. Hooker in 1894. A variant of it, filed by John Mast in 1899 and patented in 1903, is still made today in Lititz, Pennsylvania, under the Victor brand. Even the line that convinced generations of inventors, “build a better mousetrap and the world will beat a path to your door,” has an odd history: Ralph Waldo Emerson never wrote it. The mousetrap version appears in 1889, seven years after his death, in a book of quotations edited by Sarah Yule.

Four thousand better mousetraps didn't move the market an inch. The reason holds for many industries, and it is where Category Design starts.

Why does the best product often fail to win?

Because nobody buys a product in the abstract. Before choosing, customers decide what they are buying: “a mousetrap,” “management software,” “a watch.” Then they compare the options inside that box, with the criteria the box brings with it. Whoever wrote those criteria, usually the first company people remembered, starts with an advantage that a technical improvement rarely makes up.

That is how a company can be right about everything and still lose. Its trap snaps faster, its machine uses less power, its service answers sooner. But if customers measure everyone by the leader's yardstick, those advantages become spec-sheet details, and in the end the decision comes down to price.

Category Design works on this step: instead of asking how to win the comparison, it asks what customers should be comparing in the first place. We explain it in full in the guide on what Category Design is. Here we look at two cases where the mechanism is easy to see.

What happens when you compete on who does the most?

In August 1993 Apple put the Newton MessagePad on sale for $699. It was an ambitious device: a small handheld computer that recognized handwriting. It sold 50,000 units in its first three months, but handwriting recognition in the early versions was unreliable and often forced users to rewrite and redraw.

In March 1996 Palm Computing, which had meanwhile been bought by U.S. Robotics, introduced the Pilot. It did far less than the Newton. To write on it you had to learn a simplified alphabet, Graffiti: a bet, as the Computer History Museum puts it, that people would be willing to change how they wrote in order to be understood better by the machine. According to the Computer History Museum it cost $299, fit in a shirt pocket and synced with a home or office computer at the press of a single button.

During development Jeff Hawkins of Palm carried a block of wood the size of the Pilot into meetings and took pretend notes on it with a chopstick as a stylus, counting the steps the most common tasks required. The museum sums up the result this way: the PalmPilot succeeded because it was an accessory to the computer, not a replacement for it. U.S. Robotics sold about a million in the first 18 months. Apple discontinued the Newton line in 1998.

Our reading is that Palm didn't beat the Newton in the Newton's race. The Newton asked to be judged as a computer, and as a computer it was small and imprecise. The Pilot asked to be judged as a pocket organizer connected to the computer, a box where it was nearly alone and where the Newton's flaws didn't count.

How do you change the basis of comparison?

In the 1970s and 1980s Swiss watchmakers went through what is known as the quartz crisis: battery-powered watches made in East Asia were cheap and put the most established European brands under pressure. In our view, on accuracy, the ground on which Switzerland had built its reputation, quartz had made it hard to stand out.

On March 1, 1983, the first Swatch collection launched in Zurich: twelve plastic models, priced between 39.90 and 49.90 Swiss francs. An integrated production technique cut the number of parts from about 100 to 51. The name, according to the brand's history, is a contraction of “second watch”: not the good one you keep for life, but one to match your clothes and change. By the end of 1984 about 3.5 million had been sold.

Swatch could have tried to make a quartz watch slightly more accurate or slightly cheaper than the Japanese ones, and fought on ground where it started at a disadvantage. Instead it moved the question from “which watch tells time best?” to “how many watches do you want to own?”. On that second question, the comparison with the old competitors almost stops making sense.

So product quality doesn't matter?

It matters, but it does a different job from the one it is usually asked to do. Quality is what keeps the promise standing: the Pilot really had to sync with one button, and Swatch had to work even though it was cheap. If the product betrays what the category promises, customers notice on the first purchase and no new name will hold up.

What quality doesn't do is choose the race. A company that improves its product without asking which box it is being judged in risks becoming mousetrap number 4,401: better, patented, and unknown.

How can you tell if your company is playing on the wrong field?

In our work with Italian companies, the signs that come up are almost always the same. You lose customers on price to competitors you consider worse. Customers describe you in the industry leader's terms: “you're like them, but…”. You do something nobody in your market talks about, a delivery time, a service, a way of working, and you give it away without ever mentioning it. We have collected five of them in a separate article on the signs of the wrong category.

For us at Project Mantra, the question to ask is easy to write and uncomfortable to answer: if tomorrow your customers judged you on criteria you chose, what would they be? If the answer is “the same as now, only a bit better,” the work to be done concerns the category more than the product.

The Newton cost $699 and did more. The Pilot cost $299 and did less. It was the Pilot that sold a million units, and the Newton that was pulled from the market.