Don Norman: Designing For People

Nielsen Norman Group

Why do you like iPod when the shuttle is so bad?

I'd like to ask your opinion on the design of the Shuffle mp3 player by Apple from a usability point of view.

I'm asking this question because BusinessWeek quotes you as saying:

"I've been thinking hard about the Apple product-development process since I left... If you follow my [guidelines], it will guarantee good design. But Steve Jobs doesn't want good design. He wants great design, and my method will never give you that. That takes a rare leader, who can bring both the cohesion and commitment and style. And Steve has it."

I interpret this as "great design requires following the guidelines plus commitment, style, etc." Yet some of the very successful and so-called great products by Jobs and his design team have often contradicted the most basic HCI and ergonomics principles. While I appreciate Job's role in bringing design to tech products, it seems that the Shuffle violates some key HCI principles of feedback & user's control.

It is clear that good/great design is a poor predictor of market share as the eMate and the Cube have shown. So reframing the original question: Despite its potential success with consumers, is the Shuffle a so-called great product despite (or even because) it contradicts your guidelines?

I've just always been puzzled by the double standards that many HCI gurus apply when it comes to Apple's designs. They often give excuses to justify plain blunders, I remember the heated debate over the aberrant GUI design that used the same metaphor to eject a floppy and to erase files and directories. Is it still around these days after two decades of GUI evolution? Or the rather cynical reasons for the persistent one-button mouse. How are less options more options?

That's why qualifying now the blindness of the Shuffle as a 'feature' and a 'choice' is not so surprising. I wonder how this product would fare with usability experts if it came from Sony or Creative rather than Apple. Objectively, I think one should be entitled to point out the failures of Apple. It would give more merit to their virtues, which are many indeed.

Here is a long question (the actual question went on for three lengthy emails. The entire correspondence makes for worthwhile reading, but is too long for this format.

I believe that the question and the resulting debate results from a confusion between marketing decisions and usability design. Basically, the criticism is all about the marketing decisions.

Thus, my questioner asked:

Haven't we gone through enough decades of research to understand that a product with the capacity to play 240 songs (that's about 20 CDs) should provide at least a minimum level of information? I don't believe that any empirical research would conclude that users prefer to not know what song is playing at least once in a while. ... Can one keep a straight face while suggesting that radios and CD players should now come without displays to give users more choices?

But, hey, sure, people prefer to know the name of the song being played, but that's a design tradeoff. Money and space. Displays add to the cost and take space. I believe that the Apple Shuttle is an excellent compromise among the conflicting requirements of simplicity, elegance, size, battery life, and function. If you think of it as a supplementary music player rather than you main one, then the tradeoffs become more clear.

As technology changes, the tradeoffs will change. Thus, Sony has introduced a music player roughly the size of the shuttle, but with a three-line display. But note: My questioner complained that Apple products were too expensive: well, the Sony products are even more expensive.

When we talk about HCI standards, we should not confuse these with marketing decisions. The Apple Shuttle is honest. It doesn't pretend to be other than what it is. If customers don't like the feature set (or lack of features), they won't buy it.

Apple has produced an excellent system. The real product is the system, not any of the individual components. They paved the way for music licensing, they made iTunes an easy to use application, they made interaction of the players with the music players simple and easy (and fast), and they provide a wide variety of music players, with open APIs so that developers can enhance the value.

Nobody else comes close.

If you look at the other music players, there are many that are as attractive as Apple, there are many that offer more features, more memory, and lower cost. But it is the total system that makes the iPod. Some of those others deserve higher market share than they have, but they suffer because of marketing and sales issues and because they do not offer as complete a system solution.

Wake up, HCI folks: HCI is only one small piece of what it takes to make a successful product.

Am I an Apple bigot? No. I can critique their products and their customer service philosophy. But overall, they do better than any other player. Yes, I own an iPod and, yes, I am thinking of buying the shuttle. But my computers (three of them) are Windows based.

If you think the products don't match what you want from a product, don't buy it. But don't try to claim that they do not meet some mythical HCI standards. The role of marketing is to determine what features to offer. The role of HCI is two things: to ensure that the marketing decision is based upon accurate, real information about how people use these products; and then to ensure that the product implementation is as great as possible.

Do iPods have limitations? Of course. So does everything in life.

Essays

Books