Why Usable Design Matters: Most Products Fail the First Time

Usable design is not the icing on the cake. It's more like the batter.

Consider this cautionary tale of the talking soda machine from the classic The Psychology of Everyday Things  by Donald A. Norman (1988). Norman recalls a conversation he had with a designer about the high stakes of getting design right.

It usually takes five or six attempts to get a product right. This may  be acceptable in an established product, but consider what it means in a new one. Suppose a company wants to make a product that will perhaps make a real difference [emphasis added].The problem is that if the product is truly revolutionary, it is unlikely that anyone will quite know how to design it right the first time; it will take several tries. But if  product is introduced into the marketplace and fails, well that is it. Perhaps it could be introduced a second time, or maybe even a third time, but after that it is dead: everyone believes it to be a failure.

I asked him to explain, “You mean,” I said, “that it takes five or six tires to get an idea right?”

“Yes, he said, “at least that.”

But I replied, “You also said that if a newly introduced product doesn’t catch on in the first two or three times, then it is dead?”

“Yup,” he said.

“Then new products are almost guaranteed to fail, no matter how good the idea.”

“Now you understand,” said the designer. “Consider the use of voice messages on complex devices such as cameras, soft-drink machines, and copiers. A failure. No longer even tried.  Too bad. It really is a good idea, for it can be very useful the hands or eyes are busy elsewhere. But those first few attempts were very badly done and the public scoffed—properly. Now, nobody dares try it again, even in those places where it is needed.  .

Scary? Yep, sure is when you think about all the people trying to make new products for the developing world with few resources and volunteer teams. This ragtag approach has the  potential to generate cyncism about technology among the very groups and individuals the product is intended to help. And if it takes five or six times to get something right, how can a Digital Vision fellow do it right once?  However, perhaps the troubling record of new products suggests some valuable opportunities.

1. Improve upon what already exists. If someone else has built a prototype or just a few versions of a product, it probably isn't as usable as it could be. Making an improvement on just one set of features could make a huge difference.
2. Learn from others' mistakes. If if you are making a whole new product, try to find failed earlier versions. Consider those your earlier versions. Usability tests can run on existing systems so you can identify how past products didn't meet users goals or help them complete critical tasks.

3. Test, test and test that app again. It's a lot easier with software and web sites than with soda machines to build multiple versions. Before you put your prototype out, test it with real users trying real tasks to meet real goals. Run a round of testing on each version so you can release v.6 instead of v.1 or v.2 and have a much better chance for success.

Good product design is necessary for a techology's adoption and to produce a community of users  and, in the case of open source applications, developers. 

So before you start to bake your cake, take some time to think about what recipe you will use for that batter.

Question: Can anyone comment with suggestions on what improvements can be made to existing technologies appropriate for the developing world? What about failed products that should have worked, but didn't because the design wasn't user-centric enough?