As always, also transferring the discussion from Facebook, since some interesting debates developed there:

Gerald Karam: The banking one is a real winner. I use it all the time. I use prime as well. But it has become less attractive as I buy less online.

Lana Yarosh: Originally, the thing that tipped me into getting prime was the kindle book borrowing thing. But then I got sucked into all the other stuff too.

Jeff Rick: I’ve been thinking of writing a blog post that is related, but almost completely orthogonal to this point. The theme: Apple needs to stop emphasizing design and start emphasizing software engineering. While good design (whether aesthetics or user experience) is great to have, it needs to frickin’ work. When Apple products work as advertised, they are pretty awesome; however, more often in recent times, they don’t work as advertised. I’m pretty heavily invested in AirPlay (2 speaker sets). Most of the time it works, but occasionally it goes into a weird state that requires rebooting the speakers and / or the router. We have only Apple machines, an Apple router and speakers that have been certified by Apple. Everything has the latest firmware / OS. With all that control, there’s no reason that it should fail on a regular basis. If Steve Jobs had my experience, you’d bet he’d be shouting at people to fix it AND it would get fixed. Remember, he was the one who led the switch to OS X because it would be a much more stable basis for an operating system. He was the one who started a pissing match with Adobe largely because Flash would create inferior iPad experiences (too slow, draining the battery) and was responsible for most OS X crashes. While people remember him for design progress, they forget about his insistence of it working. Good design is only good if it also works.

Lana Yarosh: Working well is certainly a major part of the user experience. You can’t have a good user experience when things break down or work too slowly. But I still think that there is a fundamental that’s even MORE important than it working well, and that’s whether it is useful or not. Apple is really good at figuring out what would be useful to people (mostly, by letting others do it first), THEN they are good at making it work better. Though most people equate Apple with good design (again because of the confusion between design and “prettiness”), I think Apple is better at creative borrowing of useful ideas. Somebody still needed to do the first step of figuring out what would be useful to people (where HCI shines).\

Jeff Rick: I’m not sure that we can fully separate these elements to select a most important one. If it isn’t something worth doing, then it isn’t useful. If it doesn’t work, it isn’t useful. If it isn’t usable, it isn’t useful. Maybe a multiplicative metaphor would fit a bit better: VALUE = (QUALITY OF IDEA) x (QUALITY OF INTERFACE IMPLEMENTATION) x (QUALITY OF SOFTWARE IMPLEMENTATION). All parts are important for there to be value.

Jose Zagal: “Apple is really good at figuring out what would be useful to people (mostly, by letting others do it first)” – In other words, they’re not really good at figuring out what would be useful to people…

Lana Yarosh: @Jeff, I agree that all 3 things matter, but I still insist that usefulness is most important. People are much more willing to put up with a lack of reliability or usability if something they are using is very useful (an example here: http://dl.acm.org/citation.cfm?id=1753436) If only we could assign ratings for all three aspects and then do some empirical investigating. Is that anything you’ve seen before?

Jeff Rick: (QUALITY OF IDEA), my term, or usefulness, your term, varies more widely than the other two aspects. In that sense, it has a large effect on value (or is more important). You could think of the other two varying between 0 and 1. The closer they get to being great / working / usable (approaching 1), the more the actual value of the idea shines through. If software engineers and interface designers do a great job, then the quality of the system approaches the quality of the idea.

Jeff Rick: On a different subject, you accidentally struck a pet peeve of mine I disagree with “Somebody still needed to do the first step of figuring out what would be useful to people (where HCI shines).” HCI researchers, particularly those with an H background, often make that claim. It’s just bravado. Other areas of CS (and C-focused HCI researchers) are perfectly capable of coming up with good ideas. All of the groundbreaking work by Doug Englebart, XEROX Parc, Myron Krueger, John McCarthy, Dennis Ritchie, Tim Berners-Lee, etc. was done without HCI methods. Figuring out what would be useful to people does not need to be the first step. For many successful designers (e.g., Philippe Starck), it isn’t the first step. Furthermore, the specific HCI methods of user-based design are not universally good. They can lead to ideas that are simply evolutionary on one side (based on what the user already is familiar with) or impossible on the other side (cannot be implemented with today’s technology). Kristin Lamberty and I published an article about the benefits of starting with what the technology allows (medium-based design).

Lana Yarosh: Link to paper, please? Sounds interesting! By the way, I’m not saying that HCI researchers are the only ones coming up with useful ideas, but just that it’s one way to do it. There are plenty of people doing what I consider good HCI, without ANY training in formal HCI or HCI methods. And I certainly think that many of the great people you’ve listed are examples of that.

Jeff Rick: Note that medium-based design is not a general purpose design method. It is specifically for designing learning environments. http://home.cc.gatech.edu/je77/uploads/3/mbd-ile-published.pdf