That blog of mineThe myth that Jakob Nielsen is always right
Jakob is a swell cat. He offers up tons of useful information about web design on a regular basis, and for that, I’m grateful. But today, he’s just dead wrong.
Today’s Alertbox article is called “The Myth of the Genius Designer”. Sadly, it’s crap.
The whole article leads up to the following three points:
“To summarize:
- For a good starting point, get a good designer.
- To reduce risk, ensure that your designer works from usability data, rather than guesses.
- To improve quality, use iterative design and polish each round through usability evaluation.”
The first point is obvious - a good designer is better than a bad one. The second point is where I take exception, primarily because of the words “rather than guesses”.
Yes, good designers work off of usability data. But it’s not one rather than the other, it’s both.
Good designers also work from experience. They leverage a designer’s common sense. They work off of hunches. Not guesses. Hunches. There’s a difference.
Hunches are grounded in experience, common sense, logic, and all those other things considered good qualities for a person doing a job. Guesses are, well, guesses. They’re not grounded in anything. A guess is something you get from people who have no real idea what they’re doing.
Jakob’s third point is also problematic, because it implies that great designers don’t iterate. In my experience, all great designers iterate. It’s extremely rare that a perfect design (if there is such a thing) magically appears in a designer’s head so that he can simply document it and call it a day. What happens far more frequently (like, all the time) is that a designer starts with a hunch, turns it into a design, and iterates through it repeatedly until he’s able to sleep at night. He doesn’t stop until the design is as good as he thinks it can be. A great designer is obsesssed with making the design perfect, and he has an accute understanding that perfection is hard to come by. It’s not a given, it’s earned.
Yes, these designs can and should be validated somehow. But usability testing doesn’t always provide the answers, and it’s entirely unnecessary to run a usability study on every project. Great designers know when to rely on patterns and standards, and when to avoid them and do something new. They know how to create something makes sense in the first place. They know that the innovative stuff might benefit from some usability testing, but that the standard stuff is probably a safe bet as long as it’s implemented well. They apply good design principles and create things that solve most of the potential usability problems in the first place.
Jaokb is right that great designers get that way by studying the succeesses and failures of designs. They get that way by doing the work, learning from it, studying user behavior, and so on. And a great designer will continue doing this throughout her career. But she certainly doesn’t need to evaluate usability data on every project. That’s just a waste of time and money.
What Jakob is right about is that great designers are uncommon, and it’s unlikely one of them works for you. But if you have someone you believe would step up and become great if given the chance, make that person a design dictator for a while and see what happens.
You might be amazed. You might see a mediocre designer turn into a great one, simply by giving him ownership and getting out of his way. And you might get better designs with a great designer than you ever would with a rigorous process of usability evaluation.
Jakob isn’t always right.
Posted by Robert on May 29th, 2007
7 comments
Great post amigo. I have been bothered by several of Nielsen’s latest statements. I feel that his focus is stuck on user centered design as opposed to user experience design. In a recent BBC article he comes off as willing to sacrifice the experience for the product design. It seems to me that this is often not the wisest decision. If you have a great, easy to use product with horrible experience (lack of support, slow loading, poor out of box experience, bad design composition, etc.) your product will probably fail.
Consumers want the whole package and that package. I believe that the usability person must attempt to find a harmony for all these forces and a total sacrifice of the experience for the sole benefit of product design is going to become an antiquated view.
Well I will get off my soap box for now… and really who am I to say? Again great post!
I’m sorry, but I don’t believe that Jakob is ever right. Anyone who designs a site that looks like this (http://www.useit.com/) and claims that it’s usable has a very different definition of usability than I do. Now *this* site is usable, and your book is *fantastic*!!!
Thanks for the kind words.
I think there is definitely a difference between usability and user experience. While Nielsen’s site is “usable”, it definitely lacks a compelling user-experience. Sure, the content makes the site worth wading through, but the lack of design effort simultaneously makes it worth complaining about.
One can only hope that usability never completely overtakes the user experience. In my eyes, a good user experience depends on high usability, but usability should not be given more weight than the whole experience.
You sounds like you’re panicking LOL. Jakob is right, the era of “hunch” is over.
“What happens far more frequently (like, all the time) is that a designer starts with a hunch, turns it into a design, and iterates through it repeatedly until he’s able to sleep at night. He doesn’t stop until the design is as good as he thinks it can be.”
And? Still it’s just in his precious little mind, maybe it will work, maybe not. What he thinks and the reality can be two different things. Testing will prove it or not.
The result of this bullshit is the useless artsy-fartsy Flash crap on the Web. Game over.
Thanks for your comment.
“What he thinks and the reality can be two different things. Testing will prove it or not.”
Sure, they can be two different things, but a good designer should be able to solve the major issues in the first place with a smart design. And usability testing is no silver bullet. Useful, definitely, but flawed.
Even with really good usability testing, it takes a good designer to figure out how to fix the problems without damaging other parts of the experience. And she’ll do that using her experience, her common sense and intelligence, and her hunches. Usability testing is just one of many tools.
Also, the era of the hunch is far from over. Most companies still don’t allow time for usability testing or user research. In a huge amount of cases, a designer’s hunch is all the company has.
But it’s obvious: A designer does no usability testing. Users must do this. Designers may have good patterns and heuristic evaluation skills. But of course they (software engineers too) have a blind spot in usability. Therefore usability tests.
What worries me is the underlying tone in Nielsens words: Designers have no rigor in their work. Usability experts have.
[...] almost sounds like two mutually exclusive ways of designing: one can design from vision trusting in one’s own intuition, or one can design from facts, trusting in the results of user or market [...]