8: Handling Usability Arguments
Most discussions about usability are a waste of time - and many who claim that something is better for the user are usually lobbying for something they want for other reasons they don't want to disclose.
"What is usable?" is like a religious question - people have strongly held beliefs that have no basis in fact and cannot be proven logically. The only way to know if something is usable is to test it with users.
(EN: He doesn't even pause to mention that this entire book is his own answer to that very same question. That's not to say the book is useless, but that there are informed opinions and uninformed ones - and the opinion of someone wit ha lot of experience in observing usability testing is worth listening to. But strictly speaking, it's not always right and a test is indeed the only way to sort things out.)
What People "Like"
A favored tactic in these pointless arguments is making allegations about what people "like." Someone will claim that everyone likes videos, or nobody likes buttons, or some other absurd generalization.
In the first place, how do they know? Particularly those with no experience designing and testing websites do not have a clue what people like. Generally, they know what they like, at least in the moment, and claim other people like the same things they do. Ask them to cite the research that backs their claims, and they will fall awkwardly silent.
But even when you do have a sense, based on research, that users like something (or find it to be usable), it's historical data. It's not what people like, but what some people used to like in a specific situation. It doesn't guarantee that the same thing will be effective for the specific people who will, in the future, use your specific site for a specific task.
As in all debates, it's best to check your emotions at the door, and be ready to prove what you say. But not everyone has the discipline to do this.
Competing Agendas
Krug mentions that most development teams who come to the table to discuss a web project are from different areas, and each has its own agenda. Even if they set their personal tastes aside, they have different objectives and will argue for design choices that suit their agendas, or the agendas they want others to support.
The CEO wants something that's going to impress the board of directors, a developer wants something that is easy to code and maintain, the marketer wants something that hustles people to purchase quickly, and the designer wants something that's visually appealing.
(EN: This goes on a while, elaborating on what some of the areas are interested in, but the point is made.)
The Myth of Average
When firms realize that the battle of personal opinions and departmental agendas is going nowhere, they turn to research. They want to know what the average user is really like.
The problem is: there is no such thing. All Web site users are unique, and even the behavior of a single individual will change depending on what he is attempting to do at the moment.
Focus groups of users are also flawed in that a person describing what he would do to a group of strangers, knowing he is being observed and is recorded, is not at all what he is really going to do when he is not in that situation. He will describe an idealized version of himself - making responsible, intelligent, and well-considered choices. This has nothing to do with his real behavior.
The author suggests that testing is the only way to get past that.
(EN: He does not subject usability to the same scrutiny - and it deserves it. Usability testing is still fake and affected - it is a person who is pretending to do something on a model of a site, and he is still acting in an idealized manner because he's pantomiming the task and aware he is being observed. It's a little better, but not much ... and until you can observe real people doing real tasks, it's often the best that can be offered.)