Thanks to Jen Dehaan, royalty of the Adobe writing teams, there is now a fresh excerpt from the book up in the Adobe Design Center.

The excerpt, called “Interface surgery: Converting an implementation-model design into a mental-model design”, was taken straight from Chapter 3 of Designing the Obvious. If the sample chapter hasn’t satisfied you, check out the Adobe article.

Mark Schraad (AOL Marketplace), Sarah Bloomer (Sarah Bloomer and Co.), Christine Wodtke (Boxes & Arrows, Cuccina Media), and myself discussed Why We Should Ignore Users in an almost-booked-solid panel session in the big room at SxSW yesterday. We talked about the merits of user research, creating personas and scenarios, and we talked to audience members about their concerns with various approaches. Audience members were also given red and green cards they could hold up to agree or disagree with panelists as we spoke.

It was quite a lot of fun, and I’m really happy (and surprised and impressed) that so many people managed to make it to a panel at 10am after a Saturday night on Sixth Street in Austin, TX.

There are a few typos in this partially-complete transcript from Fixin’ Supper, but the panel ran for an hour, so I’m amazed someone was able to make as few typing mistakes as FS did. (Thanks, FS!)

If I see a podcast for the panel, I’ll offer it up on the blog as well.

In T-minus-3 hours, I’ll be headed to SxSW 2007 with the wife, a handful of copies of the book to give away, and a six-minute-40-second presentation (for the beginning of the panel discussion).

My panel, Why We Should Ignore Users, is on Sunday at 10am, and I’ll be doing two book signings as well.

The first one is Saturday at 4:30pm at the Day Stage Cafe’, and the second is Sunday at 3pm at the Peachpit booth in the SxSW Bookstore, where I’ll be signing books alongside Mr. Jeffrey Zeldman and Mr. Brendan Dawes.

I hope to see you there!

Many designers and developers believe that if they can get their mothers to use a site correctly, or even decently, they’ve succeeded in making the site usable. The idea has always seemed to be that the developer’s poor mother is aparently the worst web user on the planet, or at least close to it.

The thing no one mentions is that a lot of people have done this test, and not all of these mothers can be the web’s worst users.

What does this mean?

Mothers are not the tried-and-true test of usability because they’re the worst users. It’s because they’re typical users.

Many mothers are no worse off in computer-savviness than the average banker or lawyer or fly-fisherman. Many mothers, even those with less than 3 months experience using the web, are no worse off than people who have been using the web for 2 years.

Why? Because it doesn’t matter how long you use the web, it matters how much you understand about it. And most people never need to know how or why this big mess of modern technology we call “the web” works. They only care about getting things done.

Next time you run the Mother Test, remember afterwards that there are several million other people out there that do the same things your mother does. They make the same mistakes, they have the same misconceptions, and they care just as little as your mother about being proficient on the web.

If you design commercial software or web applications, those are your users.

I’ve said in many a presentation and many a conversation (and even this article) to avoid using long bits of text at the top of a screen to provide instructions to a user about how a page is supposed to work or what they’re supposed to do. I suspect many people have doubted this tip. After all, how else can we explain it?

This article, from UXmatters, demonstrates the point nicely and offers some best practices.

Specifically, it describes how a user skips right over instructions because her eye is drawn immediately to a control. She is in the mindset of doing, not reading. To resolve this, break up the instructive text into smaller bits and position those bits where they help the most: beneath (or very near) the fields to which they relate. This way, users will actually see the instructions. If not before they enter data into a field, they’ll see it afterwards and can correct any mistakes.