Like most good books, the author doesn’t try to strive for a unified theory of everything.  Instead, he’s practical, full of good advice on both what to do and what not to do, and tries to summarize things into easy-to-remember principles.

And like most good books, this one makes you realize you’ve made a bunch of mistakes in the past.  Moreover, it makes them seem obviously amateurish.  But it’s good to get a head-smackingly-obvious message once in a while.

It’s really tempting to think a bunch of smart people in a room can design something amazing.  The truth is that as you get more and smarter people, you get further away from your average user / customer / visitor, so your design is likely to get worse, not better.  We have to avoid this designer-self-selection sort of bias.

Another key is that good design is just like general agile development: you have to iterate on it.  Don’t agonize over getting a perfect design out the door.  Get something out, test with real users (not yourself, not employees who know the product really well), take their feedback, implement the obvious things (don’t stress over nit-picks), and repeat.

Duh.  I know.  Read the book 😉