Brian's Reviews > The Design of Everyday Things

The Design of Everyday Things by Donald A. Norman
Rate this book
Clear rating

's review

really liked it
bookshelves: principal-engineer-reading-list
Recommended to Brian by: Michael Economy

(4.0) Some good stuff in here, though it's certainly dated

I'll be looking up some of his other books to see if he's as good at predicting and suggesting product improvements as he was back then.

I think he makes concrete some really common sense ways to approach and analyze designs of products that humans use. It's certainly entertaining to point out ridiculous products, interfaces etc., but that's kind of 'negative design': what not to do. That doesn't actually help you do it right. Fortunately, he does spend a fair amount of time on how to do it right. So some good stuff to summarize:

* Make the controls/interactive elements visible: they won't used if they're not noticed
* Use cultural, intuitive clues to suggest the function of elements...e.g. choose materials, shapes, colors appropriately
* Try to make the interactive elements map to the functions they perform, particularly easy if there is some spatial component to what is controlled, arrange the elements in the same arrangement as their actions
* Think of the steps that will occur when a user interacts with a product:
-- user forms a goal
-- user translates the goal into smaller, more concrete 'intentions'
-- user enacts the intention as best he can
-- the system responds
-- the user tries to interpret the response
-- 'bugs' can occur anywhere in this chain, identifying the source can help identify the solution
* Make sure there's feedback when user does something. If the product's state has changed, make sure user can tell. Readable displays can be helpful here
* Users will make mistakes. Expect them and make them reversible, low impact
* If you need instructions for new user to operate, you're probably doing it wrong
* Try to use constraints to limit the wrong actions user can take
-- e.g. 3.5" floppy disks can only be inserted one way into drive (though they look square and top not much different from bottom)
-- cultural constraints can be used as well as physical constraints
* as last resort, turn to arbitrary standards (so even if something not intuitive, user only has to learn once and can apply to all similar devices)
-- e.g. QWERTY keyboard
* how to use technology best to improve products/processes:
-- simplify tasks, but leave them largely the same (don't automate away key steps that users will forget occur and can't troubleshoot when something's wrong)
-- make things visible that weren't visible before so state is easier to track
-- design for error, don't blame "human error" when unintuitive/broken interface leads to disaster (e.g. three mile island)

So I think a few of these can be explicit steps to take when evaluating a design:
* are the relevant features visible? is feedback visible?
* is it clear what the mapping is from interface to resulting actions?
* when user interacts with each control, is there appropriate feedback?
* can user identify when he's made an error? can he undo the error? are the 'human errors' ever catastrophic?
* are there constraints you can apply to reduce possibility of error?

He also made some cool predictions/product requests, which makes me want to read some of his more recent stuff:
* the windows/macintosh user interface would take off
* calendar/reminder book would be electronic and fit in your pocket (but didn't think they'd BE the phone...talked about connecting the calendar to the phone...but well on his way to asking for the smartphone)
* how big hypertext would get, mostly in the context of books/media...not sure he really thought of the Web as the logical extension though
* it also seems that his line of thinking was adopted into the types of user testing that i'm familiar with...let naive users play with the product with no guidance and see what sense they make of it. what mistakes do they make? when were they surprised by how the product behaved? why?

Only negative bits were that there was some material in the middle about theory of mind, memory, psychology of errors etc. that I didn't think was all that relevant. Interesting, perhaps, but a little out of place.
5 likes · flag

Sign into Goodreads to see if any of your friends have read The Design of Everyday Things.
Sign In »

Reading Progress

February 6, 2012 – Shelved
April 30, 2012 – Started Reading
May 3, 2012 – Finished Reading
March 11, 2019 – Shelved as: principal-engineer-reading-list

Comments Showing 1-9 of 9 (9 new)

dateDown arrow    newest »

Brian man really wish we allowed list tags or had some better markup....

message 2: by rivka (new)

If you need instructions for new user to operate, you're probably doing it wrong
This design philosophy probably works for hoi polloi. But for those of us who fall into the group an IT friend of mine calls "superusers", it's crap. I want to be able to customize all my settings, whether it's a word processor or a browser or a bookshelf! And once the defaults aren't enough, expecting that to be possible without instructions is unlikely to still be realistic.

Besides, some of us like manuals and tutorials -- at least if they're well-done.

Otis Chandler Great review Brian! Makes me want to re-read this - I remember loving it.

Rivka: I agree and disagree. I think for new users, Brians statement is accurate. You don't get instructions from Apple when you buy an iphone - it is just intuitive. BUT - for super-user features, having instructions is not a bad thing. My opinion is they should be built into the product though at the appropriate places, and not in a manual - nobody reads manuals.

message 4: by rivka (last edited May 04, 2012 11:34AM) (new)

rivka ← nobody


Brian ah, i left out that he says once users become expert, they should be able to do things faster (example...use menus but allow hotkeys when they really know what they're doing). he believes superusers should also be able to become more efficient than new users, customize as they would...but hide most of that from new users cause they'll just get confused.

i still think he's right though...superusers shouldn't need a manual to customize experience either. i don't like unintuitive menus (see: TiVo) nor cryptic buttons or buttons that serve 10 different purposes depending on context

like my monitor at work has these up/down arrows to adjust things (first off, you can't read what the buttons are unless you look really really's black on black where the black is glossy)...but when you want to change the brightness, you push the down arrow to increase brightness and vice versa....because for some reason they decided the down arrow should be on the right, and then decided the slider on screen when you're adjusting the brightness should be left-right (brighter to the right). took me a while to figure that out...and then realize they tried to indicate it to you by saying:

brightness: ^/v: -/+

a) the 'instructions' shouldn't be necessary
b) they should actually be clear. i didn't realize the order mattered in that till after i figured out how to turn up brightness!

Brian I read manuals too. just read one for our refrigerator in fact. but still, that's usually an exercise in frustration because "no one reads the manual" so they don't bother making it readable.

message 7: by rivka (new)

rivka Brian wrote: "he believes superusers should also be able to become more efficient than new users, customize as they would...but hide most of that from new users cause they'll just get confused."


Brian wrote: "i don't like unintuitive menus (see: TiVo) nor cryptic buttons or buttons that serve 10 different purposes depending on context"

Also agreed.

But good manuals/tutorials have their place. I like the tutorials MS makes for many of the medium-advanced features of Office, for instance.

message 8: by Michael (last edited May 10, 2012 11:42AM) (new) - rated it 3 stars

Michael Economy I HATE that iphones don't have a decent manual, I've owned one for four and a half years and i 'discover' new features from the TV commercials! thats the most round about way to educate your users.

Theres a lot of cool features that are totally non intuative, and the only way you learn about them is word of mouth! Case in point, using volume up to take photos.

message 9: by Otis (last edited May 10, 2012 02:51PM) (new) - rated it 5 stars

Otis Chandler Cool - I didn't know you could use volume up!

I think its a 80/20 or 90/10 kind of thing. MOST things should be easy to figure out and self-explanatory. But not everything can be.

back to top