Although Steve Krug’s Don’t Make Me Think was first published in 2000, his examination of web usability is still chillingly relevant. The web might have changed since 2000 – it’s now faster, swooshier and more densely populated by cats – but the fundamentals of how we use it are the same.
Kill the question marks
Users hate doubt, so use clear, descriptive text for the navigation – in most cases, basic instructions aren’t the place to play guessing games. Always think simpler – for instance, Krug berates the bookstore websites that use ‘Quick search’ and make you choose to search by title, author, or keyword. He then looks at Amazon, which only has one search box, called ‘Search’. And look where that got Amazon.
We don’t read. We scan.
You’re passionate about your product. You’ve got the best laser-cut chromium-plated, hand-finished dog brushes in the world. You could write about them for hours.
But please don’t.
Because most of the time, users scan the copy instead of reading every word. So get to the point quickly, and make sure you write in a way that can be easily scanned. And it’s not just your writing that should be split into obvious chunks – the whole design needs to conform to a visual hierarchy: make clickable things big and obvious, show people where they are, and don’t mess about too much with the basic conventions of web design.
Use billboard design
Steve Krug’s top tip: imagine users driving past your website at 70mph. If they can’t understand and use your site in the short time they have to spend on it, they’ll be gone. So make it clean, fast, and interesting.
Omit needless words
Cut the copy by half. Then cut it by half again, and you have the right number of words. Be ruthless. We like this one – it’s all about getting to that golden quarter of perfectly distilled copy. (And yes, it’s true that we sometimes want lots of detail, but that’s another story.)
Opinions are a waste of time
‘Let’s put a carousel on the homepage’ someone says. You hate carousels. But your marketing person loves them. You fight until someone pulls rank and pushes through whatever they think is the right choice.
The problem with most arguments about web design is this: they’re based on assumptions and personal preferences, rather than usability testing. Steve’s guide to testing is excellent – mainly for pointing out that lots of people will test for things that don’t really make a difference, ignoring the glaring problems that really do need testing. For instance, the colour of the navigation bar makes no difference if your users have no idea what the site is about.
But testing isn’t everything
He’s also very clear: no amount of usability testing gives you absolute proof that one thing works better than another. The results from a test can only ever guide your decision, along with your experience, judgement and common sense.
So did you scan to the bottom for this roundup bit at the end? Did you just scan the headers? Then you’re behaving exactly as you should. It’s been designed that way. Now go get the book. Or read this post properly, for the bits about cats and lasers.