I wrote a while ago about dominant design, whereby one design becomes the standard way of doing things. In that post, I gave examples of play/stop/pause icons on media players, qwerty keyboards and traffic lights, though there are plenty of other examples out there. One thing I didn’t mention was the distinction between these two:
De facto standards, where a dominant design becomes accepted voluntarily through the prevalence of its use; and de juro standards, where the dominant design is stipulated by some governing body as obligatory. So media player icons would be de facto, and traffic lights would presumably be de juro.
I was reminded of this the other day when a university library asked for feedback on possible search box designs for their new website. The library’s problem was fairly simple. When users want to search the library catalogue for a specific item, they have one thing they definitely want to do: type in keywords, like “queen elizabeth”. As an optional extra, they might want to put filters on their search, like “books only” or “journals only”.
Here are two ways a user interface could let those users select filters like that before searching:
1. A drop down list, like LSE library:
2. Tabs, like the University of Strathclyde’s library:
I’ve only made a cursory survey of university library websites, but it seems like these are two popular options. The interesting part is that Google does neither of these things – and it’s a safe bet that they’ve tested out what works fairly thoroughly. Starting at Google.com, you can refine your search after you’ve entered your keywords and hit enter, but not before.
Whether we like it or not, Google’s user interface for search has become the de facto standard. So when we use a search engine, we expect:
- one simple text box for keywords, preferably in the centre of the screen.
- no drop down list or tabs to confuse us (at least, not until we’ve done the first iteration of our search)
When we approach search differently to this standard – even slightly differently – usability often suffers as a result. Standards are like a language we all speak, and deviations require code switching on the user’s part. We’re quite good at this, but it can take a bit of effort. And each time we have to put any effort into web browsing, we deplete what Steve Krug calls our “reservoir of goodwill”. If I look at DuckDuckGo, they get lots of goodwill, as I know straight away what their site does and how to use it. Why? Because it follows the de facto standards for a search engine, as established by Google.
The web is built on a set of standards of both the de facto and de juro variety. It’s fairly clear to me that following standards increases makes things more usable. That doesn’t mean they’re the be-all and end-all of web design, but if you’re going to go your own way, decreased usability might be one of the costs.
A slightly unsettling observation related to this is that de facto standards tend to be established by internet giants and followed by the minnows trailing in their wake. It’s surely possible that enthusiasm for these standards could mean siding with the powerful at the expense of the little guys. That’s not such a problem as long as we’re talking about search box design, but there are plenty of other issues where this would leave a bad taste in the mouth. Real name policies on social media were one standard which were established by Facebook and later copied by Google+. Does that mean all future social media should follow this convention simply to meet user’s expectations? Sounds a bit short-sighted to me, not to mention a little bit Orwellian-fascist.
Aside from this, a completely standardised, homogenous world is boring and contrary to the way of Nature. Charles Darwin once published a paper called “On the Tendency of Species to form Varieties”, which seems relevant here.