Your site judged in 1/20th of a second?

— Published January 25th, 2006  |  by Brian Donohue

The study, conducted by researchers in Canada, received coverage on the BBC, among other places. The researchers compared visitors' impressions of websites when viewed for just 50 ms as compared to "longer durations". (No reports said what the longer durations were.) To everyone's surprise, including the researchers, visitors' impressions in literally the blink of an eye matched well with other visitors' impressions based on the more lengthy viewings.

Just how long is 1/20th of a second?

Really, really short. This site shows you just how quick this is. Before you go there, be sure to view the page just once at the quick speed.

What could you make out in the space of time? Not very much, no doubt. Which is of course what makes these results so interesting—and counter-intuitive.

Are the results for real?

Well, they are in a peer-reviewed scientific journal, so that certainly gives some credibility. But the journal doesn’t make its articles available online, which means unless you’re have academic library privileges, you can’t figure out the details of what and how they tested.

And of course the big question is, what do these results mean for people who design websites? The lead researcher, Gitte Lindgaard, told Reuters, “It really is just a physiological response, so web designers have to make sure they’re not offending users visually.” That seems fair enough.

And what offends users? As Lindgaard says, they don’t know: “When we looked at the websites that we tested, there is really nothing there that tells us what leads to dislike or to like.”

A case of over-generalisation

But Lindgaard also told Nature, “Unless the first impression is favourable, visitors will be out of your site before they even know that you might be offering more than your competitors.” This sounds reasonable, and is the most tempting conclusion from the study. In other words, in a split second people decide whether they like your site and whether they’ll use it.

And here’s the catch. This study focused exclusively on people’s visual impression of the site. Sites that rated well visually in a split second tended to be rated well when looked at longer. But what the researchers didn’t do is correlate these opinions with what people thought after they actual used the site. Despite the conclusion made by the lead researcher, they don’t have any idea if people will leave the site based on that blink impression.

Context matters

Jakob Nielsen made a few comments about this research in his alertbox email this week. His best point was that “users don’t see random web pages”, which is what it seems this study provided. In real life, people arrive at a site with a specific goal, which is almost never “Let me judge how much I like this site’s visual design”. This study removed the natural context of how people use websites, and by doing so, the researchers limited the usefulness of their results.

The bottom line

Anyone in science will tell you to never trust the results of a study until it’s replicated. But hey, this is just websites we’re talking about here, not cancer research. So assuming it’s all for real, what we take away is this: your visitors’ first impression of a web site will always be a visual one. And it now seems their impressions are formed in the snap of a finger. So do some simple blink testing of your site, even if it’s only informally with your friends, because it matters.

But remember that the visual impression is just the first opinion. The real beauty of a site, and users’ lasting impressions, are much more likely to rely on how easy it is to use, and whether you provide content people actually want.