Who Cares How Pretty Web Sites Are?

An internal debate I had recently about how user behavior shows that prettiness may not be what web site visitors really want.

A few weeks back, I wrote about why I think web standards are difficult to learn. I wrote that because I was spending 80% of my time getting my code into XHTML 1.0 and styling it with CSS so that it rendered consistently across 5 or 6 browsers.

What was I doing the other 20% of the time? Creating content, of course. I was putting together what a huge percentage of my site visitors come for. When I thought about it in these terms (time spent), I felt like styling with CSS was a lot of work for comparatively little gain. After all, people will still be able to find the site, read the content, and click on the links, whether or not I’ve styled it.

Why did I care so much? Why was I spending so much time to get my code to validate? Why was I applying margins and padding until my head spun? Why didn’t I simply drop the content into a couple <Hn> and <p> tags and just display it unstyled?

I’m not sure, but my internal debate went something like this:

Josh #1: Just get it styled, you lazy wimp. Be a craftsman. Look at all the pretty sites out there. You know how to do it, take the time to do it right. Patience and time, my friend. Rome wasn’t built in a day.

Josh #2: Nobody really cares about style, Josh. You know this, even though many folks argue otherwise. Don’t worry about that. The most successful sites on the web aren’t even close to pretty. Amazon, Yahoo, Ebay, CNN: their ugliness not cared about by millions of visitors every day. Why don’t you try to be more efficient with your time? Just get it into semantic code (assuming that it alone will help users) and be done with it.

Josh #1: Not so fast, Bizarro. Even Don Norman just published a book about how pretty things are more desirable. It goes for teapots and it goes for web sites. Don’t you find pretty women more desirable? And about those sites, those sites happen to be business sites and are therefore mostly successful because of their business savvy. They’ve got marketing money to tell people how great they are. People put up with their crappy web sites because they want to get stuff done. They muddle through. I’d even venture to guess that if their sites were pretty, then they’d have even more visitors than they do now.

Josh #2: Pretty women are a dime a dozen, my friend, and pretty web sites, too. Smart & pretty, now that’s desirable. Get the content down first, and then you can worry about how it’s all going to look. Give stylesheets a rest. Wouldn’t it be nice to have a stylesheet shorter than 300 lines? Wouldn’t that be desirable? Let’s get real here: user behavior shows that people were happily devouring web pages back when they were a mass of links on a gray background and years before stylesheets showed up. Thus, the same behavior predicts they will be happily devouring web sites after the next great technology comes along. It’s information they’re after, not tools. Tools aren’t great. What you build with them can be.

Josh #1: Ok, so in the last 30 minutes all we’ve done is stare out the window at brick chimneys and haven’t written a lick of code. On this pace we might even get zero done today. So let’s focus here on what is truly important. Without that XHTML we won’t have anything to write CSS for. Sounds simple enough, right? So let’s get the content done, and if we’ve got more time, then we’ll add the icing on the cake. But if we do add the icing, let’s make sure it’s that nice whipped cream cheese icing with the pretty chocolate cocoa dusted on the top.

Pretty Please?

Published: May 23rd, 2004