I just add this article in my bookmark.

I wouldn’t mind hearing more details about how you did the Crutchfield test — are you able to share them?

There might be some universal truths about left and right, but at the end of the day the main issue is connected to the individual design – and the way the design drives our eyes.

When looking at the page live on the site (http://www.crutchfield.com/S-fahcfnLTVOg/App/Product/Item/Main.aspx?I=472TOUCH16&cc=01&wm=cl), the lack of content to the right of the product is affecting your behavior.

It forces your attention in one of two other directions – either to the left to the price and the call to action or down to… absolutely nothing.

The most prominent problem with this design might well be the lack of a secondary call to action at the bottom of the page to collect the “lost” users. Having just one “call to action” on this page seems something of a gambit.

When all is said and done I do agree with everyone that challenging a convention is a challenge you want to measure the result of, not gamble you business on.

It makes you wonder, are visitors just trained that way now for web surfing? And if so, how strong is the habit? I’ve got a pretty strong habit of good penmanship (Thank You, Sister Mary Agnes! Your wooden ruler inflicted quite a bit of unintended “cursive”, if you’ll pardon the pun). That penmanship would be hard to break now, for sure. But I’ve also got a more conventional habit of putting my fork on the left when sitting to dinner — a habit I could easily by-pass with just a little effort. So is the “F” effect a deep or a shallow scanning habit? Does it come from the actual preferences of the visitors, or does the common elements of design during the web’s first decade simply imbue a certain level of “F” proclivity?

And you’re correct, Bryan and I are in fact in the midst of a new book, precisely what we were working on when the Left-Side Add-To-Cart discussion first occured. I’ll leave details for a later post once we have a better sense from the major publisher as to release date and the like.

but there’s also plenty of other research which proves completely the opposite. What to do?

As the crew at Future Now keep pointing out, conversion rates for most sites are till stuck at 2%.

We need to be careful about accepting anything as definitive when it comes to web design. With conversion rates that low, everything is up for reinvention.

The F pattern merely shows that that’s how most web designers layout their pages, which forces people to look in that pattern.

There’s another great post on this site. (the bunnyfoot sunsilk ad test) that shows that just by changing the direction of a model’s eyes, you can drastically change where people look.

Now, that’s got me thinking. The effectiveness of Calls to Action are obviously also dependent on what’s around them. So how do we go about getting some definitive tests?

As I said in my earlier reply, publishing and advertising in the west already charge a premium for right placed content.

It would be great to find out their research. Anybody know anyone in advertising who can get hold of this research. I know Ogilvy did a whole lot in the 60’s and 70’s but I lent my book of his to somebody and I never saw it again.

PS John, this is obviously a great post and hot topic. Maybe theirs another article or chapter in an upcoming book on this?

What I did assert outright was that testing is infrequently done at companies, even in 2008. And so, in the absence of any evidence to the contrary, unorthodox UI use should best be assumed as non-tested.

Nevertheless, the point of the article — perhaps missed in your rush to link to us — is not what Crutchfield is doing in the specific, but how companies in general can begin to think about testing techniques for unorthodox elements when they find themselves in a testing culture where such experiments can occur.

Then, following part 2 of the test, you’d only be looking at on-going conversion of those who did convert earlier, i.e., you’re repeat customers. Although that, presumably, is where the left-side AddToCart customers will need to build up some traffic for significance, you’re also drawing that traffic from a sub-group of your customers who’ve by definition have already shown a propensity to buy your products (repeat customer typically have a higher average conversion rate than first-time customers — atleast for businesses who have legitimately good products that satisfy customer needs). So you’re starting from a higher conversion rate population anyway, which will help tremendously with the return traffic you’ll need for significance

In other words, if your site’s conversion rate is so low that you can’t get enough data to make a start at this sort of experiment then this isn’t the sort of experiment you should be considering anyway — you should be doing much simpler optimizations that increase overall conversion (the so-called “low hanging fruit”) before tackling what may in fact be a persuasional challenge for your company’s business topology.

Once you get into the “You previously bought that, so you may like this” sort of stuff, plus tracking such data, auto-generated offers, seasonal stuff and all that jazz – moving the button from one side of the screen to the other is not always as simple as it sounds!

Sometimes there’s a neat and elegant way of doing it, and sometimes it would be easiest to run the entire system twice rather than mess around trying to be smart. Most off the shelf CMS shopping systems were designed before such testing became popular so it’s not a no-brainer by any means.

It strikes me as one of those things some bright young spark comes up with, the IT claim they can do it rather than look silly and then it all goes horribly wrong when some customers get billed twice, some lose their orders and by the way, here’s an email saying our logo is half off the screen and it looks crap in Firefox.

Of course that was the bad old day, WE are so damn smart we never have ANY bugs or screwups and everything will work perfectly what with CSS and being net 2.0 an’ all. It’s just that old-school management have heard all that before and know how quickly “There’s a problem…” features in conversations.

Hope not so see you trying to push over buildings with your bare hands but please send pictures if/when you do.

Thanks.

It’s a sad state we’re in, isn’t it? It’s not that the left-side versus right-side AddToCart is not determinable — we could actually design experiments to measure it and a host of other important varaibles….but there’s no collective corporate Will to do so.

As easy as it would be for me to walk over to 34th St and topple the Empire State Building with my bare hands.

Dream on, guys.

I was thinking you may need to test this with people who have never used the web before – and have not been conditioned to add to cart on the right. Now, that would make for an *interesting* and probably frustrating test!!