Good design considers the experience of the user

Coming up with ideas for web design on paper

Web designers often talk about user experience as a key component of their work. In this context, user experience describes what someone actively using the site thinks about the experience of using the site, and how good or bad that may be. But what about the other side of user experience? 

What knowledge, history, or baggage accompanies the visitor when they arrive at your website? It’s incredibly unlikely that yours is the very first website someone will ever visit. As such, the decisions that we make as designers often take into account conventions that we expect our users to largely already understand. These could include button styles, interface elements like forms and drop down menus and well known iconography. Established layout traditions such as horizontal navigation bars at the top of the page or logos located in the top left that link back to the homepage help new visitors to your site become quickly acclimatized, and allow them to get down to work and find the specific content they’re looking for without getting too hung up on the interface.

What about when users expect things to work a certain way based upon their previous experiences or common conventions, and the user experience that you provide is decidedly different?

Let’s look at a design trend from another arena that can point to the dangers of implementing an interface that is not fully understood by the user base.

Over the past number of years, I’ve noticed a design trend that has the potential to not only confuse its users—but could actually kill them. This new design convention comes from the automotive industry. I first noticed it on a Mazda 3 I owned nearly 10 years ago: electroluminescent dashboard lighting. Basically, the dash lights come on whenever the car is running, day or night.

It used to be that the gauge cluster in a vehicle only illuminated in conjunction with the headlights. So, when you were driving after dark, even if your daytime running lights were on (we’re in Canada, where DRLs are mandated by law), you knew that your head and tail lights were not on if you couldn’t see your speedometer. With advances in LED technology, and the omnipresence of tablet-style interfaces in vehicles, nearly every modern vehicle has a dashboard that illuminates during the day as well, creating a compelling and more easily seen instrument cluster at all hours. But, it removed one of the major visual cues drivers used to tell that their lights were on.

The trouble is that the experience (there’s that word again) of drivers is that when the dash lights are on, the headlights are also on. Their previous exposure to vehicles has shown them that this is the convention. The trouble is, many drivers don’t realize that they still need to physically turn the headlight switch when the other visual clues in the vehicle they’ve relied upon for years tell them that they’ve got lights on.

Every evening on my commute home, I see at least five vehicles with no tail lights on and only low-brightness daytime running lights in the front. I’ve witnessed dozens of near-accidents as vehicles were nearly rear ended since they were pretty much invisible from behind. Flashing lights at these drivers doesn’t help—they just think you’re an asshole for doing so. Which is true, I guess, but it doesn’t change the fact that they’re driving a car that cannot be seen easily.

Auto manufacturers need to address this issue by making the head and tail lights of any vehicle with an illuminated dash come on automatically whenever it’s dark. It should not be a user-selectable setting, since the risk to other motorists is huge. If governments are going to mandate driver aids such as backup cameras and electronic stability control, automatic headlights would seem to be a no-brainer.

Of course, it seems pretty unlikely that your lead generation focused website could kill someone if your visitors don’t understand what a hamburger menu icon is for. It does mean, however, that when you work with a design firm on your next website redesign, you’ll want to consider the interface design conventions your users are already familiar with.

These could be conventions or features established on your previous website that your users are used to and would miss. They could be interface standards utilized on large sites across the web such as Amazon, Facebook, or CNN, for example – sites used by a large swath of the population. No matter where the conventions come from, designers need to be considering what their personas may or may not know and use their judgement to design accordingly.

If site visitors are used to a hierarchical menu for a directory of services and you replace this system with a single search box, do not be surprised if the number of people who are able find these important pages decreases. A better solution might be to introduce the search box as an additional navigation option, and see how the usage compares with users who browse the directory to access the same content. If usage of the search box skyrockets, then perhaps you can begin to progressively eliminate or simplify the directory.

Changes to single site features (such as the example noted above) also apply to full site redesigns. Before entering into a major site overhaul, consider taking a more measured approach using Conversion Rate Optimization and A/B testing to see the impact of design changes on your site visitors. Use data from heatmaps and funnel analytics to make progressive design decisions rather than wholesale changes to the interface your site visitors are used to. This data will help you to create a site that continues to convert at a higher level over time.

Rather than simply blowing up the old and replacing it with an entirely new site design, consider what your site visitors are used to, make a hypothesis about how you might improve that experience, and test it. Implement those changes, create another hypothesis, and test again. If users still aren’t finding the lightswitch, consider that maybe you’re sending them the wrong signals, and reevaluate your design.

Join us in conversation…