Testing The Success of Your Web Designs

testingsuccess-01-01
Turquoise Flag-Tip

It’s Friday. I’m hungry. One of my friends recommended a new trendy restaurant and, hey, it’s right around the corner. However, knowing that “trendy” can mean “a sole sprig of lettuce and a sprinkling of truffle oil” and I wasn’t joking about this whole hunger thing, I head to the restaurant’s site to check out the menu before making any firm plans. But, when the graphic and video-heavy site finally loads, the menu is impossible to find, as are directions to the entrance, which I’ve been told is hidden down some strange back alley. In fact, I can’t even figure out how to enter the site, mired as I am in this gorgeous yet impossible to navigate landing page, with earthy music blaring through my speakers. Frustrated, I choose another old standby just around the corner (probably ten feet from this new restaurant, but who can tell?), just to spite the first one.

I know I’m not the only person who feels this way. And yet, I appreciate beautiful design just as much as the next person, and dislike those hyper-corporate sites that are all business and no personality. That’s why I think it’s so important to test the success of your web designs — not just to see whether or not users actually find the site compelling, easy to navigate, and even easier to buy from, but also to see how far a designer can push the envelope before that bottom line begins to suffer.

To accurately measure the success of a website, Google Analytics is a must. There’s a lot to it, so I highly recommend this analytics guide for both an initial grounding in the subject and for more advanced, deeper investigations. As a start, here are just a few key things to look out for when testing the success of your web designs.

Conversion Rate

When it comes to determining the effectiveness of a website’s user interface, there are few measures as telling as conversion rate. Whether it’s an actual purchase or simply downloading a pamphlet, conversion rate is an effective way to gauge just how engaged in the site users are, and where in the sale funnel they might be encountering roadblocks.

Of course, basic measures like keeping a headcount of newsletter sign-ups are useful, but to really dig into the nitty gritty, consider using Google Analytic’s Advanced Segmentation tool. As you can see in this case study of a company called WBC, advanced segmentation can really help you dig down into subtle measures that are powerful yet easy to miss. For example, this particular company found that, lurking within a generally low conversion rate were loyal users with high conversion rates. This lead to a redesign that displayed a greater range of products most desired by loyal customers, and established industry authority.

This tool is most powerful when paired with content experiments, which, despite the title, can be applied not just to content but to various elements of design as well. As a designer, you already know that on a landing page the most crucial information and any forms or other means of conversion should appear above the fold. But how much information should appear? Should the content be wordy and informational or highly visual? How clean is too clean, how packed too packed? With content experiments, you’ll randomly send visitors two or more versions of your site while tracking conversion rates, enabling you to test everything from major layout differences to the color of a headline. Whether that creative, totally new layout works or not will entirely come down to the data.

Event Tracking

Let’s say a design is working and a customer has added the desired product into the cart or started to fill out a form field. But then they get frustrated with just how long the form is, or they type something incorrectly, or the cart responds with an error message. Make no mistake: web users are fickle and these kinds of frustrations are likely to turn them away.

Google Analytics’ event tracking can both identify and help mitigate the problem. With the ga.js tracking code, you’ll be able to see and record just how users are interacting with website elements, and you can classify those interactions with web page objects. So, whether your forms are too long or your checkout process is too cumbersome, event tracking can help you identify user experience and sales funnel roadblocks and move them out of the way.

Site Speed

By most accounts, the average user expects a web page to load in no more than 2 seconds. Yep, all of those jaw-dropping photos and helpful videos and interactive features you’ve added to a site in order to up conversions and engage users (and just generally keep things fun and cool) have all of a couple of seconds to load and become totally functionally. Not only that, but they also have to work on a variety of devices; fair or not, users will blame the site for not loading on their ancient iMac, and all of that design genius of yours will be thrown out of the window.

First order of business a site speed test for every site you produce. The Site Speed menu under Content in your Google Analytics dashboard will also provide a look at specific page load times, as well as that of the overall site. If the results are disappointing and you’ve got a high percentage of visitors coming to you from around the world, consider hosting your site on a Content Distributed Network like MetaCDN. As the name implies, CDNs distribute storage of a site across a worldwide network, so that users will always be downloading page elements from the nearest server to them, rather than waiting for it to download from some server halfway across the world. CDNs also automatically account for the demands of different devices, making for an overall much speedier experience (and a higher likelihood that visitors will stick around).

Bounce Rate

Bounce rate analytics are easy to find in your GA dashboard both for landing pages and specific pages on your site. However, just what bounce rate means is a little more confusing. Strictly defined, bounce rate measures the percentage of users who leave the site rather than clicking links that bring them deeper in.

But if you’ve designed for, say, an expertise blog, this could just mean that users are finding exactly what they want and leaving. You’ll know this for sure if they’re staying awhile on the site — something you can see for certain when you take a look at site times in the Engagement tool. This all may mean that the content and layout in themselves might be great, but there may not be, for example, a nice display of related articles in the sidebar, or enough featuring of services to show the reader that there is more on offer. Other causes of high bounce rate might include loading and error issues, boring content and design and poor usability.

Google Experiments can again prove crucial not just in upping conversions but also in getting users to stay there in the first place. Use the bounce rate to identify and pin down the problem, and Experiments to determine just what to do about it.

Take-Away

Creativity, artfulness, and fun are all crucial elements of good web design. But users won’t appreciate any of that if they can’t find what they’re looking for — and fast. Rather than fearing the numbers, web designers should use them as the source of their creativity. In fact, many times the most creative and inspiring solutions are those that come from within real world constraints. In that way, analytics and web design are perfectly paired.

Discussion

We want to hear what you have to say. Do you agree? Do you have a better way to approach the topic? Let the community know by joining the discussion.