With all of the factors involved in designing and building a web site, one of the most influential questions is often overlooked:

Is your site easy to use?

Usability isn't the same as design. Just because you've hired a talented designer to craft your site and make it look great doesn't mean it's easy to use. Looking good is a completely different matter from working well! After all, plenty of beautiful sites have won design awards while losing customers by the thousands.

How many times have you gotten lost on a good-looking site or abandoned a purchase in frustration after you couldn't find the information you were looking for? If you walk into a brick-and-mortar store and can't find your favorite brand of gherkin pickles, you can simply ask an employee where they are. But on the Web, it's much easier for a customer to go to a competitor's site than to go through the trouble of sending an e-mail inquiry.

Whatever your business is about, your web site will have specific goals, such as convincing people to...

• Subscribe to your newsletter
• Fill out a survey
• Purchase your product
• Inquire about a service you offer

Usability is simply a gauge of how easy it is for your visitors to do these things.

For an e-commerce site, usability is especially crucial. If people can't follow your navigation scheme, they won't be able to find your products. And if they can't find them, how can they buy them?

Obviously, a key measure of the success of your site is its efficiency in converting visitors to buyers. Yet did you know that, according to market research from the Gartner Group, more than 50% of Web sales are lost because visitors can't find the content they're looking for? And another study by usability consultants Creative Good estimated that improving the customer experience increases the number of buyers by 40% and increases the overall order size by 10%!

With results like these, why doesn't everyone test their sites for usability? Some people mistakenly assume that usability testing is too expensive, too time-consuming, or too complicated to bother with, especially for smaller companies.

Fortunately, usability doesn't need to be any of these things. While there are high-priced consultants who can do it for you, a do-it-yourself test can be very effective.

Setting Up a Basic Usability Test

While usability testing is most efficiently done as part of the process of creating a web site, it can be done at any time to improve your site's effectiveness. If you are planning a design update or adding new elements to a site, it's crucial you begin the testing before you invest time and money in making changes.

To do a basic usability test, you just need to find a "sample group" of potential customers and ask them to perform simple tasks at your web site -- like purchasing a product, subscribing to a newsletter, or locating specific information like your guarantee -- while you watch them.

1. When to test:

You can test usability any time. In fact, even if you don't have a site yet, you can still test your initial design using rough sketches on paper that show the layout of key information and navigation links. If you're testing potential changes to an existing site, you can work from quick HTML mock-ups, or use your designer's print-outs.

Obviously, the more detailed the testing prototype, the better the results, but you'll be surprised by how much information you can gather with even the roughest layouts.

If your site is up and running already, you can test your current design to flag any potential problems and increase its efficiency. Usability testing should be an ongoing process to fine-tune your site and make sure you aren't losing customers -- and profits -- unnecessarily.

2. Set your goals:

Start by setting your testing priorities. Which of the actions your visitors perform are most important to your business? Focus on a few key things you want all visitors to be able to do, such as...

• Subscribe to your newsletter
• Become a member
• Add a product to their shopping cart
• Find answers to common questions

These basic tasks are the "script" for the test. The more complicated the site, the more detailed the script. An e-commerce site selling plumbing supplies might use a script that looks something like this:

1. Click the link for the page on which you think bathroom faucets are located.
2. Find the American Standard "Ceratop" faucet.
3. Are there any less expensive faucets?
4. Add it to a list of items to buy.
5. How much will it cost to ship the faucet to where you live?
6. How long is the warranty?
7. Complete the purchase.

As your testers work through each task, you'll be able to see how they use your site. Do they browse categories or look for a search function? Do they encounter any difficulties along the way? This is an incredible opportunity to get inside your customers' heads and watch what happens when they use your site.

You can also analyze your site's metrics to see what's not working. If an analysis of your web logs reveals that tons of people are exiting your site from one or two particular pages, for instance, usability testing can be a good way to find out what's behind the high exit rate. This is especially crucial if these pages are part of your check-out process.

Note: If you can, get a test credit card number from your merchant account or gateway provider so your testers can complete test purchases. If this isn't possible, have the testers take the check-out process as far as possible, and then ask them what they'd expect to happen next.

3. Choose the right people:

The people you choose for the test are important, as they should mimic the range of users you have (or want to have) using your site. Sit down and gather any customer demographic information you have to create a series of user profiles.

What is their level of computer experience? How old are they? What special knowledge do they have (if your site serves a specific demographic or industry). A site targeting real estate professionals will have very different user profiles than a site selling skateboard wheels, so make sure your testers mirror your actual users.

Strive for a mix of computer experience that matches the mix you'd expect of your audience. Are most of your customers already comfortable with computers? Are there some newbies in the mix? You can recruit existing customers if you're testing changes to the site, but for an existing site, look for people who haven't used your site before.

Finally, don't worry about getting a large pool of testers: You only need five or six people to identify 80% of the main problems that may be affecting your sales.

Note: It's common practice to pay testers for their time and effort. And while using Uncle Henry or Bob from accounting may save you $40, they're likely to skew the results if they don't reflect your target audience and are already familiar with your site.

4. What you'll need:

Set aside a clean, quiet place where there will be no distractions, and provide a comfortable chair for the tester. Place a chair for yourself slightly behind the tester so you can see where they're clicking as they complete each task.

Have your tasks and questions -- your script -- written down, and be ready to take notes. If you have a video camera, you can also tape the test (with the camera looking over the tester's shoulder towards the screen). Before you start the actual test, run through the script yourself to make sure all the links are working, that the tasks make sense, and that the video equipment captures the detail you'll need to see.

5. Running the test:

Before you start the test, explain to your testers that it's the site you're testing, not them. Let them know that they can't do anything "wrong," and tell them to surf the same way they normally would. The more relaxed and natural they are during the test, the better your results.

Then, ask them a few questions about their level of experience, how often they use the Web, and what they know about your company and products, so you can better understand their reactions.

Start at your homepage, and ask them what they think your site is about. This can be a good way of judging how successfully you're welcoming new visitors.

Throughout the test, encourage your testers to think aloud while they work through the tasks you've set out for them, so you can get a sense of their expectations.

Next, work through your prepared script. Ask the tester to attempt various tasks and answer the questions you've prepared, while checking their expectations with questions such as: "What do you think you'll be able to do here?" and "Before you press that button, tell me what you expect to see next." While you should take notes and follow the script, be flexible enough that you can pursue any responses that may take you by surprise.

During the test, be sure not to guide the subject. Watch that you don't provide any hints, suggestions, or even answers that will influence their actions. If they can't complete a task, simply ask them what they expected to have happen and how they'd fix the problem, then move on to the next task.

If testers have a problem or become confused, don't assume you know why. Ask what the problem is, and then paraphrase their answer back to them to make sure you aren't bringing your own bias into the test.

6. What to watch for:

• Hesitation: If their mouse cursor hovers over a link, ask them what they're thinking. Hesitation often means they're trying to figure something out, and usually indicates a problem. In a perfect design, the user doesn't have to think -- everything makes sense and the next step is always clear.
  
  
• Backing up: When users back out of a page (using either their browser's "back" button or the site's navigation) it's often a hint that their expectations weren't met. Perhaps they thought the link would take them somewhere else, or they've lost track of where they are in the site.
  
  
• Unexpected routes: Did your tester take a different route through the site to accomplish a task than you expected? People tend to have different ways of navigating web sites. Did they use their browser's back button three times to retrace their steps rather than clicking once on your navigation links? It may be a sign they've lost their way or haven't noticed the links.
  
  
• Extended reading: Unless your page is a long salesletter or has a newsletter, users shouldn't have to read too many instructions to make their way through the site. Usability isn't just about buttons and navigation; it's also an important test of your copy. Can your visitors find the information they're looking for, and do they understand it?

Making Changes and Testing Results

Once you've thanked your guinea pigs for their time and the tests are finished, go over your notes. You're looking for general patterns and behaviors, not details or specific statistics. Did most users get stuck at the same place? Did more than one person hesitate over the same button?

The biggest sticking points should reveal themselves pretty quickly. Once you've identified the main roadblocks, use your testers' suggestions about how they'd fix them or what they'd expect to find as a basis for a solution, and then test the solution -- before you implement it!

As with any testing, make sure you change only one thing at a time so you always know exactly what's responsible for any improvement.

And throughout the testing process -- from coming up with the script to implementing the changes -- try to keep an open mind and trust your users. Their feedback is not a criticism of you or a reflection of how much time you've spent on your site. In fact, the more time you've spent working on it, the less objective you may be about how it works.

Note: If you rely on third-party solutions like shopping carts or payment systems, you can't always change the way they work to improve usability. If testing reveals serious problems, it may be worth investigating -- and testing -- other solutions, even if they're more expensive. After all, a poorly designed shopping cart system that's causing half of your customers to abandon their purchases is no bargain!

Usability Tips

A big part of usability testing involves looking at your site from the customer's point of view. Sure, your programmer or Web designer may have a bunch of perfectly valid technical reasons for setting up things the way they are, but your goal shouldn't be to make things easier for your programmer or designer at the expense of your customers' experience.

As you surf the Web over the next couple of weeks, keep an eye out for usability issues you come across on other sites -- basically anything that makes you back up, curse, stop to figure out the next step, or stare blankly at your screen! Make a note and bookmark these sites for future reference.

And make sure your site isn't guilty of common usability blunders like these:

• If a potential customer forgets to fill in their zip code when they submit an order form, will they lose all the information they already entered and have to start over again? If so, you'll likely lose a number of potential customers at this point.
  
  
• Your site's navigation scheme must be clear and intuitive. If your users have to guess at the meaning of vague icons or have to squint to read an obscure typeface, you're making them work too hard.
  
  
• Usability also takes into account other issues, such as load time. Research shows that if the time between a viewer's click and the appearance of a new page is more than six seconds, they get distracted and are likely to move on -- probably to your competition!
  

Final Thoughts:

These days, there are certain expectations regarding how a web site should look and how it should work. For instance, research shows that most people expect to see a "home" link in the top left corner of a page, and that they look for internal links down the left as well.

Now, you could argue that internal links look better or make more sense along the right side, but in the end, usability isn't about what "makes sense" or looks good to you, it's about what works for your average visitor. And if 90% of your users expect to find your navigation along the left side of the page, then the left side is what works!

Sometimes the simplest solution is the best. Links that look like buttons get clicked on more often -- simply because they look like something that can be clicked on. The first thing anyone who surfs the Web learns is that blue, underlined text is a link. If you start making your links look different for the sake of prettying up your site, you risk losing functionality.

Finally, don't reinvent the wheel just for the sake of being trendy. Your web site is a business tool first and foremost. Study sites that have a similar function to yours and look for common approaches. Amazon.com, for instance, has helped set standards and expectations for how an e-commerce site should be organized.

While you don't want to simply copy successful sites, it makes sense to adopt some of the same navigation techniques. After all, with millions and millions of customers using a site like Amazon.com, chances are your visitors will be familiar with their approach. Take advantage of this familiarity and apply the usability strategies other successful sites have found to be effective to your site -- then focus on testing to fine-tune the way your own site works.

Still not convinced you should test your site's usability?

Make no mistake: If you don't test your site, your visitors and customers will "test" it every day! The problem is, if they're having trouble using your site, they won't take the time to send you a note offering helpful suggestions -- they'll just check out your competition!