Using Real People to Test a Web Site

Knowing how customers interact with a ecommerce website can be greatly beneficial. There’s a business that offers real-time user evaluations, where individuals navigate through a website and record their experiences. The results can identify possible issues, which merchants can then adjust to enhance their conversion rates.

The business is UserTesting.com. We talked with its own vice president of marketing, Chris Hicken, about the organization and the user-testing procedure.

Practical eCommerce: What’s UserTesting.com?

Chris Hicken:“It is much easier to understand what we do in the context of this problem that we’re trying to solve. Most websites have a three percent conversion rate. Meaning that for every 100 visitors that come to your website, 3 of them will go to place an order, or sign up for a newsletter, or anything you decide to be a conversion.

“The question for site operators to answer is,’Why are the other 97 percent of the people leaving the site without buying?’ That is where our support comes in. We make it simple for website owners to seek out people which are a target market and watch them perform common tasks in their sites. Some tasks might include trying to find a solution or placing an order or registering for a newsletter. Through that process, you’re likely to discover how and why people get stuck on your own site and why they wind up leaving your site without buying. That is what we do.”

Chris Hicken

PEC: Is it mainly conversion testing that you’re considering, or can it be other kinds of testing?

Hicken:“I’d say it is heavily focused on endurance. But from the context of your viewers, most people who are online retailers will use us to determine why people are leaving the site without purchasing. So it is really conversion testing support.”

PEC: What is that testing procedure going to cost our subscribers, ecommerce merchants?

Hicken:“It is $39 per person that tests your website. Typically a standard usability evaluation will be approximately 15 minutes long. All that you will need to get started is a URL that you need to check and a list of jobs that you need someone to carry out. By way of instance, lets say a company like Amazon.com would like to watch five individuals place an order for a TV on its site. They will instruct our users to begin on Amazon.com and they will probably say,’Use whatever method you would normally use to look for a TV that you would purchase for yourself. Then go through the process of purchasing it.’ That is really it.

“Then, we have a panel of individuals located all across the U.S., Canada, and the U.K., and on a first-come first-served basis people will accept that work from Amazon.com and they’ll record their computer screens as they perform those common tasks at Amazon.com.”

PEC: You have just mentioned Amazon. Can you tell us other ecommerce customers using your services?

Hicken:“A lot of the large online retailers utilize us, such as Walmart, Amazon, Staples, Dell, HP, Groupon. But because our support is relatively cheap, we also have startup businesses using us to get some early comments about how good or bad their site is. And medium-sized retailers often use us to boost their site usability and the conversion prices.”

PEC: There are other testing procedures, such as split testing, multivariate testing, that type of thing. Google provides these kinds of tests at no cost on its Website Optimizer. Why should merchants cover your testing when they could do those other kinds of testing at no cost?

Hicken:“I would strongly recommend all merchants use Google Analytics and Google Website Optimizer — those tools which are free. They work really well with our services. It’s possible to think about user testing for a problem discovery tool. So as you’re watching people go through your website, you’re likely to find where people are getting stuck and why. That gives you an idea of these things on your website that will need to change first. So you will determine what the low hanging fruit is original.

“You can detect, for instance, the navigation on your site is truly confusing for users by using our services. Then you will go on to construct a split test, an A version and a B version of your website — maybe multiple variations. You will use Google’s Website Optimizer to check the functioning of the new design which you produce.”

PEC: Can you give us some examples of ecommerce businesses which have used UserTesting.com and made adjustments to their websites based on what they heard?

Hicken: “Yes. A significant online merchant used our service to do some testing of a new shopping cart that it had published and it found two key problems with the cart. The first one was that all Internet Explorer users received a security warning from the shopping cart. The error said something like,’This page isn’t secure. Do you want to continue?’ And that’s a significant no-no in shopping cart development. So that was the primary problem they found.

“The next one was as users went through the shopping cart processthey could either check out as a guest or they could check out with their current accounts. Our testers found when they clicked the’check out as a guest’ button, they were taken to a webpage where they had been asked to fill in information as though they were registering for a normal user. That was really irritating people as they went through the checkout process, since they clicked’check out as a guest’ they expected to not have to register for an account. Hence the merchant went on to correct both of those issues. It experienced significant improvements in customers getting through their shopping cart procedure.”

PEC: For instance, how many testers did the merchant use to determine those findings?

Hicken:“It took five users to detect those issues. The merchant did run another group of ten users to confirm the findings. But the first findings were done with only five users.”

Listen

Listen to the complete audio interview with Chris Hicken.

PEC: Walk through the process of doing all this with your business. How can a merchant register? How long does it take? How are the results delivered? That type of thing.

Hicken:“You Begin by visiting UserTesting.com. There’ll be a button which says,’Try It.’ The four pieces of information which you will need to have are: (a) the URL you want users to begin, (b) the situation you want users to assume, such as’Imagine you want to buy a TV,’ (c) a list of jobs you want the users to execute, and (d) the amount of testers that you want to use.

“For an early evaluation, if it is your first time doing so. I would recommend just starting with three testers to be certain that you get some interesting results. Of course if you want it, you can keep doing more tests.

“In terms of what you can expect to return, you will find a video recording of each tester’s monitor and sound. You’ll have the ability to hear them speaking their thoughts out loud as they use your website. You get about a 15-minute video recording and you get written answers — up to four questions which are answered after the exam is finished. So in case you wish to ask some questions about how hard or easy it was to use the website, you can do that.”

PEC: How long can that take?

Hicken:“Generally you’ll receive results back within 1 hour of placing the order on the website. You may place an order almost any time in the day and get results back in one hour.”

PEC: How are these results delivered to the merchant?

Hicken:“You will find an email notification that the results are ready for you to review. To watch them, you may visit UserTesting.com and log into your account.”

PEC: Talking of UserTesting.com, tell us about the company. Where’s it located? When was it founded?

Hicken: “The company was founded in 2007 by Dave Garr and Darrell Benatar. We’re located in Silicon Valley — Mountain View, California.”

PEC: Anything else?

Hicken:“Our instrument is utilized in three common ways. First, individuals test their competitors’ sites to find out what their competitors do better. You can watch someone perform three tests on your website and then another 3 tests on your opponents’ and see what their sites do better than yours. That is the first thing.

“The second issue is site operators will watch testers begin their searches on a search engine — such as Google or Bing. For a good deal of online retailers, they understand when they get customers coming to their websites, but they do not know why someone has clicked in their pay-per-click advertising or someone has clicked on their link in the search engine results. That means you can have users start their testing procedure on Google. You can see which connection captures their attention and why and which ones they clicked on.

“The next thing is we also have a mobile device testing support. If you want to check your mobile website and find out how easy or hard it’s to use in an iPhone or Android device, you can use our service to do that.”