Why people are leaving your website?
How can you perform low-cost remote usability testing for your website or mobile app?
How to find your site’s biggest problem?
Just watch NOW this interview with Darrell Benatar CEO and co-founder of Usertesting.com!
About Darrell Benatar:
Darrell Benatar, a serial entrepreneur, founded Surprise.com in 1999 and hired Dave Garr. Darrell was not a believer in usability testing until he saw Dave doing it.
He was shocked at the problems people were having with things that he’d always assumed worked fine. So he began doing user testing for all Surprise.com redesigns. Since then, Surprise.com was selected by Time magazine as one of the “50 Best Websites.” Seeing its effectiveness on his site, he’s decided to make it easy for other website owners to do user testing.
Raw Transcription:
Marco: I lost you, Darrell. I lost your camera. Darrell, where are you? Darrell, where are you? Oh, you’re back. Okay.
Darrell: I was thinking, I didn’t really want to be able to see myself during the interview because it’s distracting. I guess I’ll just have to not look.
Marco: Cover it with the hand.
Darrell: Yeah, okay.
Marco: Hello everyone. Marco Montemagno here, founder of Tech Alchemist. Today I’ve an absolutely great guest, and I’ll tell you why I’m so happy about it. I have here with me, Darrell Benatar, CEO and Co-founder of UserTesting.com. Hello, Darrell? How are you doing?
Darrell: Hi, Marco.
Marco: We had some efforts to create and to set-up everything. I was talking to Darrell before about how difficult, still difficult to record video interviews. I hope you, Darrell, will create or develop some kind of platform in the near future.
Today I want to talk about usability and testing, and how it’s important to test your website, or your website on mobile, or your mobile app before going online. In this way, you can have a lot of tips.
I want to start immediately with this one. I’ve been using User Testing for the Tech Alchemist website, so I tried and tested on my skin. One funny thing I immediately understood is that on ten- people testing, now I will ask you to explain a bit UserTesting.com, it works. But on ten people that I tried, all ten people couldn’t pronounce my surname. Everyone, they couldn’t say “Montemagno,” everyone says “Monte Magno,” “Magno Magno,” like impossible.
Then I had the genius idea of creating on Sound Cloud how to pronounce my surname with my voice. When you go there, you don’t know how to pronounce? Good. Click here. It was good.
I think UserTesting.com is a great, really great service. We have not [inaudible 02:33] so it’s just that I tested it and I say good, this works and everyone should use it in my opinion. Can you explain it in 20 seconds what UserTesting.com does for the users?
Darrell: Yes. The basic service is we created a large panel of users who will record their screen and their voice while they use your website. You can come to UserTesting.com, and say you have a travel website. You can fill out a form saying, “I want to see ten people who are planning a vacation in the next six months, and I want them to use my website and try and book a vacation in San Francisco.”
Within minutes, members of our panel are recording their screen and their voice while they use that website. They’ll be saying things like, “Well, I would type San Francisco in the search box.”
In about an hour, you get back a video where you see the user screen, you hear them speaking their thoughts, and you see their mouse movements and their key strokes. You start to understand where users are getting stuck on your website.
Marco: Right. I’ll show, Darrell, I will insert in the video a sample of this because for me it was very powerful.
Recording: “Okay. I’m at Petco’s home page, and the first task is to find the store closest to me. I’m looking for some sort of store link. Okay. I’ll click on “Store Locator.” All right.
Okay, so I can enter either my city and state or ZIP code. I usually use my ZIP, because it’s faster to type, so I’ll put that in. Click “Find,” let’s see. Wow. There are tons of stores. I can’t really tell which one is closest. That’s not too helpful. Maybe it would’ve been better if the default view was zoomed in closer and showed fewer stores.”
All right. I can hover over these things for details. I guess all these red and blue things are all Petco logos, but maybe something smaller would’ve been better.
Marco: When I came on UserTesting.com the first time, and I saw a video where the video basically showed the reaction and the kind of video that you were expecting from the users. I said, “Gosh.” In the beginning it was a little bit scary and afraid, because it was a little bit rough, like straight.
Not brutal, but I got reactions, excellent feedbacks by the way. Excellent feedbacks. I don’t know, how can you get such high quality feedback? In the beginning I was-you listen to a person while he’s watching your website and giving you feedback, like, “Oh gosh, this Italian, he’s bald.”
It’s really excellent. How can you have this quality of feedback? How do you select the users?
Darrell: First of all, these are just average people. To become a member of our panel-and we have around 10,000 on our panel. To be on our panel, the user actually does a sample video. We check to see if they’re able to speak their thoughts continuously and follow instructions.
Then, once they’re in our panel, when our clients run test requests, they can give a one to five-star rating on how helpful that tester was. Then we have an algorithm, and when we distribute the jobs, we give a little bit of priority to those users, those testers that tend to be more verbal and have helpful things that they say.
They’re not at all experts in any way. They’re really just average people who are good at talking while they’re doing something.
Marco: Excellent. Another question, how can you be so fast? Because what really impressed me from UserTesting.com is that I was just asking for a test, and then in I don’t know, 30 seconds, I could see people getting the job. How can you be so fast?
Darrell: We have thousands and thousands of people that are signed up to do this. There are a lot of people that are just at home during the day, playing virtual games or on social networks. They get pinged from us that this is available, and they grab it.
Even though we distribute your request for a person to small-ish groups, within that group it’s first come first serve. People tend to be pretty aggressive in grabbing tests because if they don’t get it they won’t.
The economics of this are that you pay $39 per participant, so in your case you had ten participants, so that’s $390. $10 of each $39 goes to the actual participant. They’re just doing this for about 20 minutes, so for a lot of people that’s a very attractive financial offer. I think more importantly, people find it . . . they really get the psychic benefit of saying what they think and knowing that their opinion matters to somebody.
One of the features we have that you might not have experienced is after the test, you have the ability to ask the user follow-up questions through our website, like a chat session. Sometimes our clients use that as a place to compliment or thank the tester for their feedback. That thank you, psychic reward is really powerful and valuable to our panelists.
Marco: Absolutely. To recap, for people watching this video. Usertesting.com, $39 and you get a test from one person. I want to go with you deeper on the kind of demographic that you can have on the tester. Anyway, you can have one test from one person, and you will get a video.
While the person is testing your website, you create all the questions that you him want to answer. I don’t know, “Have a look at the website for five seconds and then tell me what’s your feedback.” Then you also get the answers to all the questions, so you can read also the additional notes, and then you can also follow back.
About the demographics, I’ve been testing male, female. I saw panels from U.S., U.K., and Canada if I remember well. It’s for English websites, let’s say, only?
Darrell: Right now, yes.
Marco: What kind of demographic can you offer to users? What’s your focus?
Darrell: There’s two things. There’s the demographics, which you mentioned some of. Age, gender, income, experience on the web, what country they’re from, if they participate in particular social networks, if they play certain types of games. Then, you can also apply a screener to the selection process. That is a questionnaire that the user answers some questions, and if they answer the way you intend, they are able to do the test, otherwise they don’t.
For example, let’s say you just wanted people who were users of Orange or AT&T telephone networks. You could ask, “What telephone network do you use,” and you could list all the competitors. Only the ones that selected Orange and AT&T would get through. That’s called a screener. You can apply the screener.
Now, we have our basic retail offering, which is what you saw on our website, and we have an enterprise offering, which is a little bit more expensive [for] your commitment, and you get more features with that. You get to use a more extensive screener. You get to get access to a larger pool of testers. That’s one of the areas that is an advantage to signing up with our enterprise program.
Marco: That’s great. I want to give you a few examples more to make people understand exactly what this kind of service is giving. Obviously, this is the topic of usability in my opinion. I think for how many years, I’ve been doing stuff online, and so less testing. Maybe you think, “Okay, we’ll see, we just go out, get the feedback from the users, and that’s it.” In this way you can avoid so many mistakes before.
I was thinking about my experience, and I’ve been testing. Is it better, a photo or a video on the homepage? What is better? I don’t know. The bio, is it a good bio; is it a bad bio? Do you understand what the website is about? Yes or no?
So many things you can test, and in this way it helps to improve the website, the product or the service.
Darrell: Yeah, and I have to comment on that, because people are much more familiar with doing AB testing using their analytics packages. What we find is usability testing is very complementary to AB testing, and in this way, when you want to change something on your website to see if it increases your conversion rate, you run an AB test, one of the questions you have is what should I test?
Oftentimes, I think people are guessing at things that they think might help. In general, 80% of the time when you run an AB test, A wins. The new idea fails. When you use our service or you do usability testing on your website, you find areas that appear to be really problematic. You can run tests to see if changes to those areas increase your conversion rate.
For example, on your video versus photo, it might be that if you ran an AB test, the photo wins. It could be just that people don’t understand the video, or they find the first 20 seconds of the video is not that helpful, so they don’t watch the whole thing.
If you ran usability tests on the video, you might find “Oh, people don’t understand this part of what I’m trying to communicate here.” Fixing the video might be something that increases your conversion rate. You can test that using analytics.
Marco: Just to recap about AB testing for people who are not so inside AB testing. It’s a way of testing-I don’t know a title, I try to write a title with a different word or a different color, so I test two versions. I come up in my mind services like Optimizely or these kinds of tools that they let you run this kind of test.
Darrell: The core thing there is half the people see one version and half the people see the other version, and you compare what the conversion rates are for those two.
Marco: Perfect. How about mobile, Darrell? I’ve been testing it on a website. I didn’t try the mobile version. Do you allow to test your website on mobile and to test a mobile app, or only website? Which is the experience of your website on a mobile device?
Darrell: Yeah. You can test anything on a mobile device. I’d like to quickly explain how we do that, so your listeners understand. What we’ve done is we send out to hundreds of our testers, I think we’re getting close to around 1,000 people, who have our webcam on a stand that we’ve sent them.
They connect that to their computer, and for the test they hold their mobile device underneath the camera. They use their fingers like they normally would. The video feed for that camera is going onto the computer screen for the user and getting recorded just like our normal service works. That video gets uploaded and everything’s the same.
Marco: Excuse me, Darrell. Excuse me to interrupt you. You send to 1,000 people more or less a webcam on a stand that they use, and the webcam is just connected with their laptop, and they can record everything that is under their webcam?
Darrell: That’s right.
Marco: Okay. Cool.
Darrell: What you get back is a video where you see the person’s hand holding the device, and you see their fingers doing gestures on the device. That means it doesn’t matter what platform, or whether you’re testing an app or a mobile website. It can be done on an iPad or a tablet or any other tablet. It covers all the bases.
We find that it’s really valuable to have to be able to see the actual person’s hand and the physical phone. Not only do you get the gestures but you have the gyro. If they’re playing a game, you can see them tipping the phone. It has GPS, so if the application is GPS, you can see that.
You get feedback, like, if the person’s playing a game for example, but it’s designed such that their fingers are too large relative to the items on the phone, and they can’t get it to work. It’s important I think to be able to see their hands and see real usage.
Marco: Amazing. Can you ask those users also, I don’t know if you want to test only on Android, test only on iPad, you can do it?
Darrell: Yeah. When you go through the process of setting up the test, you can specify which device you want it to be tested on.
Marco: Excellent. How do they record their audio? Through the webcam?
Darrell: Yes, the webcam has a built-in microphone.
Marco: All right, excellent. Yeah, in this way you solve the problem to record the screen. From inside, you really don’t get I think the experience of the user. I think this is super-cool.
All right. I think UserTesting.com also got $3 million funding in September?
Darrell: Yes.
Marco: Okay. You’re growing. What’s your next plan for the company?
Darrell: It’s mainly just growing all the offerings that we currently have. We are having good success selling our offering into larger companies. When we started the business, we were thinking usability testing is something that’s very expensive, and only the biggest companies can afford to do it, so we’re going to create something really simple and fast, inexpensive, so everybody can do this kind of testing.
Then later on we started to realize a lot of large companies were using our service, and we started to provide them with additional features and services that they wanted. We’ve had good success selling to those companies. We’re continuing to add more and more services and things that those companies are looking for.
Marco: Okay. Another thing that I saw on your blog, on UserTesting.com blog, was an integration with the prototyping tool. I think it was pretty interesting, prototyping tools are tools like Balsamic [SP] or Just In Mind I think is the one you’re integrating. Tools that help people, companies to create a prototype as the names say.
In this way, you don’t have to code from zero your prototype. You just use this tool and then if the prototype is good, then you will code it. You create the proper app. And you create the integration, so I thought it was pretty interesting about that.
Darrell: Yeah. I think it’s important to point out that this is a tool that is not meant to only be used on a production website. The famous words of advice are “test early, test often.” We want to try and make it easier for people to start to test things earlier in the process.
The reason for that advice of course is the further along you get in the development of the new feature on your website, the harder it is to make changes both technically and psychologically. You become invested in what you’ve created, and it’s scary to find out that all that time you invested was headed down the wrong path.
You really want to try and test early. One of the advantages of testing with our panel is our panel is comfortable with instructions that say, ‘Hey, this is not a fully functional website here. This is just a partially functional; I’m trying to understand some things.’
An example would be, here’s a website. Only the top level navigation works, and nothing else on the page works. That’s intended. What I want to know, the [inaudible 21:05] wants to know is, if you were trying to find this particular piece of information on my website, where would you click on in the navigation to get there? You can start by testing that out and getting that right before you build out the rest of your website.
With this latest integration, you can upload still images, you can upload a PDF file or a JPEG, and you can have people look at three different versions of a mock-up of a home page you have and give you some feedback on that.
Or you can work with a partner like Just In Mind, and you can build a functional prototype which allows you to click around, even though things don’t necessarily work, and have our testers try your prototype out and get some early feedback on that.
Marco: When UserTesting.com is launching new stuff, are you testing on UserTesting.com or not?
Darrell: We do. We use our own service. It’s a little bit harder because our existing panel of users are already familiar with our website. You don’t get as fresh eyes. But we are able to test lots of things on our site.
I want to say that usability testing is one great tool for getting feedback, but there are other important ways to solicit feedback from your customers. Giving people an easy way to click on something on your site and give you feedback, like pop-up surveys.
We use an exit survey. When you leave the site, we quickly show you a pop-up that says, “Hey, if you didn’t find what you were looking for here, could you tell us why? What happened?”
Marco: About this, Darrell. So many years that you are inside this world of usability and testing. Can you say, “Okay, I have three things that absolutely, someone running a business online has absolutely to do before launching his website?” If you should suggest at least three things, three kinds of tests, or your personal framework to test your stuff online before you launch it, what would it be?
Darrell: Well, I’m going to give you those three things, and I’m going to briefly say that when you run a test like this, there’s a spectrum of feedback that you get. On one end is can people just, are they able to accomplish fundamental tasks on your site? Can they get through there?
The other is persuasion. Are you successfully motivating the user to try and accomplish this task on your website? That gets me to the three suggestions that I would have. First, it’s really valuable to run what we call a five-second test on your website, to have users look at your site for five to ten seconds, and then look away and try and explain, what do they remember? What do they understand the site does?
Because there’s so many situations where that’s really how long you have to motivate a user to move forward. The common situation is that clicking on a pay-per-click ad on Google. They’ve read the ad, and they expect something. The question in their mind is, when they get to that website, they’re quickly trying to assess, is this relevant? Does this match what I was expecting? If it doesn’t, they leave, and you’ve wasted your click.
We make it really easy on UserTesting.com when you run a test with our service; we have templates set up, so you can select pre- created questions and a test for users. One of them is that five- seconds test. We’ll actually show them your website and then not show them your website.
We’ll show the user the website for five, seven seconds, and then we won’t show it, and then we’ll give them questions. That’s all set up [inaudible 25:14].
Marco: First tip is five-second test. For me it was very, very useful. I’ve been changing the website because I understood that it was not so clear, the message, then when I did the change it got a good impact. All right.
Darrell: I would say the second thing I recommend is can a user do the most common, important thing on your website easily and successfully? For a lot of retailers, that is can they find the product they’re looking for, and can they successfully make it through the check-out process?
It’s really hard to create an easy check-out process. Users can get easily frustrated and leave your website. You want to try and simplify your forms, so you’re not asking them to do anything unnecessarily. The fewer fields that there are in a check-out form, the more likely you are to have success.
The check-out form is the most likely place that you’re going to lose the user. Just testing the basic function, can they do the basic things they’re supposed to do on your website? I think a mistake that probably is commonly made is, of course they can check-out on my website. People are doing it all the time. I would’ve known if it were a problem.
In fact, we all have gotten used to accepting 3% conversion rates to get the visitors to our websites. 97% of the time, something’s going wrong. There’s plenty to learn there. The best websites have 20%, 25% conversion rates. It goes to show you it is possible. There’s tons of room for improvement.
Then I guess the third thing I would say applies to a lot of websites is testing the search on your website. Search is the lifeline for the user. If they get to your website and they’re kind of confused, their first instinct is to go to that search box. You want to be delivering very relevant search results. If the user doesn’t see what they’re looking for on the first search effort that they make, half the users will leave the site right at that point.
Marco: That’s funny.
Darrell: Yeah. If they try again, it keeps going. Half will leave right away, so you’ll lose three-quarters on the first two attempts. Search can often be improved by better handling of unusual characters. Predictive search has become really important these days.
If you don’t have the budget to build a really great search system, Google actually offers a free search solution where you can put some JavaScript on your page, and it shows a search box, and then you get a page of search results with Google.
That’s sometimes a good way to start because you learn from that what people are searching on and what’s really important. If you build your own search, you start with having some data there.
Marco: Right. Predictive search, I just remembered to viewers, like the one you find in Google now, Google Instant, or if you go on Quora, you find the kind of search, the kind of queries that people are looking for. It’s very interesting.
Darrell: Yeah. As you’re typing, the search suggests possible things that you might be looking for.
Marco: Yeah. Yeah. Are there any specific tools, online tools that you would recommend to run several tests? Not only UserTesting.com, but are there any specific tools that you use or you would recommend? I’m also thinking of, what do they call the hit map, to see where the user is clicking. Are there any specific tools that you would suggest?
Darrell: Yeah, there are a number of lightweight tools that are valuable. I could recommend a site called Usability Hub.
Marco: Usability Hub?
Darrell: Yeah, and they’ve got a number of tools to help you with things like having a large number of users go to your site and asking them to do something, and then seeing where their first click is. You can get a map showing where all the first clicks were.
There’s a tool called Attention Wizard, and it’s kind of interesting, because what they’ve done is they have studied actually heat maps. Heat maps are where you have a large number of users use a website, and you keep track of where their mouse moves. Actually, the true heat map is you have a large number of users that are hooked up with eye-tracking devices, so it can tell where their eyes are looking on the page.
That can be a very valuable tool, but it’s a very expensive, time- consuming thing to do. It can give you some insights as to where users are focusing on the page. Attention Wizard has done a detailed analysis of many, many of these eye-tracking heat maps, and it has built an algorithm to predict based on the colors and the photos and the position of things on the page, predict what an eye-tracking heat map will probably look like for that page.
Marco: Wow. Amazing.
Darrell: Yeah, so you can just upload your homepage on attentionwizard.com, and you’ll get back a heat map that shows you colors on your home page, indicating where red is where people are most frequently looking, and orange, also very popular, but not as much, and yellow. You get that kind of information.
Marco: Amazing, amazing. Very interesting. Last question, Darrell, and then I’ll let you go. I know you’re super-busy. What will happen next in usability? I heard a recent interview from Mark Zuckerberg. He was only talking about mobile. Google for ages now, they only say mobile first, mobile, mobile, mobile. Everything looks like it’s changing. iPad and so on. What will change in the usability world if anything? Or everything will stay the same, or with all these devices we should think about a totally different way of designing our online presence?
Darrell: What we’re seeing actually already is that website owners are recognizing that their users are interacting with their web presence on multiple devices. You may be out shopping and see an item that you’re interested in, and you might look it up on Amazon.com.
Then, when you get back to your office, you might go to research that item a little bit more and then complete the purchase on Amazon.com. That’s one example. We see that people who are responsible for websites or an online presence are going to need to make sure that the user can move between devices and maintain the same session, and that things are consistent, so that they stop thinking that there’s much of a difference between being on their computer or on their mobile device. Moving across screens is what we see the future of testing, you’re going to need to test across all devices.
Marco: Okay. Excellent. Darrell, thank you so much. Darrell Benatar, CEO and Co-founder of UserTesting.com. Have a look guys, because it’s a great website. Also we’ll check the couple of websites that you suggested, very interesting. Thanks much, good luck with everything, Darrell.
Darrell: Yeah. Thank you so much, Marco. It was a pleasure.
Marco: Thank you.
Darrell: Bye-bye.