1 00:00:00,000 --> 00:00:04,600 [Usability Testing with Dan Gorgone] 2 00:00:04,600 --> 00:00:06,230 [Dan Gorgone @dangorgone] Hey everyone, I'm Dan Gorgone. 3 00:00:06,230 --> 00:00:08,560 Thanks for joining me here at this workshop. 4 00:00:08,560 --> 00:00:10,610 I'm going to present about usability testing 5 00:00:10,610 --> 00:00:12,800 and do a little Q&A at the end. 6 00:00:12,800 --> 00:00:16,030 I want to talk about the benefits and the purpose 7 00:00:16,030 --> 00:00:18,340 of doing usability testing and the process, 8 00:00:18,340 --> 00:00:20,140 what happens before you test, 9 00:00:20,140 --> 00:00:23,230 what happens during a test, what will happen after a test 10 00:00:23,230 --> 00:00:26,020 and then some ways you can take your testing 11 00:00:26,020 --> 00:00:28,890 to the next level. 12 00:00:28,890 --> 00:00:32,140 Now, a recent course I just published 13 00:00:32,140 --> 00:00:34,350 is Usability Foundations. 14 00:00:34,350 --> 00:00:39,400 It's something I believe in very strongly as a designer 15 00:00:39,400 --> 00:00:43,670 of really anything, whether you're talking about websites or apps 16 00:00:43,670 --> 00:00:46,190 or you're talking about light switches on the wall 17 00:00:46,190 --> 00:00:49,170 or your car or anything. 18 00:00:49,170 --> 00:00:55,750 If you want someone to use something that you're designing, 19 00:00:55,750 --> 00:00:59,620 it's got to be usable, and if you want them to use it a lot, 20 00:00:59,620 --> 00:01:04,340 it has to really work, and they have to be satisfied with the experience. 21 00:01:04,340 --> 00:01:07,080 Part of what I talk about in Usability Foundations, 22 00:01:07,080 --> 00:01:09,360 which is this new course you can check out, 23 00:01:09,360 --> 00:01:15,140 is defining what usability is, how it factors into usability experience, 24 00:01:15,140 --> 00:01:19,440 but then getting into some of the details about how you can improve 25 00:01:19,440 --> 00:01:22,260 the usability of your websites, your apps, 26 00:01:22,260 --> 00:01:25,710 and use different things like testing and frameworks 27 00:01:25,710 --> 00:01:29,830 and other stuff, especially the mobile angle, 28 00:01:29,830 --> 00:01:36,040 because the mobile angle has certainly changed design as well as testing. 29 00:01:36,040 --> 00:01:39,940 Many of you may be wondering 30 00:01:39,940 --> 00:01:43,800 is usability testing really worth it? 31 00:01:43,800 --> 00:01:46,460 And what is the purpose? 32 00:01:46,460 --> 00:01:48,700 What are the benefits behind testing? 33 00:01:48,700 --> 00:01:52,020 The first thing you might think is that we want to make sure 34 00:01:52,020 --> 00:01:55,570 that the site or the app is working. 35 00:01:55,570 --> 00:01:57,120 That's true. 36 00:01:57,120 --> 00:02:00,980 That's a good thing that you want to know. 37 00:02:00,980 --> 00:02:03,740 But a lot of people have this misconception 38 00:02:03,740 --> 00:02:07,350 that testing is something you do right before you roll something out. 39 00:02:07,350 --> 00:02:12,050 Well, it is a good time to test certainly, 40 00:02:12,050 --> 00:02:15,540 but the idea is you want to have done testing 41 00:02:15,540 --> 00:02:17,450 throughout the entire process. 42 00:02:17,450 --> 00:02:19,960 The way I think about it, I use this analogy. 43 00:02:19,960 --> 00:02:23,900 If you're a runner and you're getting ready to run a race 44 00:02:23,900 --> 00:02:26,780 the last thing you want to do 45 00:02:26,780 --> 00:02:31,160 as the starter is counting down the seconds for you to start 46 00:02:31,160 --> 00:02:33,790 is to check and see if your shoelaces are tied or not. 47 00:02:33,790 --> 00:02:35,290 You should have done that. 48 00:02:35,290 --> 00:02:38,350 You should have checked your equipment, checked everything you needed to do 49 00:02:38,350 --> 00:02:40,560 before you run this race, and you need to be ready 50 00:02:40,560 --> 00:02:42,720 when the launch is going to happen. 51 00:02:42,720 --> 00:02:46,140 Right before you launch is not the only time 52 00:02:46,140 --> 00:02:48,070 you should be doing testing, and in fact, 53 00:02:48,070 --> 00:02:50,000 I think you shouldn't be doing testing at all then. 54 00:02:50,000 --> 00:02:52,100 You should be concentrating on the launch. 55 00:02:52,100 --> 00:02:55,950 That's why you should have it spread out throughout the design and development process, 56 00:02:55,950 --> 00:02:57,490 from the very beginning. 57 00:02:57,490 --> 00:03:00,830 By doing testing throughout that process 58 00:03:00,830 --> 00:03:04,720 you will improve the design, and you will improve your development process 59 00:03:04,720 --> 00:03:10,170 throughout the entire thing, make things much more efficient. 60 00:03:10,170 --> 00:03:14,750 By doing testing, you shouldn't have this fear 61 00:03:14,750 --> 00:03:18,010 that you have to make sure that everything works 62 00:03:18,010 --> 00:03:20,640 every time you do testing. 63 00:03:20,640 --> 00:03:22,920 Testing is a great opportunity 64 00:03:22,920 --> 00:03:27,160 to pinpoint specific features on a site or app, 65 00:03:27,160 --> 00:03:31,210 and I'm going to use site and app pretty interchangeably here, 66 00:03:31,210 --> 00:03:35,550 so whatever I say about sites you can apply to apps as well. 67 00:03:35,550 --> 00:03:37,550 You can pinpoint these features 68 00:03:37,550 --> 00:03:40,780 and make sure that the most important features 69 00:03:40,780 --> 00:03:43,630 or parts of your site are working. 70 00:03:43,630 --> 00:03:46,120 Think about it this way, and we'll get into it in more detail. 71 00:03:46,120 --> 00:03:48,070 But think about it this way. 72 00:03:48,070 --> 00:03:52,650 If specific parts of your site were not working, 73 00:03:52,650 --> 00:03:55,810 what are the ones you could get away with not working, 74 00:03:55,810 --> 00:03:58,750 and what are the ones that would absolutely cripple you? 75 00:03:58,750 --> 00:04:02,030 Making sure that the essential features 76 00:04:02,030 --> 00:04:05,590 are working is part of what you should be doing with testing. 77 00:04:08,470 --> 00:04:10,580 Another incredible benefit 78 00:04:10,580 --> 00:04:15,990 of testing is gaining an outside perspective. 79 00:04:15,990 --> 00:04:23,560 Design and development can be a very solitary sort of process. 80 00:04:23,560 --> 00:04:26,760 Even when you're working with a team you're very insulated 81 00:04:26,760 --> 00:04:29,790 from the outside world, even insulated from other departments 82 00:04:29,790 --> 00:04:33,160 if you're in a company, especially if you're by yourself, 83 00:04:33,160 --> 00:04:35,590 you're freelancing, things like that. 84 00:04:35,590 --> 00:04:38,290 You're stuck in a cube where you're really buried in your laptop 85 00:04:38,290 --> 00:04:40,450 and you're building stuff. 86 00:04:40,450 --> 00:04:44,210 It happens that—and it's only natural— 87 00:04:44,210 --> 00:04:47,200 that you use your instincts as a designer, 88 00:04:47,200 --> 00:04:50,500 and if you've been hired to do things, you've done lots of research, 89 00:04:50,500 --> 00:04:53,830 you've had lots of training, you should know how to design things 90 00:04:53,830 --> 00:04:57,680 because you can anticipate what other people are thinking 91 00:04:57,680 --> 00:05:02,740 or will need when they use the thing that you're designing. 92 00:05:02,740 --> 00:05:07,000 But gaining that outside perspective, once you've actually put something together, 93 00:05:07,000 --> 00:05:10,330 is incredibly powerful. 94 00:05:10,330 --> 00:05:15,330 And it's really something that you have to see it to believe it, 95 00:05:15,330 --> 00:05:17,530 and you can trust me on this. 96 00:05:17,530 --> 00:05:21,530 I've had the opportunity to test different sites before, 97 00:05:21,530 --> 00:05:26,910 and when you see someone use something differently for the first time, 98 00:05:26,910 --> 00:05:29,150 it's a real eye opener. 99 00:05:29,150 --> 00:05:31,680 But it's an incredible opportunity 100 00:05:31,680 --> 00:05:34,620 to learn something really valuable. 101 00:05:34,620 --> 00:05:36,550 It can be a positive interaction 102 00:05:36,550 --> 00:05:38,000 or a negative interaction. 103 00:05:38,000 --> 00:05:41,190 Maybe they used something, and they were really happy. 104 00:05:41,190 --> 00:05:43,410 They got a big smile on their face, and that's great. 105 00:05:43,410 --> 00:05:46,860 It's really validating as a designer. 106 00:05:46,860 --> 00:05:51,870 Or they could use something, and you see there's a real stumbling block there. 107 00:05:51,870 --> 00:05:53,400 There's a mistake. 108 00:05:53,400 --> 00:05:55,780 Well, this is also a positive for you 109 00:05:55,780 --> 00:05:59,040 because you know what's broken, and you can fix it. 110 00:05:59,040 --> 00:06:04,570 An outside perspective is very, very valuable. 111 00:06:04,570 --> 00:06:08,140 And another purpose behind testing is to ensure, 112 00:06:08,140 --> 00:06:10,910 like I said at the beginning, what a lot of people think of first, 113 00:06:10,910 --> 00:06:15,330 to ensure that your site or app is working correctly. 114 00:06:15,330 --> 00:06:19,080 Now, this is a benefit that I see 115 00:06:19,080 --> 00:06:21,990 is valuable after a launch. 116 00:06:21,990 --> 00:06:24,720 You want to make sure that your site continues to work, 117 00:06:24,720 --> 00:06:27,320 and you don't want to leave things to chance 118 00:06:27,320 --> 00:06:30,430 or think that no news is good news. 119 00:06:30,430 --> 00:06:34,190 Like if you haven't heard anything about a new section of the site 120 00:06:34,190 --> 00:06:37,900 or a new feature that you've rolled out that you can't assume 121 00:06:37,900 --> 00:06:40,280 that it's okay or that it's working correctly. 122 00:06:40,280 --> 00:06:42,850 You do have to test these things. 123 00:06:42,850 --> 00:06:45,120 You have to ensure that everything is working, 124 00:06:45,120 --> 00:06:50,020 and then you have that peace of mind going forward. 125 00:06:50,020 --> 00:06:53,190 Let's talk about the testing process. 126 00:06:53,190 --> 00:06:57,180 Now, if you have a site or an app 127 00:06:57,180 --> 00:06:59,540 that you are responsible for, 128 00:06:59,540 --> 00:07:03,680 that you are building, you do want to make sure that you test, 129 00:07:03,680 --> 00:07:07,640 and you want this to be part of the design and development process. 130 00:07:09,450 --> 00:07:13,030 You need access to something that works. 131 00:07:13,030 --> 00:07:17,450 Now, you can theoretically test a mockup 132 00:07:17,450 --> 00:07:22,430 or a screenshot or something you rough up in Photoshop. 133 00:07:22,430 --> 00:07:24,390 Even something on a cocktail napkin. 134 00:07:24,390 --> 00:07:26,280 You can ask someone's opinion 135 00:07:26,280 --> 00:07:30,790 about the things you design or a potential new design, 136 00:07:30,790 --> 00:07:32,370 a fix for something. 137 00:07:32,370 --> 00:07:35,880 Getting information from someone else, that outside perspective I talked about, 138 00:07:35,880 --> 00:07:38,110 is incredibly valuable. 139 00:07:38,110 --> 00:07:40,170 But to get the most information you can 140 00:07:40,170 --> 00:07:43,690 and to get the most relevant and actionable information you can, 141 00:07:43,690 --> 00:07:45,710 you have to have something that's working. 142 00:07:45,710 --> 00:07:47,830 You have to have a site that is live 143 00:07:47,830 --> 00:07:51,370 or in development, something on a staging server 144 00:07:51,370 --> 00:07:55,790 or an app that you can give a mobile device to someone 145 00:07:55,790 --> 00:07:59,280 and they can start using it or manipulating it 146 00:07:59,280 --> 00:08:02,330 or creating content on it, whatever it's used for. 147 00:08:02,330 --> 00:08:05,230 You have to have something that works. 148 00:08:05,230 --> 00:08:09,240 Dealing in the generalities of giving people screenshots, 149 00:08:09,240 --> 00:08:13,480 and you can ask questions like "Where would you click?" 150 00:08:13,480 --> 00:08:17,760 And you would get an answer, and you can try and get some more information that way. 151 00:08:17,760 --> 00:08:20,550 But to lead them through an entire process, 152 00:08:20,550 --> 00:08:24,220 take them from screen to screen to screen to screen, 153 00:08:24,220 --> 00:08:28,050 from beginning to finish, from the home page to the ending of a shopping cart, 154 00:08:28,050 --> 00:08:32,250 that's the spot where you get the most detail, 155 00:08:32,250 --> 00:08:36,000 and you get the most information that will help you. 156 00:08:36,000 --> 00:08:38,400 Determining the tasks to test. 157 00:08:38,400 --> 00:08:41,190 This is vital for your testing process, 158 00:08:41,190 --> 00:08:45,900 and it's something that I did talk about a little earlier. 159 00:08:45,900 --> 00:08:48,340 What are those things that really drive your site? 160 00:08:48,340 --> 00:08:54,910 What are the things that really tie into the metrics of success, if you will? 161 00:08:54,910 --> 00:09:01,060 Revenue, leads that you generate from a contact form, 162 00:09:01,060 --> 00:09:06,350 clients that you get from your portfolio site or your personal site, 163 00:09:06,350 --> 00:09:09,140 sales or clicks or downloads, 164 00:09:09,140 --> 00:09:12,580 whatever the metric is that is important to you, 165 00:09:12,580 --> 00:09:14,940 there is a process, there is a feature, 166 00:09:14,940 --> 00:09:17,920 there is a section of the site that is tied to it 167 00:09:17,920 --> 00:09:20,910 so that when people click something or download something 168 00:09:20,910 --> 00:09:25,090 or submit a form that's a success to you. 169 00:09:25,090 --> 00:09:28,430 That is the process that you want to test. 170 00:09:28,430 --> 00:09:35,120 Chances are, you have 1, 2, perhaps 3 of those things that are very important to you 171 00:09:35,120 --> 00:09:39,290 and your company on the site, and that's fine. 172 00:09:39,290 --> 00:09:42,980 If you had, say, 3 tasks that you wanted to test 173 00:09:42,980 --> 00:09:45,330 at a particular time, that's a great number. 174 00:09:45,330 --> 00:09:49,500 You don't want to test every single page on the site. 175 00:09:49,500 --> 00:09:51,570 Maybe you have a small site, and it's possible. 176 00:09:51,570 --> 00:09:55,790 But you have to think about how you can scale these tests out 177 00:09:55,790 --> 00:09:58,490 and about how much capacity you have, 178 00:09:58,490 --> 00:10:00,420 how many resources you have. 179 00:10:00,420 --> 00:10:03,260 I'll talk more about administering the tests 180 00:10:03,260 --> 00:10:06,700 and some of those numbers a little later. 181 00:10:06,700 --> 00:10:08,890 But you want to stay focused here. 182 00:10:08,890 --> 00:10:11,430 What are the most important things for your site? 183 00:10:11,430 --> 00:10:15,050 Focus on those so you can test. 184 00:10:15,050 --> 00:10:17,470 Identifying personas. 185 00:10:17,470 --> 00:10:20,960 I don't want to scare anyone off with the persona buzzword. 186 00:10:20,960 --> 00:10:25,240 It's really all comes down to who is going to use your site? 187 00:10:25,240 --> 00:10:27,980 If it's a site that's online, 188 00:10:27,980 --> 00:10:30,490 who uses your site right now? 189 00:10:30,490 --> 00:10:32,500 Who are they? Who are these people? 190 00:10:32,500 --> 00:10:36,220 If your site is about a particular type of content 191 00:10:36,220 --> 00:10:39,370 it's clearly going to be people that like that content. 192 00:10:39,370 --> 00:10:43,860 If you have a site about greatest movie villains of all time, 193 00:10:43,860 --> 00:10:46,110 these are people that love movies. 194 00:10:46,110 --> 00:10:48,280 They love entertainment. 195 00:10:48,280 --> 00:10:51,910 If this is a site about learning how to code, these are people that are learning how to code 196 00:10:51,910 --> 00:10:56,500 or refreshing their skills, kind of like teamtreehouse.com. 197 00:10:56,500 --> 00:11:01,830 But figure out based on their needs 198 00:11:01,830 --> 00:11:06,150 and their interests as well as demographic data if you have it. 199 00:11:06,150 --> 00:11:09,030 So male or female, a particular age group. 200 00:11:09,030 --> 00:11:11,620 Are they educated? 201 00:11:11,620 --> 00:11:14,940 Do they live in a certain part of the world? 202 00:11:14,940 --> 00:11:19,140 Do they live in a certain part of your town? 203 00:11:19,140 --> 00:11:21,970 You can use all this information to your advantage, 204 00:11:21,970 --> 00:11:25,230 and you want to target those people 205 00:11:25,230 --> 00:11:28,720 who would generally use your site or that you want to use your site. 206 00:11:28,720 --> 00:11:32,210 If you get some of those types of people 207 00:11:32,210 --> 00:11:37,570 to test your site, you're getting the most insightful data that you can. 208 00:11:37,570 --> 00:11:43,100 If you have people that wouldn't normally use your site testing it, 209 00:11:43,100 --> 00:11:47,900 you can get some information, but it's not quite as valuable 210 00:11:47,900 --> 00:11:50,090 as picking the right people. 211 00:11:50,090 --> 00:11:52,780 But I will say this: testing with 1 person 212 00:11:52,780 --> 00:11:56,510 is always better than testing with nobody. 213 00:11:56,510 --> 00:12:00,170 Before you test, you want to locate a place in which to test. 214 00:12:00,170 --> 00:12:02,910 This is really important. 215 00:12:02,910 --> 00:12:05,950 You want to have a controlled environment 216 00:12:05,950 --> 00:12:07,730 in which to test. 217 00:12:07,730 --> 00:12:12,840 And what I mean by that is if you want to test in a public place, 218 00:12:12,840 --> 00:12:14,840 you're going to have distractions. 219 00:12:14,840 --> 00:12:18,000 If you're at the local coffee shop 220 00:12:18,000 --> 00:12:22,970 or on a bus or at a bus stop or out and about somewhere, 221 00:12:22,970 --> 00:12:27,880 again, you can get information from people and helpful feedback. 222 00:12:27,880 --> 00:12:31,850 But if you have a conference room at your office 223 00:12:31,850 --> 00:12:37,550 or you have a closed off area or you have even your house, 224 00:12:37,550 --> 00:12:41,560 asking a friend or family or girlfriend or spouse or whatever 225 00:12:41,560 --> 00:12:45,390 to take a look at a site and asking them questions 226 00:12:45,390 --> 00:12:50,240 in the privacy of your own home or asking a colleague 227 00:12:50,240 --> 00:12:52,730 in a conference room questions about things, 228 00:12:52,730 --> 00:12:56,300 those are controlled spaces because you know you're not going to have 229 00:12:56,300 --> 00:12:58,430 the distractions of the outside world. 230 00:12:58,430 --> 00:13:00,080 You can control that space. 231 00:13:00,080 --> 00:13:02,160 You can set up equipment to record. 232 00:13:02,160 --> 00:13:05,610 You can set up your computer or mobile device, whatever you're testing on. 233 00:13:05,610 --> 00:13:10,090 It is so much more easy 234 00:13:10,090 --> 00:13:14,760 and less stressful than trying to test in public 235 00:13:14,760 --> 00:13:17,120 and having who knows what will happen 236 00:13:17,120 --> 00:13:19,130 as you're trying to ask questions. 237 00:13:19,130 --> 00:13:21,190 You want to get as much information as you can, 238 00:13:21,190 --> 00:13:25,040 so you need to be focused, and the test subjects need to be focused as well. 239 00:13:25,040 --> 00:13:27,460 So once you have an idea 240 00:13:27,460 --> 00:13:31,110 about what you want to test, you have your site 241 00:13:31,110 --> 00:13:34,230 or your app, you have some tasks in mind, 242 00:13:34,230 --> 00:13:37,420 you know what type of people you want to test with, 243 00:13:37,420 --> 00:13:41,900 the first thing you want to do is set up that test environment. 244 00:13:41,900 --> 00:13:44,670 Now, that includes the physical space that I just talked about, 245 00:13:44,670 --> 00:13:49,600 but it also includes the site or the app itself. 246 00:13:49,600 --> 00:13:53,370 Make sure that you have access to it 247 00:13:53,370 --> 00:13:55,480 when you're going to be ready to test. 248 00:13:55,480 --> 00:13:57,270 And if we're talking about the day of the test, 249 00:13:57,270 --> 00:14:00,200 well, make sure it works, 250 00:14:00,200 --> 00:14:02,000 but be ready for it. 251 00:14:02,000 --> 00:14:05,290 Think about these things the day before or the week before. 252 00:14:05,290 --> 00:14:09,320 Make sure you have access, so if you're working in a company, for example, 253 00:14:09,320 --> 00:14:14,340 and you have to reserve a conference room for an entire day, do that in advance. 254 00:14:14,340 --> 00:14:17,320 Don't spring that upon people. 255 00:14:17,320 --> 00:14:22,290 If you are testing a development version of a website 256 00:14:22,290 --> 00:14:26,230 and you want to have access to it, and you need to talk to some developers 257 00:14:26,230 --> 00:14:29,880 or server guys or something to make sure you have access, 258 00:14:29,880 --> 00:14:31,300 talk to those people. 259 00:14:31,300 --> 00:14:33,950 Let them know about it. 260 00:14:33,950 --> 00:14:37,440 It's important to also do this in advance. 261 00:14:37,440 --> 00:14:39,620 You have to draft test users. 262 00:14:39,620 --> 00:14:42,130 You have to identify those people 263 00:14:42,130 --> 00:14:45,460 that based on those personas we discussed, 264 00:14:45,460 --> 00:14:48,630 the demographics, what are their interests, what are their needs, 265 00:14:48,630 --> 00:14:51,740 things like that, start targeting people. 266 00:14:51,740 --> 00:14:53,480 See who you can find, 267 00:14:53,480 --> 00:14:56,280 and I'll tell you about an example. 268 00:14:56,280 --> 00:14:58,680 I'll tell you about a scenario that I went through. 269 00:14:58,680 --> 00:15:02,700 I was testing the old version of the Treehouse site last year, 270 00:15:02,700 --> 00:15:06,540 and I needed some test users, and we're here in Orlando. 271 00:15:06,540 --> 00:15:11,450 But I didn't want to use anyone who was already here in the office 272 00:15:11,450 --> 00:15:14,840 and knows a lot about our website. 273 00:15:14,840 --> 00:15:17,840 I wanted to get people who were relatively new to it 274 00:15:17,840 --> 00:15:20,640 or completely new to the site, 275 00:15:20,640 --> 00:15:26,440 and I wanted to ask them questions about how easy it is to go through 276 00:15:26,440 --> 00:15:30,610 and learn different things, were there any stumbling blocks, things like that. 277 00:15:30,610 --> 00:15:33,740 Well, the way that I reached out to find people 278 00:15:33,740 --> 00:15:36,470 was I went through our Meetup.com group. 279 00:15:36,470 --> 00:15:40,020 Meetup.com, if you haven't been there, 280 00:15:40,020 --> 00:15:46,810 is a site where you can set up events, and you can set up a schedule of different things 281 00:15:46,810 --> 00:15:50,100 for organizations or local groups, things like that, 282 00:15:50,100 --> 00:15:52,000 and invite them, and you can hold events, 283 00:15:52,940 --> 00:15:56,010 and it has a nice schedule for you, RSVPs online, things like that. 284 00:15:56,010 --> 00:16:00,840 I put out a call for test users for usability testing, 285 00:16:00,840 --> 00:16:03,990 and I gave the instructions, and I said, 286 00:16:03,990 --> 00:16:08,940 "I'm looking for people who are interested in web design 287 00:16:08,940 --> 00:16:14,810 or development or making apps," which is what we do at teamtreehouse.com. 288 00:16:14,810 --> 00:16:21,080 "I'm looking for people who are not members of our site," 289 00:16:21,080 --> 00:16:26,300 so I tried to get people who had not been there before or maybe visited once or twice. 290 00:16:26,300 --> 00:16:29,700 "And can you be available on this date, 291 00:16:29,700 --> 00:16:35,710 this particular time, or can you give me an hour of your time on this day?" 292 00:16:35,710 --> 00:16:38,930 That's actually what I did, because I ended up getting 293 00:16:38,930 --> 00:16:44,410 maybe 10 or 12 different people who responded positively. 294 00:16:44,410 --> 00:16:47,600 They said, "Oh, yeah, I'm available, and I can be there, 295 00:16:47,600 --> 00:16:50,640 and these are the hours that I can be available," 296 00:16:50,640 --> 00:16:54,800 whether they're on a lunch break or it's the morning or the afternoon. 297 00:16:54,800 --> 00:16:59,470 Based on that, I was able to get some information from them 298 00:16:59,470 --> 00:17:01,490 by setting up a survey. 299 00:17:01,490 --> 00:17:03,240 I went into surveymonkey.com, 300 00:17:03,240 --> 00:17:07,160 a great online survey tool, which I hope many of you are familiar with. 301 00:17:07,160 --> 00:17:11,079 It's surveymonkey.com. I set up a survey and asked them questions. 302 00:17:11,079 --> 00:17:13,119 I asked them have they been to the site before. 303 00:17:13,119 --> 00:17:17,060 I asked them what technologies are they interested in, 304 00:17:17,060 --> 00:17:20,400 and I listed a number of things that we offer on the site 305 00:17:20,400 --> 00:17:23,349 and a few things that we don't. 306 00:17:23,349 --> 00:17:25,349 And then I asked what their availability was, 307 00:17:25,349 --> 00:17:28,540 and then based on all of that information I was able to see 308 00:17:28,540 --> 00:17:32,670 who could look at specific types of content 309 00:17:32,670 --> 00:17:35,260 we had on the site, who would be available, 310 00:17:35,260 --> 00:17:37,380 who would fill in the slots on my schedule here, 311 00:17:37,380 --> 00:17:40,430 and I was able to schedule 6 people, 312 00:17:40,430 --> 00:17:44,490 an hour each, 3 before lunch and 3 after, 313 00:17:44,490 --> 00:17:48,770 and I was able to bring them all in and get some great information. 314 00:17:48,770 --> 00:17:51,950 That's how I was able to find 315 00:17:51,950 --> 00:17:56,310 people that met the criteria that I had set 316 00:17:56,310 --> 00:18:01,360 based on who we would expect to use the site 317 00:18:01,360 --> 00:18:06,740 and some other demographics as well. 318 00:18:06,740 --> 00:18:09,050 That's one example of how you can draft people 319 00:18:09,050 --> 00:18:12,170 who happen to be in the area and have the same interests 320 00:18:12,170 --> 00:18:15,440 or same needs that you're offering with your site or app. 321 00:18:15,440 --> 00:18:21,500 It's something that a number of people who are interested and sign up for those sites 322 00:18:21,500 --> 00:18:24,010 are already interested. 323 00:18:24,010 --> 00:18:26,450 If you give them an opportunity like that, 324 00:18:26,450 --> 00:18:30,350 you'll find that there's plenty of people that would take you up on it. 325 00:18:30,350 --> 00:18:35,360 Before you test, again, it's important not to insulate yourself 326 00:18:35,360 --> 00:18:37,330 from the rest of the company. 327 00:18:37,330 --> 00:18:40,490 Testing should be something that you share 328 00:18:40,490 --> 00:18:44,860 with other people, especially if, say, you're a designer 329 00:18:44,860 --> 00:18:46,760 and you have a development team over there, 330 00:18:46,760 --> 00:18:50,050 or you could be on the marketing side, 331 00:18:50,050 --> 00:18:54,150 online marketing, and you have the people that build the site over there. 332 00:18:54,150 --> 00:18:57,920 You want to be able to talk with them, communicate to them 333 00:18:57,920 --> 00:18:59,570 about the testing process, 334 00:18:59,570 --> 00:19:02,280 because after you do testing, 335 00:19:02,280 --> 00:19:04,090 you're going to have results for them. 336 00:19:04,090 --> 00:19:06,620 You're going to have some great information to share, 337 00:19:06,620 --> 00:19:13,320 and it's a little disconcerting when you're part of the team that works on the site 338 00:19:13,320 --> 00:19:15,510 and makes this thing happen and you're passionate about it 339 00:19:15,510 --> 00:19:18,430 and then someone comes to you from out of nowhere and says, 340 00:19:18,430 --> 00:19:20,180 "All right, I've got a bunch of changes for the site, 341 00:19:20,180 --> 00:19:24,570 and you're going to make them," or something like that. 342 00:19:24,570 --> 00:19:28,500 Without being part of that process it can be difficult, 343 00:19:28,500 --> 00:19:35,590 and this is more of an interpersonal, interdepartmental sort of diplomacy type deal, 344 00:19:35,590 --> 00:19:39,200 but you can get some great insight from them as well 345 00:19:39,200 --> 00:19:42,200 because there may be things they really want to test, 346 00:19:42,200 --> 00:19:44,430 but they haven't had an opportunity, 347 00:19:44,430 --> 00:19:49,020 or they may be very curious about what people think about a new section 348 00:19:49,020 --> 00:19:51,560 or a potential redesign of something. 349 00:19:51,560 --> 00:19:54,900 You may have some great stuff to test, 350 00:19:54,900 --> 00:19:58,110 and giving that opportunity, since you're setting something up already, 351 00:19:58,110 --> 00:20:01,770 they can be a great resource for you. 352 00:20:01,770 --> 00:20:04,990 One of the last things before setting up your test 353 00:20:04,990 --> 00:20:08,050 is to write up a testing script. 354 00:20:08,050 --> 00:20:13,740 Now, this is one of those things that I think some people might overlook, 355 00:20:13,740 --> 00:20:17,640 because they have these ideas, they know what they want to test, 356 00:20:17,640 --> 00:20:20,010 and maybe it's like a little checklist or something like that. 357 00:20:20,010 --> 00:20:23,410 They have 3 questions they want to ask or something like that. 358 00:20:23,410 --> 00:20:28,900 But if you have a script, a script can put you at ease. 359 00:20:28,900 --> 00:20:31,830 It can put the test subject at ease. 360 00:20:31,830 --> 00:20:33,440 It's very, very helpful. 361 00:20:33,440 --> 00:20:37,580 A place where you can get a script to use is right here, 362 00:20:37,580 --> 00:20:40,950 Steve Krug's "Don't Make Me Think." 363 00:20:40,950 --> 00:20:42,360 This is a fantastic book. 364 00:20:42,360 --> 00:20:44,500 It came out in 2000. 365 00:20:44,500 --> 00:20:48,490 It is one of the industry-changing books, I'll say, 366 00:20:48,490 --> 00:20:51,750 about usability, its importance to your site, 367 00:20:51,750 --> 00:20:55,970 and how you can improve through testing. 368 00:20:55,970 --> 00:21:01,160 Included in that book and also on Steve Krug's website 369 00:21:01,160 --> 00:21:05,540 sensible.com, he has a number of resources you can download 370 00:21:05,540 --> 00:21:09,510 including a test script, which is really a template you can fill in 371 00:21:09,510 --> 00:21:12,900 all the information you have. 372 00:21:12,900 --> 00:21:14,330 Also a recording consent form. 373 00:21:14,330 --> 00:21:17,070 We've talk about recording in a minute. 374 00:21:17,070 --> 00:21:19,300 He's got some other stuff on there too. 375 00:21:19,300 --> 00:21:26,470 But the importance of the script is that as you're administering these tests 376 00:21:26,470 --> 00:21:29,070 you've got people coming in from the outside, 377 00:21:29,070 --> 00:21:35,880 you've got a testing environment, and you're trying to keep everything together, 378 00:21:35,880 --> 00:21:39,560 and you've got a lot of things happening, and this is probably generally not what you'd do 379 00:21:39,560 --> 00:21:42,060 from day to day, so you've already got lots to think about. 380 00:21:42,060 --> 00:21:45,100 Relieve some of the stress 381 00:21:45,100 --> 00:21:49,310 and have a script all written out for yourself 382 00:21:49,310 --> 00:21:52,510 once you start going through the process. 383 00:21:52,510 --> 00:21:56,720 And really, it's as simple as looking at it and reading it 384 00:21:56,720 --> 00:21:58,890 and actually telling the person who came in, 385 00:21:58,890 --> 00:22:00,820 "I'm going to read from the script, 386 00:22:00,820 --> 00:22:03,260 "because I don't want to forget anything. 387 00:22:03,260 --> 00:22:06,520 I'm going to go through point by point, and if you have any questions, let me know." 388 00:22:06,520 --> 00:22:09,250 But the script is the easiest way for you to organize your thoughts 389 00:22:09,250 --> 00:22:12,960 before you even get to this craziness of doing tests, 390 00:22:12,960 --> 00:22:16,900 and it can be very stressful for people. 391 00:22:16,900 --> 00:22:19,010 But the script is a great crutch. 392 00:22:19,010 --> 00:22:22,630 It's very helpful, and you can put everything you want in there, 393 00:22:22,630 --> 00:22:29,880 and then you repeat the process with each new test user that comes in. 394 00:22:29,880 --> 00:22:33,370 During your test, or I should say right before your test, 395 00:22:33,370 --> 00:22:35,660 you're going to prep that controlled space. 396 00:22:35,660 --> 00:22:39,050 You're going to get everything ready 397 00:22:39,050 --> 00:22:41,560 from your laptop, your mobile device, 398 00:22:41,560 --> 00:22:44,430 get a schedule together, 399 00:22:44,430 --> 00:22:47,330 print out forms like recording consent forms, 400 00:22:47,330 --> 00:22:49,650 print out your script that you've prepared, 401 00:22:49,650 --> 00:22:53,550 and one other thing I know was on a slide earlier 402 00:22:53,550 --> 00:22:56,380 when we talked about drafting users 403 00:22:56,380 --> 00:22:58,930 there's also the issue of compensating them. 404 00:22:58,930 --> 00:23:01,200 A lot of people are more than happy to help, 405 00:23:01,200 --> 00:23:04,200 they're just busy, and their time is worth something. 406 00:23:04,200 --> 00:23:06,090 It really is worth something to them. 407 00:23:06,090 --> 00:23:09,590 But their opinion is also worth something to you as well, 408 00:23:09,590 --> 00:23:13,170 so compensate them for that. 409 00:23:13,170 --> 00:23:16,120 A standard thing could be $50 or $100. 410 00:23:16,120 --> 00:23:18,590 It could be a gift card. It could be cash. 411 00:23:18,590 --> 00:23:22,120 If you happen to be right near a coffee shop, 412 00:23:22,120 --> 00:23:25,030 maybe pick up a Starbucks gift card and say, "Here you go. 413 00:23:25,030 --> 00:23:28,260 Thanks for your help," but some kind of compensation. 414 00:23:28,260 --> 00:23:31,580 Maybe a free membership to the site that you're testing, 415 00:23:31,580 --> 00:23:35,220 maybe a discount code or something like that. 416 00:23:35,220 --> 00:23:37,620 Make sure you compensate them for their time, 417 00:23:37,620 --> 00:23:40,920 because they'll appreciate it, and they'll also be more willing 418 00:23:40,920 --> 00:23:44,590 to give up an hour of their life or however long it is 419 00:23:44,590 --> 00:23:47,670 and spend it with you testing a site. 420 00:23:47,670 --> 00:23:50,750 Now, a part of prepping that space 421 00:23:50,750 --> 00:23:55,450 is getting all that stuff, whether you have to talk to a financial person 422 00:23:55,450 --> 00:23:59,500 in your office or go to the bank and get some cash. 423 00:23:59,500 --> 00:24:01,850 Get all those things ready to go, 424 00:24:01,850 --> 00:24:04,600 because then you can even have a folder 425 00:24:04,600 --> 00:24:06,580 set up for each test user that comes in. 426 00:24:06,580 --> 00:24:08,550 You just open the next folder. 427 00:24:08,550 --> 00:24:10,960 Here's the script. Here's the thing you have to sign. 428 00:24:10,960 --> 00:24:12,800 Here's that compensation. 429 00:24:12,800 --> 00:24:14,460 Once we get all this stuff in there, 430 00:24:14,460 --> 00:24:16,620 folders closed, move onto the next one. 431 00:24:16,620 --> 00:24:19,970 It makes it nice and easy. 432 00:24:19,970 --> 00:24:21,900 Start recording. 433 00:24:21,900 --> 00:24:23,390 Oh, my goodness, if you're recording, 434 00:24:23,390 --> 00:24:27,310 I mean if you're doing a test and you don't record the test, 435 00:24:27,310 --> 00:24:30,990 you're absolutely missing out on something. 436 00:24:30,990 --> 00:24:35,100 Recording what happens on the screen, 437 00:24:35,100 --> 00:24:39,110 recording the person's face and all that stuff 438 00:24:39,110 --> 00:24:42,400 is very helpful after the fact. 439 00:24:42,400 --> 00:24:45,030 Being able to go back and watch a replay 440 00:24:45,030 --> 00:24:49,200 of what the test subject was doing, 441 00:24:49,200 --> 00:24:52,360 what they were talking about, 442 00:24:52,360 --> 00:24:55,470 where their mouse was going, their body language, 443 00:24:55,470 --> 00:24:57,420 their face, all these different things, 444 00:24:57,420 --> 00:24:59,050 and especially the things they say, 445 00:24:59,050 --> 00:25:02,720 any mistakes they make, 446 00:25:02,720 --> 00:25:05,330 all these things that happen during the test, 447 00:25:05,330 --> 00:25:09,340 they can easily be missed because you're trying to administer this thing. 448 00:25:09,340 --> 00:25:11,090 And again, lots of stuff is happening. 449 00:25:11,090 --> 00:25:13,750 You've got paperwork. 450 00:25:13,750 --> 00:25:16,970 You're trying to listen. You're trying to engage them. 451 00:25:16,970 --> 00:25:23,060 Record. Record, and again, take some of that stress away from this whole process. 452 00:25:23,060 --> 00:25:27,050 You can go back and watch a recording of the test as many times as you want, 453 00:25:27,050 --> 00:25:34,180 and you can get some great insight that you missed the first time. 454 00:25:34,180 --> 00:25:37,750 I'm going to give you an example of a testing setup that we had. 455 00:25:37,750 --> 00:25:41,870 When we did the test of the teamtreehouse.com site last year, 456 00:25:41,870 --> 00:25:44,640 I went into a conference room here in Orlando, 457 00:25:44,640 --> 00:25:47,350 set up my laptop, 458 00:25:47,350 --> 00:25:49,970 and I had my notes, and I had my folder, 459 00:25:49,970 --> 00:25:53,860 my glass of water right next to me, 460 00:25:53,860 --> 00:25:58,440 and sitting with me there is one of our friends that came in to test the site. 461 00:25:58,440 --> 00:26:01,170 He has a computer in front of him 462 00:26:01,170 --> 00:26:05,330 where he's looking at the Treehouse site, 463 00:26:05,330 --> 00:26:07,070 and you can see a screenshot of it there, 464 00:26:07,070 --> 00:26:09,000 and I apologize if that's pixelated. 465 00:26:09,000 --> 00:26:11,710 It's not what actually happened with the recording. 466 00:26:11,710 --> 00:26:15,990 That's a screenshot of a screenshot of a video recording 467 00:26:15,990 --> 00:26:18,550 of a pixelated something. 468 00:26:18,550 --> 00:26:22,850 But what you can see there is that the website is there. 469 00:26:22,850 --> 00:26:24,650 That's an old version of the Treehouse site. 470 00:26:24,650 --> 00:26:29,230 And sort of picture in picture you have a web cam 471 00:26:29,230 --> 00:26:32,940 recording of our test user. 472 00:26:32,940 --> 00:26:36,770 Going back 1 slide, you might be able to see on my laptop 473 00:26:36,770 --> 00:26:41,440 I actually have Camtasia loaded up on my laptop, 474 00:26:41,440 --> 00:26:45,470 and that is the software I used to record that session. 475 00:26:45,470 --> 00:26:47,650 Camtasia 2 is out now. 476 00:26:47,650 --> 00:26:50,250 That's a software that I recommend. 477 00:26:50,250 --> 00:26:53,940 It is a great multi-purpose tool actually. 478 00:26:53,940 --> 00:27:00,440 I think they probably developed for user testing and screen recording, things like that, 479 00:27:00,440 --> 00:27:04,200 but you can use it for recording yourself off of web cams 480 00:27:04,200 --> 00:27:06,990 to do presentations. 481 00:27:06,990 --> 00:27:10,550 You can use it for video editing and audio editing. 482 00:27:10,550 --> 00:27:13,060 It's certainly no Final Cut Pro or anything like that, 483 00:27:13,060 --> 00:27:15,860 but as an affordable piece of software 484 00:27:15,860 --> 00:27:19,910 that can allow you to record the screen 485 00:27:19,910 --> 00:27:24,960 during tests like this and have that web cam shot of the user so you can see them too, 486 00:27:24,960 --> 00:27:29,390 for me it's a bargain, and like I said, it's a great multipurpose tool. 487 00:27:29,390 --> 00:27:34,220 But that is what I used to record the test, 488 00:27:34,220 --> 00:27:42,100 and I was able to go back and watch those recordings any time that I wanted. 489 00:27:42,100 --> 00:27:44,970 And of course, make sure to leave yourself a note 490 00:27:44,970 --> 00:27:48,000 to start recording. 491 00:27:48,000 --> 00:27:50,150 It's a nice thing to build into your script, 492 00:27:50,150 --> 00:27:53,120 and that's why I say use your script and take notes 493 00:27:53,120 --> 00:27:55,230 throughout the entire process. 494 00:27:55,230 --> 00:27:57,100 The script is there to help you. 495 00:27:57,100 --> 00:27:59,070 Make whatever notes you want 496 00:27:59,070 --> 00:28:01,000 inside the script. 497 00:28:01,000 --> 00:28:03,540 You don't have to read every single thing there. 498 00:28:03,540 --> 00:28:05,900 If there's a big, giant note in there 499 00:28:05,900 --> 00:28:09,190 in all caps and it's this big and it says, "Start recording," 500 00:28:09,190 --> 00:28:11,130 that's a great reminder. 501 00:28:11,130 --> 00:28:14,450 And then at the end you can say, "Stop recording" when it's done, 502 00:28:14,450 --> 00:28:18,730 and you'll get everything that you wanted to get. 503 00:28:18,730 --> 00:28:21,800 Try to take notes as much as you can 504 00:28:21,800 --> 00:28:23,980 during the tests. 505 00:28:23,980 --> 00:28:27,010 That's a great time to take notes is during the test. 506 00:28:27,010 --> 00:28:30,450 It's also a great time after the fact 507 00:28:30,450 --> 00:28:32,570 when you can go back and review those things 508 00:28:32,570 --> 00:28:38,750 and watch replays of the sessions and be able to match up those notes 509 00:28:38,750 --> 00:28:43,070 that you took live with maybe some new notes that you take 510 00:28:43,070 --> 00:28:45,170 when you're watching a recording. 511 00:28:45,170 --> 00:28:47,100 Taking notes is great. 512 00:28:47,100 --> 00:28:50,260 However, you have to make this experience 513 00:28:50,260 --> 00:28:54,880 as pleasing, as effortless, 514 00:28:54,880 --> 00:28:58,770 as easy as you can for whoever comes in to visit you. 515 00:28:58,770 --> 00:29:01,370 Make sure that you listen 516 00:29:01,370 --> 00:29:03,000 to your test users. 517 00:29:03,000 --> 00:29:06,700 Make sure that you ask them some questions about themselves, 518 00:29:06,700 --> 00:29:13,120 and don't look at them merely as a cog in the machine or something. 519 00:29:13,120 --> 00:29:16,390 They are real human beings. 520 00:29:16,390 --> 00:29:18,780 This is one of the huge points 521 00:29:18,780 --> 00:29:21,900 of testing is to get that other person's perspective. 522 00:29:21,900 --> 00:29:24,260 They are there to help you, 523 00:29:24,260 --> 00:29:29,100 and they're giving up their time to look at whatever site or app you've built. 524 00:29:29,100 --> 00:29:32,090 Listen to them and engage with them. 525 00:29:32,090 --> 00:29:34,200 Have a conversation with them. 526 00:29:34,200 --> 00:29:36,480 Don't take up too much time. You're there for a purpose. 527 00:29:36,480 --> 00:29:39,180 But make them feel at ease. 528 00:29:39,180 --> 00:29:42,280 I know when we had a couple people come in for our test 529 00:29:42,280 --> 00:29:44,670 1 or 2 of them were nervous 530 00:29:44,670 --> 00:29:48,350 because they almost felt like they were the ones being tested, 531 00:29:48,350 --> 00:29:50,390 and that's something you can tell them. 532 00:29:50,390 --> 00:29:53,030 You can explicitly tell them, "Don't worry. 533 00:29:53,030 --> 00:29:57,680 "You can't make any mistakes, because you are not the one that's being tested. 534 00:29:57,680 --> 00:30:02,980 "The site is being tested, and you are here to give us the great benefit 535 00:30:02,980 --> 00:30:07,890 of your opinion and feedback, and we really appreciate you being here." 536 00:30:07,890 --> 00:30:11,690 Tell them that, and what's more, 537 00:30:11,690 --> 00:30:16,790 believe that, because they are a huge benefit to you. 538 00:30:16,790 --> 00:30:20,970 Lastly, during your test this is something to prepare, of course, before. 539 00:30:20,970 --> 00:30:25,090 But during your test, should anything come up 540 00:30:25,090 --> 00:30:27,030 make sure you have a backup plan. 541 00:30:27,030 --> 00:30:30,810 A classic example of that is when a test user doesn't show up, 542 00:30:30,810 --> 00:30:36,800 and I said earlier that I had 6 people signed up, 543 00:30:36,800 --> 00:30:42,330 reserved for times during our all-day test of teamtreehouse.com. 544 00:30:42,330 --> 00:30:44,340 One of them did not show up. 545 00:30:44,340 --> 00:30:46,370 One of them I think emailed me 546 00:30:46,370 --> 00:30:50,240 and said they couldn't make it for whatever reason. 547 00:30:50,240 --> 00:30:56,940 And that's fine, because I overcompensated beforehand. 548 00:30:56,940 --> 00:31:01,610 I would be more than happy to have 3 to 5 test users come in, 549 00:31:01,610 --> 00:31:06,130 and that number actually is based on research by Jacob Nielsen 550 00:31:06,130 --> 00:31:08,920 from way, way back. 551 00:31:08,920 --> 00:31:12,860 Having 3 to 5 people come in and test features on your website 552 00:31:12,860 --> 00:31:18,270 statistically is about the perfect range 553 00:31:18,270 --> 00:31:20,580 of people that you can have, 554 00:31:20,580 --> 00:31:26,020 because less than that number, so 1 or 2 people, 555 00:31:26,020 --> 00:31:30,830 you're going to find less things that 3, 4, or 5 people will, 556 00:31:30,830 --> 00:31:33,730 and based on the data and research Jacob Nielsen did 557 00:31:33,730 --> 00:31:36,520 once you start going past 5 people 558 00:31:36,520 --> 00:31:40,300 the amount of effort that it takes 559 00:31:40,300 --> 00:31:44,590 to run these tests no longer returns 560 00:31:44,590 --> 00:31:47,910 the benefit of doing the testing. 561 00:31:47,910 --> 00:31:50,280 It's the law of diminishing returns right there. 562 00:31:50,280 --> 00:31:54,150 Once you get to 5 test users, 563 00:31:54,150 --> 00:31:57,070 you will probably find as many things 564 00:31:57,070 --> 00:31:59,010 as you are going to find. 565 00:31:59,010 --> 00:32:02,680 If you have 10 test users, 20 test users, 566 00:32:02,680 --> 00:32:05,150 you may find 1 or 2 more things, 567 00:32:05,150 --> 00:32:07,980 but the big ticket items that you have to fix 568 00:32:07,980 --> 00:32:11,710 you probably already found in the first 3, 4, 5 test users. 569 00:32:11,710 --> 00:32:15,080 Don't make so much work for yourself. 570 00:32:15,080 --> 00:32:18,030 Limit yourself to that, and you're going to have more than enough information 571 00:32:18,030 --> 00:32:20,360 from those few people. 572 00:32:20,360 --> 00:32:23,280 Trust me, I did. 573 00:32:23,280 --> 00:32:26,530 Have a backup plan 574 00:32:26,530 --> 00:32:29,120 for different things that may come up. 575 00:32:29,120 --> 00:32:32,440 One of the things that did come up during our test 576 00:32:32,440 --> 00:32:34,980 was for some reason the wi-fi in our office 577 00:32:34,980 --> 00:32:38,790 was slow that day, which really doesn't help 578 00:32:38,790 --> 00:32:42,590 when you have a website that serves up these big videos, 579 00:32:42,590 --> 00:32:46,170 instructional videos to teach you how to do HTML. 580 00:32:46,170 --> 00:32:51,070 The video was kind of jumpy, and one of the things I had ready 581 00:32:51,070 --> 00:32:55,230 was I had the videos I wanted them to watch downloaded on my laptop, 582 00:32:55,230 --> 00:33:00,450 and so I tried to give them the perfect experience, 583 00:33:00,450 --> 00:33:03,410 the classic experience that people would have on the site. 584 00:33:03,410 --> 00:33:05,660 But when I saw that it just wasn't working, 585 00:33:05,660 --> 00:33:10,200 and you can tell that people are getting a little antsy 586 00:33:10,200 --> 00:33:14,950 and they're thinking, "Am I wasting my time? 587 00:33:14,950 --> 00:33:19,310 Should I leave?" give them something that they can do. 588 00:33:19,310 --> 00:33:21,670 Have that as part of your backup plan, 589 00:33:21,670 --> 00:33:24,270 and that way the opportunity isn't wasted, 590 00:33:24,270 --> 00:33:26,650 especially for you. 591 00:33:26,650 --> 00:33:28,030 You brought them in there for a reason. 592 00:33:28,030 --> 00:33:33,350 Get something from them, and if you don't use your backup plan, 593 00:33:33,350 --> 00:33:38,930 save it for the next time, but that backup plan could be you have a version of the site on your laptop 594 00:33:38,930 --> 00:33:40,760 or the computer where you're testing. 595 00:33:40,760 --> 00:33:43,840 Potentially you could have screenshots of the different things, 596 00:33:43,840 --> 00:33:46,680 mockups, things like that. 597 00:33:46,680 --> 00:33:51,560 That could be part of your backup plan right there. 598 00:33:51,560 --> 00:33:56,110 After you test, certainly thank these people for their time. 599 00:33:56,110 --> 00:33:58,070 That's one thing. 600 00:33:58,070 --> 00:34:00,450 Show that you value them. 601 00:34:00,450 --> 00:34:04,290 But after your test, you will really see the true value 602 00:34:04,290 --> 00:34:08,949 of all this effort that you've put into the testing. 603 00:34:08,949 --> 00:34:11,719 Reviewing the recordings and your notes 604 00:34:11,719 --> 00:34:17,030 will be your top priority, and it should be something that you do very quickly. 605 00:34:17,030 --> 00:34:21,770 It doesn't have to be that day, especially if you have a full day of testing that you've done. 606 00:34:21,770 --> 00:34:25,590 But it should be at least that week or the very next day 607 00:34:25,590 --> 00:34:28,590 where you start looking at everything, 608 00:34:28,590 --> 00:34:31,600 because if you let it go for too long, 609 00:34:31,600 --> 00:34:33,750 it's going to be difficult to get back into it, 610 00:34:33,750 --> 00:34:36,270 and also things may change on the site 611 00:34:36,270 --> 00:34:38,770 that are no longer applicable. 612 00:34:38,770 --> 00:34:40,050 Get right to it. 613 00:34:40,050 --> 00:34:44,389 Show some value for what you've produced there. 614 00:34:44,389 --> 00:34:46,389 Produce a list of results. 615 00:34:46,389 --> 00:34:49,679 By going through the recordings and the notes, 616 00:34:49,679 --> 00:34:52,380 you will begin to see patterns. 617 00:34:52,380 --> 00:34:56,739 You'll see positive patterns like I asked them to do this, 618 00:34:56,739 --> 00:34:58,630 and they did it with no problem, 619 00:34:58,630 --> 00:35:00,500 and that test user did it, 620 00:35:00,500 --> 00:35:03,440 #2, #3, #4, they all did it successfully. 621 00:35:03,440 --> 00:35:05,570 Maybe they did it slightly different ways, 622 00:35:05,570 --> 00:35:08,450 but it all worked out, and it was perfect. 623 00:35:08,450 --> 00:35:10,320 That's great. 624 00:35:10,320 --> 00:35:12,720 Share those results. 625 00:35:12,720 --> 00:35:14,100 Don't just share problems that you find 626 00:35:14,100 --> 00:35:18,790 like there was this typo or people had problems 627 00:35:18,790 --> 00:35:20,820 performing whatever task. 628 00:35:20,820 --> 00:35:22,150 Don't just report the bad news. 629 00:35:22,150 --> 00:35:24,730 People want to hear the good news as well, and besides, 630 00:35:24,730 --> 00:35:29,230 good news makes bad news a little easier to take sometimes. 631 00:35:29,230 --> 00:35:33,570 By producing a list of results based on the different tasks 632 00:35:33,570 --> 00:35:38,290 you can easily organize that information, 633 00:35:38,290 --> 00:35:41,200 and whoever is going to be looking at this information 634 00:35:41,200 --> 00:35:43,240 will be able to see some patterns. 635 00:35:43,240 --> 00:35:46,300 They'll be able to see where certain things broke down. 636 00:35:46,300 --> 00:35:51,080 You can also break down the results by the test user. 637 00:35:51,080 --> 00:35:55,220 You can look for patterns there as well 638 00:35:55,220 --> 00:36:00,990 where one type of user who maybe had no experience with your site or content or features 639 00:36:00,990 --> 00:36:05,020 or things like that, maybe they had consistent problems with each task. 640 00:36:05,020 --> 00:36:08,800 Maybe someone who was more experienced had no problems with it. 641 00:36:08,800 --> 00:36:11,780 Maybe that says something about your site or app. 642 00:36:11,780 --> 00:36:15,650 These patterns could be based on anything. 643 00:36:15,650 --> 00:36:17,820 It could be demographics. It could be age. 644 00:36:17,820 --> 00:36:21,580 It could be experience. It could be interest in the content or the features. 645 00:36:21,580 --> 00:36:24,190 You just don't know until you look at the data. 646 00:36:24,190 --> 00:36:30,950 Now, you can have very, very granular results. 647 00:36:30,950 --> 00:36:36,270 I know I had a few pages of bullets and different notes 648 00:36:36,270 --> 00:36:39,030 for the test that we ran with 5 users, 649 00:36:39,030 --> 00:36:43,110 and I asked them to perform 3 different tasks, 650 00:36:43,110 --> 00:36:45,020 or 2, 2 or 3 different tasks. 651 00:36:45,020 --> 00:36:47,430 I had plenty of data to share 652 00:36:47,430 --> 00:36:49,950 with the design team here at Treehouse. 653 00:36:49,950 --> 00:36:54,550 But I was also able to take those results 654 00:36:54,550 --> 00:36:59,280 and translate them into the main ideas, 655 00:36:59,280 --> 00:37:06,340 the main takeaways, and I remember I had 3 specific things that I knew had to be fixed. 656 00:37:06,340 --> 00:37:08,290 I knew there was 1 particular page 657 00:37:08,290 --> 00:37:11,340 that was confusing for them where they started. 658 00:37:11,340 --> 00:37:18,130 There was 1 particular task where there was confusion about where to start, 659 00:37:18,130 --> 00:37:20,020 but once they got started, it was fantastic. 660 00:37:20,020 --> 00:37:23,670 I was able to look at all the different evidence that I had 661 00:37:23,670 --> 00:37:26,860 and collate that data and present something 662 00:37:26,860 --> 00:37:29,870 a little easier to digest 663 00:37:29,870 --> 00:37:35,780 for whoever was involved, and then whoever was going to get directly involved, 664 00:37:35,780 --> 00:37:37,380 get their hands dirty in fixing things, 665 00:37:37,380 --> 00:37:39,390 could look at all that granular data 666 00:37:39,390 --> 00:37:42,530 and see specifics about where things broke, 667 00:37:42,530 --> 00:37:44,440 what link, what page, 668 00:37:44,440 --> 00:37:49,160 what video, whatever. 669 00:37:49,160 --> 00:37:51,970 Generating recommendations based on your results 670 00:37:51,970 --> 00:37:55,800 is something that could be the difference 671 00:37:55,800 --> 00:38:06,030 between your efforts actually turning into something or not. 672 00:38:06,030 --> 00:38:08,740 Being able to take what you've done 673 00:38:08,740 --> 00:38:12,490 and translate it into something actionable 674 00:38:12,490 --> 00:38:16,690 really goes a huge, long way 675 00:38:16,690 --> 00:38:21,410 toward fixing your site or fixing your app. 676 00:38:21,410 --> 00:38:28,710 Too many times when different people have done tests 677 00:38:28,710 --> 00:38:31,870 they've reported the results, 678 00:38:31,870 --> 00:38:33,190 and then they've stopped there, 679 00:38:33,190 --> 00:38:37,320 because maybe they think they're not qualified to come up with the fixes, 680 00:38:37,320 --> 00:38:41,500 or they think that's all they have to generate, 681 00:38:41,500 --> 00:38:43,210 and their work is done. 682 00:38:43,210 --> 00:38:47,990 If you find things that are wrong with a site or an app, 683 00:38:47,990 --> 00:38:51,770 being able to report those is good. 684 00:38:51,770 --> 00:38:53,110 That's a good first step. 685 00:38:53,110 --> 00:38:56,740 But suggesting a fix can be even better. 686 00:38:56,740 --> 00:39:00,700 The fix may not be the right thing, and that's fine, 687 00:39:00,700 --> 00:39:05,340 and this comes up in any professional or project environment. 688 00:39:05,340 --> 00:39:08,090 Finding a problem, reporting it 689 00:39:08,090 --> 00:39:12,790 and suggesting a fix can lead to the correct fix, 690 00:39:12,790 --> 00:39:16,250 whoever is fixing the problem there. 691 00:39:16,250 --> 00:39:21,160 But stopping at the results is a bad idea. 692 00:39:21,160 --> 00:39:25,380 Make some recommendations, and don't be afraid to say, 693 00:39:25,380 --> 00:39:28,400 "I don't know how to fix this, but this is clearly a problem." 694 00:39:28,400 --> 00:39:30,990 Maybe that means you go online and you do some research 695 00:39:30,990 --> 00:39:35,080 about other sites that have similar features 696 00:39:35,080 --> 00:39:41,180 or content, and you see how they treat those things that are a problem on your site. 697 00:39:41,180 --> 00:39:45,630 Maybe you do some research and you find some services, 698 00:39:45,630 --> 00:39:48,480 some paid services that do those things for you. 699 00:39:48,480 --> 00:39:50,990 Maybe you don't have to buy those services 700 00:39:50,990 --> 00:39:52,750 and spend a lot of money. 701 00:39:52,750 --> 00:39:54,380 Maybe you're able to build something in-house 702 00:39:54,380 --> 00:39:56,500 and fix it for yourself, and that's great. 703 00:39:56,500 --> 00:39:59,700 At least it's led to some kind of solution, 704 00:39:59,700 --> 00:40:06,870 and that's what usability testing is about is taking us from our current state 705 00:40:06,870 --> 00:40:11,390 to something that is better for everyone involved, 706 00:40:11,390 --> 00:40:14,480 for the users and for yourself. 707 00:40:14,480 --> 00:40:16,360 Another thing about recommendations, 708 00:40:16,360 --> 00:40:18,540 if I can add this as well, 709 00:40:18,540 --> 00:40:21,500 is that it helps to break down recommendations 710 00:40:21,500 --> 00:40:24,040 sometimes into different categories. 711 00:40:24,040 --> 00:40:27,810 You could have the absolute needs 712 00:40:27,810 --> 00:40:30,540 of things that need to be fixed or changed 713 00:40:30,540 --> 00:40:33,700 or edited somehow, and then you could have a wish list. 714 00:40:33,700 --> 00:40:35,630 So the needs and the wants. 715 00:40:35,630 --> 00:40:37,240 You can have a wish list. 716 00:40:37,240 --> 00:40:42,570 You can say this particular part of the site works. 717 00:40:42,570 --> 00:40:48,170 It's okay, but the test user mentioned that they saw something like this 718 00:40:48,170 --> 00:40:51,050 on something something dot com, 719 00:40:51,050 --> 00:40:54,690 and if you know they have a system 720 00:40:54,690 --> 00:40:56,650 or a content management system or a service 721 00:40:56,650 --> 00:40:59,320 or something like that, maybe you can make the recommendation 722 00:40:59,320 --> 00:41:02,250 of maybe that's something we should consider, 723 00:41:02,250 --> 00:41:06,640 investing in some type of technology 724 00:41:06,640 --> 00:41:10,450 that makes our site better, that makes the experience better for the users. 725 00:41:10,450 --> 00:41:12,340 You can have needs and wants. 726 00:41:12,340 --> 00:41:14,450 You can also have short-term recommendations 727 00:41:14,450 --> 00:41:16,520 and long-term recommendations. 728 00:41:16,520 --> 00:41:18,970 This might actually be even better 729 00:41:18,970 --> 00:41:21,120 if you're going to be sharing these things 730 00:41:21,120 --> 00:41:24,580 with the people that work on the site, because you can give them the list 731 00:41:24,580 --> 00:41:27,870 and say, "Here are those easy things, easy fixes, 732 00:41:27,870 --> 00:41:31,160 "that I know you can knock out maybe in a day. 733 00:41:31,160 --> 00:41:35,310 "And then here's a list of other things that are going to take a lot more time and effort, 734 00:41:35,310 --> 00:41:39,780 maybe some more capital, some more resources, things like that." 735 00:41:39,780 --> 00:41:47,240 That can be helpful rather than a big, long checklist of oh, my goodness, 736 00:41:47,240 --> 00:41:49,000 your eyes start glazing over. 737 00:41:49,000 --> 00:41:51,590 It's a lot to do. 738 00:41:51,590 --> 00:41:53,580 Sharing results with the team. 739 00:41:53,580 --> 00:41:55,420 That's what I was just talking about. 740 00:41:55,420 --> 00:41:57,700 Don't sit on these results, 741 00:41:57,700 --> 00:42:00,310 unless you're the one doing all the work, and that's fine. 742 00:42:00,310 --> 00:42:04,960 Then you can take that and action that stuff, start fixing things. 743 00:42:04,960 --> 00:42:08,090 That's great if you're a 1-man team or whatever. 744 00:42:08,090 --> 00:42:10,660 But if you have other teams that do all the work, 745 00:42:10,660 --> 00:42:12,190 hopefully you took my advice. 746 00:42:12,190 --> 00:42:14,630 You talked to them beforehand. 747 00:42:14,630 --> 00:42:16,660 Share those results. 748 00:42:16,660 --> 00:42:18,010 You've already created the expectation. 749 00:42:18,010 --> 00:42:20,770 You're saying, "I'm doing testing. I'm going to let you know how it goes," 750 00:42:20,770 --> 00:42:22,040 or "You can be involved. 751 00:42:22,040 --> 00:42:24,010 You can sit in," whatever. 752 00:42:24,010 --> 00:42:26,640 But share the results with the teams that would be interested in it, 753 00:42:26,640 --> 00:42:32,670 and then the hope is they can start to see the value in this testing, 754 00:42:32,670 --> 00:42:36,460 because one of the things that can be a struggle, 755 00:42:36,460 --> 00:42:40,320 especially if you're working in a company environment, 756 00:42:40,320 --> 00:42:44,590 is just getting the sign off, getting the approval 757 00:42:44,590 --> 00:42:48,150 from your boss or your manager or your team or whoever 758 00:42:48,150 --> 00:42:50,290 to do this testing in the first place, 759 00:42:50,290 --> 00:42:52,760 because there's an expense involved. 760 00:42:52,760 --> 00:42:55,610 There's time, especially compensation 761 00:42:55,610 --> 00:42:57,530 if you're going to be compensating the test users. 762 00:42:57,530 --> 00:43:01,850 You do have to sell this. 763 00:43:01,850 --> 00:43:05,480 But there's no better way to sell this 764 00:43:05,480 --> 00:43:09,990 than showing them the data from a successful test. 765 00:43:09,990 --> 00:43:13,240 I know that's kind of backwards. 766 00:43:13,240 --> 00:43:15,800 You have to do the test to get the data 767 00:43:15,800 --> 00:43:17,610 to show that it's valuable. 768 00:43:17,610 --> 00:43:19,990 You can do something under the radar. 769 00:43:19,990 --> 00:43:22,460 You can do something cheap, for free, 770 00:43:22,460 --> 00:43:24,850 a small scale example of testing 771 00:43:24,850 --> 00:43:28,170 to show that results can be found. 772 00:43:28,170 --> 00:43:32,450 If you had a little more money or you had access to other things, 773 00:43:32,450 --> 00:43:36,300 other tools, other resources, you can get even more information, 774 00:43:36,300 --> 00:43:42,000 and it could mean even more success for your site or app. 775 00:43:42,000 --> 00:43:45,050 Lastly, don't lose steam. 776 00:43:45,050 --> 00:43:48,060 If you have a successful test, you have people come in 777 00:43:48,060 --> 00:43:52,830 that give you great insight, schedule that next set of tests. 778 00:43:52,830 --> 00:43:55,220 Get it done. 779 00:43:55,220 --> 00:43:58,570 If you test once a month 780 00:43:58,570 --> 00:44:00,510 or even once every other month, 781 00:44:00,510 --> 00:44:08,040 that can lead to a really great process internally 782 00:44:08,040 --> 00:44:11,290 during design and development or even after the launch 783 00:44:11,290 --> 00:44:18,990 leading up to further redesigns or rollouts of other features, things like that. 784 00:44:18,990 --> 00:44:22,080 It can be such a great process 785 00:44:22,080 --> 00:44:25,630 to have the regular expectation 786 00:44:25,630 --> 00:44:29,150 of testing happen, because other people will get involved, 787 00:44:29,150 --> 00:44:31,700 other people can participate. 788 00:44:31,700 --> 00:44:33,860 Maybe other people can administer it. 789 00:44:33,860 --> 00:44:40,150 But any channels that you build 790 00:44:40,150 --> 00:44:43,980 with, say, groups of people who might be interested 791 00:44:43,980 --> 00:44:48,570 in being test users, you can have a whole group of people online 792 00:44:48,570 --> 00:44:52,070 or locally in the area to help you out. 793 00:44:52,070 --> 00:44:56,640 You can even reach out to some of those test users again 794 00:44:56,640 --> 00:44:58,770 if they're especially helpful and bring them in 795 00:44:58,770 --> 00:45:01,850 and ask them questions again. 796 00:45:01,850 --> 00:45:05,210 But don't lose steam. 797 00:45:05,210 --> 00:45:07,350 Schedule another set of tests, 798 00:45:07,350 --> 00:45:13,270 and when you do tests a second time, a third time and so on, 799 00:45:13,270 --> 00:45:17,290 you don't have to test the exact same thing every single time. 800 00:45:17,290 --> 00:45:19,340 In fact, I wouldn't recommend that. 801 00:45:19,340 --> 00:45:23,540 You might want to test the most important task again, 802 00:45:23,540 --> 00:45:25,780 because that's always great to get insight on. 803 00:45:25,780 --> 00:45:29,770 But there may be other parts of the site that could use some attention. 804 00:45:29,770 --> 00:45:31,600 There may be some new features, 805 00:45:31,600 --> 00:45:33,810 and there may be other things that maybe the design team 806 00:45:33,810 --> 00:45:38,020 or development comes to you and says, "We'd really like to get this in front of some other people 807 00:45:38,020 --> 00:45:43,930 and get their perspective on it," and so being able to get their insight on it 808 00:45:43,930 --> 00:45:45,720 would be fantastic. 809 00:45:48,960 --> 00:45:50,230 That's testing. 810 00:45:50,230 --> 00:45:52,290 If you want to take testing to the next level, so to speak, 811 00:45:52,290 --> 00:45:57,540 here's something that if people haven't thought about it, it definitely blows their minds. 812 00:45:57,540 --> 00:46:00,740 Test competitor sites or apps. 813 00:46:00,740 --> 00:46:03,830 It's okay to test other websites. 814 00:46:03,830 --> 00:46:06,200 You can test your own site, 815 00:46:06,200 --> 00:46:08,650 and that's great, and you can improve your site and everything. 816 00:46:08,650 --> 00:46:11,500 But go out and look at your competitors, 817 00:46:11,500 --> 00:46:15,650 see what they're doing, and run a usability test on them. 818 00:46:15,650 --> 00:46:18,560 Get some other people's insight on them, 819 00:46:18,560 --> 00:46:22,070 because you may be too close to the situation. 820 00:46:22,070 --> 00:46:24,190 Your site does this, 821 00:46:24,190 --> 00:46:26,100 and their site does almost the same thing. 822 00:46:26,100 --> 00:46:31,430 You already know a lot about how certain things work. 823 00:46:31,430 --> 00:46:33,570 Get another person's perspective on it. 824 00:46:33,570 --> 00:46:35,310 Get an outsider's perspective 825 00:46:35,310 --> 00:46:37,120 and see what they think of it, 826 00:46:37,120 --> 00:46:40,300 because again, just like your own site, 827 00:46:40,300 --> 00:46:42,290 they may see things that you don't see 828 00:46:42,290 --> 00:46:49,100 because their experience, what they bring from past experience using other sites, 829 00:46:49,100 --> 00:46:52,340 sites that you may have never used before, 830 00:46:52,340 --> 00:46:55,780 really the experience of trying to fulfill their own needs 831 00:46:55,780 --> 00:46:58,880 as an outsider is going to be different than yours. 832 00:46:58,880 --> 00:47:02,450 Test competitor sites. 833 00:47:02,450 --> 00:47:04,050 It's okay. 834 00:47:04,050 --> 00:47:06,140 You're not breaking any laws. 835 00:47:06,140 --> 00:47:08,890 You're trying to get more information 836 00:47:08,890 --> 00:47:13,000 and gain insight so you can build an even better site than them, 837 00:47:13,000 --> 00:47:16,270 because you can guarantee that they will look at your site 838 00:47:16,270 --> 00:47:22,910 and maybe test yours, and they will try to outdo you as well. 839 00:47:22,910 --> 00:47:26,000 Some other things that you can do, and these are actually a couple things. 840 00:47:26,000 --> 00:47:28,230 You can add some automated tools to your site, 841 00:47:28,230 --> 00:47:32,240 or you can use some paid services, and sometimes they're both. 842 00:47:32,240 --> 00:47:38,070 Automated tools, the things like Google Analytics that are going to give you analytics, 843 00:47:38,070 --> 00:47:41,760 but there are other tools you can add to your site 844 00:47:41,760 --> 00:47:44,290 that can track user sessions. 845 00:47:44,290 --> 00:47:47,450 They can give you a little bit more data 846 00:47:47,450 --> 00:47:51,510 about what goes on in general 847 00:47:51,510 --> 00:47:53,750 but also for specific user sessions. 848 00:47:53,750 --> 00:47:56,040 There are apps and services out there 849 00:47:56,040 --> 00:48:00,220 that can do that, and there are paid services that allow you to set up 850 00:48:00,220 --> 00:48:04,490 usability tests that can give you even more detail, 851 00:48:04,490 --> 00:48:06,380 and it can give you actually quick access 852 00:48:06,380 --> 00:48:12,280 to data quicker than you setting up a whole session 853 00:48:12,280 --> 00:48:15,850 and spending a whole day and spending a week generating results, things like that, 854 00:48:15,850 --> 00:48:17,750 and that's fine. 855 00:48:17,750 --> 00:48:19,820 And paid services are great, 856 00:48:19,820 --> 00:48:22,980 especially ones like UserTesting, 857 00:48:22,980 --> 00:48:27,160 IntuitionHQ, Verify, which is the service from Zurb. 858 00:48:27,160 --> 00:48:29,250 And these are just a few of them. 859 00:48:29,250 --> 00:48:32,590 These are the ones that I know of, and I know user testing. 860 00:48:32,590 --> 00:48:37,110 I've spoken with the people from IntuitionHQ before. 861 00:48:37,110 --> 00:48:41,790 They all offer different ways that you can test your site. 862 00:48:41,790 --> 00:48:47,050 And they really vary, so it's important to check those out, 863 00:48:47,050 --> 00:48:50,620 and if you want more information on paid services, 864 00:48:50,620 --> 00:48:54,120 there's a great article from Smashing Magazine 865 00:48:54,120 --> 00:48:56,150 that I actually shared on my Twitter feed, 866 00:48:56,150 --> 00:48:58,640 @dangorgone on Twitter. 867 00:48:58,640 --> 00:49:03,140 I just shared it before we went on the live stream here, 868 00:49:03,140 --> 00:49:05,370 so there are a couple of articles there. 869 00:49:05,370 --> 00:49:07,650 One about a lot of these different services, 870 00:49:07,650 --> 00:49:10,700 a lot of comparison of what they offer, 871 00:49:10,700 --> 00:49:13,690 so if you're looking for something specific, 872 00:49:13,690 --> 00:49:16,570 definitely check that out, 873 00:49:16,570 --> 00:49:21,030 and it's a couple years old, so people have been updating the comments 874 00:49:21,030 --> 00:49:24,240 with newer stuff as well, so check that out as well. 875 00:49:24,240 --> 00:49:28,230 There's another article about the benefits of usability testing, 876 00:49:28,230 --> 00:49:31,930 and I believe it's an app called Hello Sign 877 00:49:31,930 --> 00:49:35,480 that they recently did some usability testing, 878 00:49:35,480 --> 00:49:42,000 and by doing the testing they were able to change the log in screen, 879 00:49:42,000 --> 00:49:45,290 the welcome screen of the app, and they made a huge difference 880 00:49:45,290 --> 00:49:48,220 in the usability of the app. 881 00:49:48,220 --> 00:49:52,670 You would think the most important stuff is inside the app, right? 882 00:49:52,670 --> 00:49:55,450 But if you have something that's behind the log in wall, 883 00:49:55,450 --> 00:49:59,310 that is something you have to keep in mind when you're testing 884 00:49:59,310 --> 00:50:02,020 is that people visiting your site for the first time 885 00:50:02,020 --> 00:50:04,070 are not already going to be logged in. 886 00:50:04,070 --> 00:50:07,440 They're going to come to the logged out version of your site, 887 00:50:07,440 --> 00:50:09,800 and they're going to see what that looks like 888 00:50:09,800 --> 00:50:14,930 and try to deal with that and make their own idea 889 00:50:14,930 --> 00:50:17,330 about whether they want to enter the site 890 00:50:17,330 --> 00:50:19,720 and do some more stuff. 891 00:50:22,750 --> 00:50:24,990 I'm going to wrap it up there. 892 00:50:24,990 --> 00:50:28,000 But I know we've got some questions 893 00:50:28,000 --> 00:50:34,200 that we're going to look at, but I do want to add that getting another person's perspective 894 00:50:34,200 --> 00:50:40,600 by doing testing is something that is incredibly valuable. 895 00:50:40,600 --> 00:50:42,790 If you haven't gotten that impression from me yet, 896 00:50:42,790 --> 00:50:44,250 I'm going to say it again. 897 00:50:44,250 --> 00:50:46,980 Asking other people to check out the work that you've done 898 00:50:46,980 --> 00:50:51,230 is always an eye opener. 899 00:50:51,230 --> 00:50:56,530 And whether you are validating the work that you do through testing, 900 00:50:56,530 --> 00:51:00,910 like you're getting positive reinforcement and it's working exactly as you thought, 901 00:51:00,910 --> 00:51:05,660 or you are discovering problems and mistakes 902 00:51:05,660 --> 00:51:07,350 and things that you can fix, 903 00:51:07,350 --> 00:51:10,210 either way it is not a waste of time. 904 00:51:10,210 --> 00:51:12,330 It's not a waste of resources at all 905 00:51:12,330 --> 00:51:16,310 because when your manager or your CEO 906 00:51:16,310 --> 00:51:19,450 walks in and says, "How's the site doing?" 907 00:51:19,450 --> 00:51:21,650 you know exactly what to say, 908 00:51:21,650 --> 00:51:24,350 although you'd probably say, "Hey, it's doing fine" either way. 909 00:51:24,350 --> 00:51:28,290 But if there are problems on the site, you know exactly what to fix. 910 00:51:28,290 --> 00:51:32,110 If everything is working great, you can prove it with data. 911 00:51:35,800 --> 00:51:40,450 I know a couple of the questions are about third-party software for doing testing. 912 00:51:40,450 --> 00:51:43,360 I know I've recommended a couple of those things, but definitely check out 913 00:51:43,360 --> 00:51:47,260 the Smashing Magazine article that I shared on Twitter. 914 00:51:47,260 --> 00:51:49,770 Again, that's @dangorgone on Twitter. 915 00:51:49,770 --> 00:51:51,370 You'll see the last couple things I posted. 916 00:51:51,370 --> 00:51:55,910 Check out those articles, because those can be helpful. 917 00:51:55,910 --> 00:51:58,470 Let's see. 918 00:51:58,470 --> 00:52:01,010 If you're doing prototyping on paper, 919 00:52:01,010 --> 00:52:03,850 what's the best way to test interaction? 920 00:52:03,850 --> 00:52:06,220 This is a question from Fernando. 921 00:52:06,220 --> 00:52:10,560 You can absolutely test when you have paper prototypes 922 00:52:10,560 --> 00:52:12,660 or even something on a whiteboard, 923 00:52:12,660 --> 00:52:15,840 but paper is even better, 924 00:52:15,840 --> 00:52:19,600 whether it's something you mock up using a tool like Mockingbird 925 00:52:19,600 --> 00:52:22,060 or whether you're actually drawing it out. 926 00:52:22,060 --> 00:52:25,130 What I would try to do with those 927 00:52:25,130 --> 00:52:30,420 is anticipate what you need to test 928 00:52:30,420 --> 00:52:34,570 as you're going through the whole step-by-step process. 929 00:52:34,570 --> 00:52:38,910 Try not to leave it at 1 screen and ask a question. 930 00:52:38,910 --> 00:52:44,960 If it's possible, give people more of a real-life experience 931 00:52:44,960 --> 00:52:49,930 by having the other pages or other designs 932 00:52:49,930 --> 00:52:51,550 there ready to go. 933 00:52:51,550 --> 00:52:56,870 Whether it's a map that you have or whether you've got those things in a folder 934 00:52:56,870 --> 00:53:00,370 and you say, "What would you click?" and then they say, "I would click on that," 935 00:53:00,370 --> 00:53:03,820 and then you pull out another piece of paper and say, "Good," or you have it in a notebook 936 00:53:03,820 --> 00:53:05,460 or something like that. 937 00:53:05,460 --> 00:53:08,350 You can do it. You can get insight that way. 938 00:53:08,350 --> 00:53:14,890 Certainly that's something that happens obviously before a launch. 939 00:53:14,890 --> 00:53:17,530 But that can be a helpful way 940 00:53:17,530 --> 00:53:24,990 I've actually found to get some insight about weighing designs against each other. 941 00:53:24,990 --> 00:53:29,490 You can have 2 versions of a design 942 00:53:29,490 --> 00:53:32,310 on a piece of paper printed out, and you put them side by side. 943 00:53:32,310 --> 00:53:35,340 And you can say, "Check this out, look at this." 944 00:53:35,340 --> 00:53:39,570 Is there a navigation menu style 945 00:53:39,570 --> 00:53:41,690 that you think is better? 946 00:53:41,690 --> 00:53:44,510 And then when they point to one, ask them why 947 00:53:44,510 --> 00:53:49,420 and get that information from them. 948 00:53:49,420 --> 00:53:53,520 Mohammed asked the best way to collect feedback from a large number of users. 949 00:53:53,520 --> 00:53:55,800 What is the best way? 950 00:53:55,800 --> 00:54:00,190 I would say one of those add-on tools 951 00:54:00,190 --> 00:54:05,360 or one of the paid services that I mentioned. 952 00:54:05,360 --> 00:54:10,810 They will allow you to get feedback from many more users 953 00:54:10,810 --> 00:54:12,880 than you would in person. 954 00:54:12,880 --> 00:54:16,760 Now, understand that those paid services are paid, 955 00:54:16,760 --> 00:54:18,180 so they are going to cost money. 956 00:54:18,180 --> 00:54:23,320 Some of them have steps to the cost, 957 00:54:23,320 --> 00:54:26,900 so it could be free up to a certain amount of users 958 00:54:26,900 --> 00:54:29,800 or a certain amount of tests or page views or something. 959 00:54:29,800 --> 00:54:32,030 You have to look at each one and see 960 00:54:32,030 --> 00:54:34,760 what you will actually pay for 961 00:54:34,760 --> 00:54:42,150 and how much feedback you can actually get from each one. 962 00:54:42,150 --> 00:54:46,040 Mouth asks are there any benefits to testing on a live server 963 00:54:46,040 --> 00:54:49,880 versus a local setup? 964 00:54:49,880 --> 00:54:57,950 Well, the live server will give you the true, real-life experience of testing. 965 00:54:57,950 --> 00:55:01,700 You'll know that if something is on the live site 966 00:55:01,700 --> 00:55:04,760 that it's guaranteed to be there, 967 00:55:04,760 --> 00:55:07,470 that it's guaranteed that if a link doesn't work, 968 00:55:07,470 --> 00:55:09,140 all right, that link definitely doesn't work. 969 00:55:09,140 --> 00:55:11,000 We have to fix that. 970 00:55:11,000 --> 00:55:15,400 If it's a local setup, the thing that I would do— 971 00:55:15,400 --> 00:55:17,740 and I would recommend this for any of the tests— 972 00:55:17,740 --> 00:55:21,420 is I would go through every test you're going to do personally 973 00:55:21,420 --> 00:55:27,320 to see if the pages that you're going to, the pages you would expect the user to go to 974 00:55:27,320 --> 00:55:30,690 are there, are working, all the images work. 975 00:55:30,690 --> 00:55:34,180 You don't have to test every single link and everything, 976 00:55:34,180 --> 00:55:40,490 but generally eyeball it and say I think they're going to go to the products page, 977 00:55:40,490 --> 00:55:42,730 and you click to the products page, and you see that. 978 00:55:42,730 --> 00:55:44,720 And I think they're going to go to this next. 979 00:55:44,720 --> 00:55:48,760 Try to make sure those things are all working in advance, 980 00:55:48,760 --> 00:55:52,390 because a local setup could be a great backup plan for sure. 981 00:55:52,390 --> 00:55:57,310 And if that happens that all of a sudden your wi-fi goes down 982 00:55:57,310 --> 00:55:59,950 or the site reboots for some reason 983 00:55:59,950 --> 00:56:03,410 and you have to rely on a local setup, that's fine. 984 00:56:03,410 --> 00:56:05,550 Be honest with the test user as well 985 00:56:05,550 --> 00:56:07,450 and say, "Oh, looks like our Internet is down. 986 00:56:07,450 --> 00:56:09,430 I'm going to use this local setup." 987 00:56:09,430 --> 00:56:12,450 But it shouldn't be any different from what you expect, 988 00:56:12,450 --> 00:56:17,790 and then watch them and observe them and see if there really is any difference, 989 00:56:17,790 --> 00:56:20,180 because there shouldn't be if it's a local setup. 990 00:56:23,470 --> 00:56:26,610 Shakara asks do you have any templates 991 00:56:26,610 --> 00:56:29,540 you recommend for sharing results of user testing 992 00:56:29,540 --> 00:56:32,110 with the rest of your team? 993 00:56:32,110 --> 00:56:35,980 There are a number of templates out online 994 00:56:35,980 --> 00:56:37,750 from different places that you can find. 995 00:56:37,750 --> 00:56:42,490 But really for me it varies. 996 00:56:42,490 --> 00:56:45,520 I think you have to do testing at least 1 time 997 00:56:45,520 --> 00:56:48,580 with whatever site or app you have 998 00:56:48,580 --> 00:56:50,560 to see what you're going to get 999 00:56:50,560 --> 00:56:55,770 and to see how that information is really going to shake out. 1000 00:56:55,770 --> 00:57:01,570 If you have 3 different tasks that you're going to administer 1001 00:57:01,570 --> 00:57:06,250 to 3 different people, than you know you're going to get basically 9 sets of results, 1002 00:57:06,250 --> 00:57:09,100 3 tasks, 3 people, 3 times 3 is 9. 1003 00:57:09,100 --> 00:57:13,760 Whether you want to set things up in a grid, you could do that. 1004 00:57:13,760 --> 00:57:15,250 You could set it up bullet style. 1005 00:57:15,250 --> 00:57:20,080 You can do what I did where I had totally granular notes, 1006 00:57:20,080 --> 00:57:23,970 like everything they did and everything they said all in a Google doc. 1007 00:57:23,970 --> 00:57:26,260 But then at the top of that Google doc 1008 00:57:26,260 --> 00:57:31,470 was the summary, and I recommend whatever template you have 1009 00:57:31,470 --> 00:57:35,760 or are going to use, however you share these results, 1010 00:57:35,760 --> 00:57:39,630 at least have an executive type summary at the top, 1011 00:57:39,630 --> 00:57:44,330 because you want to be able to give people a brief, at a glance idea 1012 00:57:44,330 --> 00:57:46,730 of what happened in the tests, 1013 00:57:46,730 --> 00:57:51,550 how successful were people, is there some general sentiment 1014 00:57:51,550 --> 00:57:55,780 that you're getting from things, and what are the things that need to be addressed? 1015 00:57:55,780 --> 00:57:59,730 That's the other thing as well. 1016 00:57:59,730 --> 00:58:02,800 Mohammed asks once we have the results of our tests, 1017 00:58:02,800 --> 00:58:06,090 how do we prioritize which changes to make 1018 00:58:06,090 --> 00:58:08,080 based on the suggestions of our users? 1019 00:58:08,080 --> 00:58:10,260 This is a great question 1020 00:58:10,260 --> 00:58:14,530 and something that you have to gauge. 1021 00:58:14,530 --> 00:58:18,070 You want to get all the information you can from them 1022 00:58:18,070 --> 00:58:21,480 and all the feedback, and as you're going through the tests, 1023 00:58:21,480 --> 00:58:26,080 some of the test users might get a little self-confidence, 1024 00:58:26,080 --> 00:58:28,930 and they may start suggesting things that you should change. 1025 00:58:28,930 --> 00:58:31,780 It's not really based on anything in particular that's happening. 1026 00:58:31,780 --> 00:58:35,750 Like, "Oh, I see your logo is blue. 1027 00:58:35,750 --> 00:58:38,410 If it was green, I think it would be cool." 1028 00:58:38,410 --> 00:58:41,370 That's not really relevant here. 1029 00:58:41,370 --> 00:58:43,080 It doesn't help. 1030 00:58:43,080 --> 00:58:45,130 Make a note of it. You never know. 1031 00:58:45,130 --> 00:58:47,160 Something could come up where the design team sees that, 1032 00:58:47,160 --> 00:58:49,420 and they're like, "Oh, yeah. 1033 00:58:49,420 --> 00:58:51,180 We've been trying to change it to green forever." 1034 00:58:51,180 --> 00:58:53,690 You never know where some of this information will come up. 1035 00:58:53,690 --> 00:58:56,680 But keep all that stuff definitely in a granular version 1036 00:58:56,680 --> 00:59:04,090 of changes, but to prioritize, to get back to your question, 1037 00:59:04,090 --> 00:59:07,570 you want to look at those primary tasks that you tested first. 1038 00:59:07,570 --> 00:59:10,550 You have to know if those things are working or not, 1039 00:59:10,550 --> 00:59:14,800 and I'll give you a classic example. 1040 00:59:14,800 --> 00:59:17,360 When I used to teach at Full Sail University 1041 00:59:17,360 --> 00:59:20,420 and I taught usability in testing 1042 00:59:20,420 --> 00:59:25,410 in my very first month where I had my students do testing on their sites 1043 00:59:25,410 --> 00:59:29,160 because they had their own sites and businesses and things, 1044 00:59:29,160 --> 00:59:32,630 this guy was part of a design agency, 1045 00:59:32,630 --> 00:59:37,800 and his most recent client was a sporting good apparel company, 1046 00:59:37,800 --> 00:59:40,380 and so they had their site all set up and everything. 1047 00:59:40,380 --> 00:59:45,910 He goes through and of course identifies the primary task we're going to test here 1048 00:59:45,910 --> 00:59:52,240 is can people go through and find, say, a set of sneakers, 1049 00:59:52,240 --> 00:59:56,780 some running shoes that they want, and can they go through the whole shopping cart process and buy it? 1050 00:59:56,780 --> 01:00:02,400 He even came up with a dummy credit card or something 1051 01:00:02,400 --> 01:00:04,210 that the system would accept. 1052 01:00:04,210 --> 01:00:06,200 That's not something everyone has the ability to do, 1053 01:00:06,200 --> 01:00:08,850 but he was able to do that, so he gave them the information, 1054 01:00:08,850 --> 01:00:11,570 so they were able to go through the whole process. 1055 01:00:11,570 --> 01:00:14,460 Well, as luck would have it, 1056 01:00:14,460 --> 01:00:18,180 he goes through, and the shopping cart was broken. 1057 01:00:18,180 --> 01:00:22,300 The shopping cart would not work at all. 1058 01:00:22,300 --> 01:00:24,490 You get through, you enter the information, 1059 01:00:24,490 --> 01:00:28,820 you hit submit, and then some kind of error came up. 1060 01:00:28,820 --> 01:00:31,240 And he was absolutely shocked. 1061 01:00:31,240 --> 01:00:33,120 He couldn't believe it. 1062 01:00:33,120 --> 01:00:37,110 Now, it was some minor thing, of course, 1063 01:00:37,110 --> 01:00:39,030 but when you're talking priorities, 1064 01:00:39,030 --> 01:00:41,680 this was the first thing that he identified, 1065 01:00:41,680 --> 01:00:45,930 the top thing that's important to the success of this site, 1066 01:00:45,930 --> 01:00:48,490 the success of the business, tested it first, 1067 01:00:48,490 --> 01:00:51,740 tested it for each user, found out that it failed. 1068 01:00:51,740 --> 01:00:53,770 That became priority #1. 1069 01:00:53,770 --> 01:00:56,060 Not just in the report that he submitted to me 1070 01:00:56,060 --> 01:00:58,440 but when he submitted that Sunday night, 1071 01:00:58,440 --> 01:01:00,900 the first thing he did the very next morning, Monday morning, 1072 01:01:00,900 --> 01:01:02,870 was he got together with his development team, 1073 01:01:02,870 --> 01:01:04,080 and he fixed that problem. 1074 01:01:04,080 --> 01:01:06,100 You better believe it that he did. 1075 01:01:06,100 --> 01:01:08,550 Some of these things are very obvious, 1076 01:01:08,550 --> 01:01:12,880 and they really mean the difference between success and failure for a site. 1077 01:01:12,880 --> 01:01:17,050 But the other way to prioritize is think about is it a short-term thing, 1078 01:01:17,050 --> 01:01:19,020 is it a long-term thing? 1079 01:01:19,020 --> 01:01:23,270 Figuring out whether it's needs or wants, 1080 01:01:23,270 --> 01:01:25,670 that can be a little more fuzzy. 1081 01:01:25,670 --> 01:01:28,540 But if you are ever in question 1082 01:01:28,540 --> 01:01:31,440 about whether something works or not 1083 01:01:31,440 --> 01:01:33,580 or whether something is a priority or not, 1084 01:01:33,580 --> 01:01:35,240 test it again. 1085 01:01:35,240 --> 01:01:38,130 Test it again and get another person's perspective on it, 1086 01:01:38,130 --> 01:01:42,420 and you may start to see a pattern emerge. 1087 01:01:42,420 --> 01:01:45,380 Sometimes 3 to 5 users is not enough 1088 01:01:45,380 --> 01:01:50,690 to determine whether something needs to go a certain way or not. 1089 01:01:50,690 --> 01:01:54,030 The design or the feature or the content, whatever it is. 1090 01:01:54,030 --> 01:01:56,290 Sometimes it does take more testing, 1091 01:01:56,290 --> 01:01:59,540 and that is just the way it is with testing. 1092 01:01:59,540 --> 01:02:04,600 It's something that you can AB test 1093 01:02:04,600 --> 01:02:07,570 if things are too close to one another. 1094 01:02:07,570 --> 01:02:11,350 If you have the possibility of AB testing 1095 01:02:11,350 --> 01:02:13,290 in design or a feature or something like that, 1096 01:02:13,290 --> 01:02:17,380 that can be another way, and whether you can do it yourself 1097 01:02:17,380 --> 01:02:21,760 or through one of the paid services, that can be a way to get some resolution 1098 01:02:21,760 --> 01:02:23,940 on that question. 1099 01:02:23,940 --> 01:02:28,340 And the last question here, and again, I thank everyone for sticking with me here. 1100 01:02:28,340 --> 01:02:32,330 Any plans for more UX content on the Treehouse site? 1101 01:02:32,330 --> 01:02:34,040 Absolutely, absolutely. 1102 01:02:34,040 --> 01:02:38,690 UX and UI I believe are a couple of the subjects 1103 01:02:38,690 --> 01:02:40,660 that I know we started at one point, 1104 01:02:40,660 --> 01:02:43,890 and we switched some gears, and like I said, 1105 01:02:43,890 --> 01:02:48,020 I've got Usability Foundations that I just put out, 1106 01:02:48,020 --> 01:02:50,940 and I've heard from a whole bunch of students about it, 1107 01:02:50,940 --> 01:02:54,620 and I'm very happy that they dig it. 1108 01:02:54,620 --> 01:02:57,340 But UX and UI will definitely have that, 1109 01:02:57,340 --> 01:03:05,000 and one of the spots where we regularly announce new content for teamtreehouse.com 1110 01:03:05,000 --> 01:03:09,800 is on our Roadmap, so if you go to teamtreehouse.com/roadmap, 1111 01:03:09,800 --> 01:03:12,480 you can see exactly what's coming up, 1112 01:03:12,480 --> 01:03:18,430 and if there are any subjects that you're looking for, 1113 01:03:18,430 --> 01:03:20,810 definitely drop us a line. 1114 01:03:20,810 --> 01:03:23,770 I want to thank you for watching this workshop on usability testing. 1115 01:03:23,770 --> 01:03:26,840 If you have any questions, anything further, 1116 01:03:26,840 --> 01:03:29,120 be sure to find me in the Treehouse forum, 1117 01:03:29,120 --> 01:03:31,950 or you can find me on Twitter, @dangorgone. 1118 01:03:31,950 --> 01:03:33,000 Thanks for watching.