WEBVTT 00:00:01.400 --> 00:00:05.580 Good morning. Almost noon. 00:00:05.580 --> 00:00:08.190 It's exciting to be here today. 00:00:08.190 --> 00:00:10.440 I want to hopefully convince you all to 00:00:10.440 --> 00:00:12.750 quit what you're doing and start doing service design. 00:00:12.750 --> 00:00:16.410 [LAUGHTER] This is some work 00:00:16.410 --> 00:00:19.110 I've done with my colleague Jordi Forlizzi. 00:00:19.110 --> 00:00:23.820 She and I have given this talk in different versions, 00:00:23.820 --> 00:00:26.265 a couple of times and met with 00:00:26.265 --> 00:00:29.160 much more hostile responses than we expected. 00:00:29.160 --> 00:00:32.190 We keep trying to iterate, see and so on. 00:00:32.190 --> 00:00:34.730 I'm curious about your reaction 00:00:34.730 --> 00:00:36.290 and I want to encourage you, 00:00:36.290 --> 00:00:38.510 if I say something you disagree with, 00:00:38.510 --> 00:00:40.325 please shout out because 00:00:40.325 --> 00:00:42.740 that's a big part of what I'm trying to understand, 00:00:42.740 --> 00:00:45.715 is where is what we're saying upsetting people. 00:00:45.715 --> 00:00:48.800 Because that either means we're making 00:00:48.800 --> 00:00:50.255 progress or we're not 00:00:50.255 --> 00:00:53.185 articulating our argument very well. 00:00:53.185 --> 00:00:59.270 I want to just start with a claim that HCI today, 00:00:59.270 --> 00:01:01.385 the people doing the work to design 00:01:01.385 --> 00:01:04.240 interfaces for people to work with computers, 00:01:04.240 --> 00:01:07.820 is fundamentally based on an identity that comes from 00:01:07.820 --> 00:01:11.660 user-centered design and user experience design. 00:01:11.660 --> 00:01:15.415 Hopefully, that's not too radical an idea. 00:01:15.415 --> 00:01:19.130 HCI has a long history of evolving its methods and 00:01:19.130 --> 00:01:23.870 practices to fit the kinds of things that it makes. 00:01:23.870 --> 00:01:26.210 As we make new kinds of things 00:01:26.210 --> 00:01:29.690 in interaction, there's a dialogue. 00:01:29.690 --> 00:01:30.950 It's a dialogue between 00:01:30.950 --> 00:01:33.845 the practice community and the research community. 00:01:33.845 --> 00:01:39.275 From that emerges the identity of HCI as a field. 00:01:39.275 --> 00:01:44.190 The problem as I see it is that, 00:01:44.190 --> 00:01:49.145 today we're making things that fundamentally conflict 00:01:49.145 --> 00:01:51.410 with what we mean by 00:01:51.410 --> 00:01:55.690 user-centered design and user experience design, 00:01:55.690 --> 00:01:57.710 and that we have to change 00:01:57.710 --> 00:02:00.830 our identity or get rid of these things. 00:02:00.830 --> 00:02:01.850 But these things we're making 00:02:01.850 --> 00:02:03.890 are actually pretty exciting. 00:02:03.890 --> 00:02:06.950 Of the three, the first is 00:02:06.950 --> 00:02:11.355 that today we're making services, 00:02:11.355 --> 00:02:13.750 we're not making products, 00:02:13.750 --> 00:02:16.730 but user experience design and 00:02:16.730 --> 00:02:21.680 user-centered design follow a product-centric process. 00:02:21.680 --> 00:02:24.845 Social computing systems which are 00:02:24.845 --> 00:02:29.585 super exciting and huge growth area in HCI, 00:02:29.585 --> 00:02:33.200 are really about trying to shape people's behavior, 00:02:33.200 --> 00:02:34.700 not meet their needs, 00:02:34.700 --> 00:02:35.780 violate some of 00:02:35.780 --> 00:02:38.900 the core principles of user-centered design. 00:02:38.900 --> 00:02:43.430 There's some exciting new area around thinking of 00:02:43.430 --> 00:02:45.590 computers and computing as 00:02:45.590 --> 00:02:50.070 a potential source for social innovation. 00:02:50.070 --> 00:02:52.490 There we're really actually trying to radically 00:02:52.490 --> 00:02:54.920 change people's behavior by putting 00:02:54.920 --> 00:02:57.290 systems in the world which really has nothing to do with 00:02:57.290 --> 00:03:00.120 meeting their needs and desires, 00:03:00.120 --> 00:03:03.050 and that if we really want to engage in these things, 00:03:03.050 --> 00:03:06.245 we have to think about the identity we move forward with. 00:03:06.245 --> 00:03:07.880 My colleague, Jordi and I, 00:03:07.880 --> 00:03:10.310 really feel we need to add 00:03:10.310 --> 00:03:15.920 a service design lens and expand the definition of HCI. 00:03:15.920 --> 00:03:19.730 By doing that, these three areas of work 00:03:19.730 --> 00:03:22.250 become much more central and fit with 00:03:22.250 --> 00:03:25.155 how we might define HCI. 00:03:25.155 --> 00:03:29.840 I just want to talk like how did we get to where we are? 00:03:29.840 --> 00:03:32.920 How did HCI's identity get here? 00:03:32.920 --> 00:03:34.970 These are what I see right now is 00:03:34.970 --> 00:03:36.635 the three critical lenses 00:03:36.635 --> 00:03:38.960 to which we want to add service design, 00:03:38.960 --> 00:03:40.250 usability, 00:03:40.250 --> 00:03:44.045 user-centered design, and user experience design. 00:03:44.045 --> 00:03:48.440 If you step backwards in time to the '70s, 00:03:48.440 --> 00:03:51.940 there were no interaction designers. 00:03:51.940 --> 00:03:56.045 What we think of an interface was fundamentally done by 00:03:56.045 --> 00:03:57.710 industrial designers and they took 00:03:57.710 --> 00:04:00.394 their sense of form and aesthetic, 00:04:00.394 --> 00:04:03.930 and applied it to sets of controls. 00:04:05.330 --> 00:04:10.425 They did a good job and they made many beautiful things. 00:04:10.425 --> 00:04:11.720 They weren't the only ones. 00:04:11.720 --> 00:04:13.340 We had a lot of engineers doing this as 00:04:13.340 --> 00:04:15.260 well and they were really 00:04:15.260 --> 00:04:20.750 focused on giving controls for functional devices, 00:04:20.750 --> 00:04:23.875 but there were just a lot of controls at the buttons. 00:04:23.875 --> 00:04:25.760 I don't know if any of you recognize 00:04:25.760 --> 00:04:30.320 the UVW deck from Sony for high-end video editing, 00:04:30.320 --> 00:04:31.880 I learned to edit using one of 00:04:31.880 --> 00:04:34.805 these controls for a nuclear power plant. 00:04:34.805 --> 00:04:37.205 Things don't look like this anymore, 00:04:37.205 --> 00:04:39.200 which is an improvement. 00:04:39.200 --> 00:04:41.090 But what happened? 00:04:41.090 --> 00:04:44.345 Well, what happened is the personal computer really, 00:04:44.345 --> 00:04:49.430 that came out and it became a space that 00:04:49.430 --> 00:04:53.060 industrial designers or engineers working separately on 00:04:53.060 --> 00:04:54.890 different kinds of things suddenly 00:04:54.890 --> 00:04:57.065 needed to think about things together. 00:04:57.065 --> 00:05:02.645 The other force I see strongly was the programmable VCR, 00:05:02.645 --> 00:05:05.420 onscreen programming, that actually moved 00:05:05.420 --> 00:05:08.725 a very simple programming task into many people's homes. 00:05:08.725 --> 00:05:11.650 Designers, thinking about interaction, 00:05:11.650 --> 00:05:14.525 needed an entirely new way to think about this. 00:05:14.525 --> 00:05:17.075 If you read Bill Moggridge's books, 00:05:17.075 --> 00:05:18.950 he talks about, I came up 00:05:18.950 --> 00:05:20.990 with the term interaction design. 00:05:20.990 --> 00:05:23.270 When I was designing this grid computer, 00:05:23.270 --> 00:05:25.760 they brought us this thing and asked us to 00:05:25.760 --> 00:05:29.030 design the case and the keyboard. 00:05:29.030 --> 00:05:32.600 He had the immediate reaction that you could 00:05:32.600 --> 00:05:35.960 not separate the design of the software 00:05:35.960 --> 00:05:37.610 from the design of the tools you're 00:05:37.610 --> 00:05:39.710 using to control the software and that we 00:05:39.710 --> 00:05:42.410 needed a new design practice 00:05:42.410 --> 00:05:44.485 that brought these things together. 00:05:44.485 --> 00:05:47.435 This is, I think the first lens, 00:05:47.435 --> 00:05:50.450 the first definition of HCI where there was 00:05:50.450 --> 00:05:53.780 design involved and there was a usability lens. 00:05:53.780 --> 00:05:55.940 The goal was to really 00:05:55.940 --> 00:05:59.455 make things that help users complete a task. 00:05:59.455 --> 00:06:01.970 It was to understand 00:06:01.970 --> 00:06:05.810 the model the user brought to the system and make sure 00:06:05.810 --> 00:06:08.930 the system functioned in such a way that the common sense 00:06:08.930 --> 00:06:10.370 the user brought was 00:06:10.370 --> 00:06:13.115 fulfilled in the way the system worked. 00:06:13.115 --> 00:06:17.435 The process at the time, so scenario-based design. 00:06:17.435 --> 00:06:20.390 There was a lot of testing without users, 00:06:20.390 --> 00:06:23.210 so heuristic evaluations and cognitive walkthroughs, 00:06:23.210 --> 00:06:25.070 and then of course, think a lot. 00:06:25.070 --> 00:06:27.350 At this time you didn't go out and 00:06:27.350 --> 00:06:29.680 talk to users before making a thing. 00:06:29.680 --> 00:06:33.620 You made early prototype versions and then you tested 00:06:33.620 --> 00:06:35.510 people's cognitive abilities to see 00:06:35.510 --> 00:06:38.920 if it was a good fit for what the product was. 00:06:38.920 --> 00:06:42.005 A great example of this style of 00:06:42.005 --> 00:06:45.800 thinking is the Apple's guides. 00:06:45.800 --> 00:06:48.320 I'm guessing no one's ever actually seen this. 00:06:48.320 --> 00:06:51.190 It's a really ridiculously old piece of work. 00:06:51.190 --> 00:06:53.210 But the challenge was, 00:06:53.210 --> 00:06:55.805 they were making multimedia databases 00:06:55.805 --> 00:06:57.320 before the web and 00:06:57.320 --> 00:06:59.330 they weren't really thinking hypertext. 00:06:59.330 --> 00:07:01.760 They thought people would get lost when 00:07:01.760 --> 00:07:06.475 navigating these large information spaces. 00:07:06.475 --> 00:07:08.960 One of the design ideas they had 00:07:08.960 --> 00:07:11.270 was to create these guides, 00:07:11.270 --> 00:07:15.350 these little people who would lead you through a body of 00:07:15.350 --> 00:07:17.330 work and let you know when you've 00:07:17.330 --> 00:07:20.030 left one topic and move to another. 00:07:20.030 --> 00:07:22.700 What was so interesting about this was, 00:07:22.700 --> 00:07:26.930 people quit reading this as an encyclopedia and they 00:07:26.930 --> 00:07:28.820 started to read it as this is 00:07:28.820 --> 00:07:32.030 the opinion of this guide on this material, 00:07:32.030 --> 00:07:33.875 was the unintended consequence. 00:07:33.875 --> 00:07:37.880 It actually was one of the pieces of work that 00:07:37.880 --> 00:07:42.000 raised this idea of interface agents. 00:07:42.000 --> 00:07:44.390 Like we could have agents that are helping us and 00:07:44.390 --> 00:07:47.030 you see it embodied the theme of that, 00:07:47.030 --> 00:07:50.650 carried out and still in use in products today. 00:07:50.650 --> 00:07:53.270 That was the usability lens. 00:07:53.270 --> 00:07:55.790 The problem was the usability lens 00:07:55.790 --> 00:07:57.380 didn't get us very far, 00:07:57.380 --> 00:08:00.365 so it was very good at modeling the problem, 00:08:00.365 --> 00:08:04.000 but there were still a number of breakdowns. 00:08:04.000 --> 00:08:06.635 A big part of this was the idea that work is 00:08:06.635 --> 00:08:09.470 social and it's not just about getting the task done, 00:08:09.470 --> 00:08:11.780 but there's much more meaning going 00:08:11.780 --> 00:08:15.505 on that determines people's behaviors and choices. 00:08:15.505 --> 00:08:19.380 This is where the user-centered design lens comes from. 00:08:19.380 --> 00:08:21.615 It didn't get rid of usability. 00:08:21.615 --> 00:08:23.690 It's simply expanded the focus 00:08:23.690 --> 00:08:25.660 of the work that was taking place. 00:08:25.660 --> 00:08:27.710 Here the goal was to 00:08:27.710 --> 00:08:30.020 understand the work process and the contexts, 00:08:30.020 --> 00:08:32.645 contexts being that particularly new piece. 00:08:32.645 --> 00:08:35.330 To understand users needs and goals, 00:08:35.330 --> 00:08:37.040 not just their tasks, 00:08:37.040 --> 00:08:40.460 and to envision how IT can be 00:08:40.460 --> 00:08:44.240 automated to really improve workflow. 00:08:44.240 --> 00:08:45.820 I would say this is also like, 00:08:45.820 --> 00:08:47.540 where can we get rid of users 00:08:47.540 --> 00:08:49.565 and replace them with computers. 00:08:49.565 --> 00:08:52.170 This is our history. 00:08:54.210 --> 00:08:57.370 But one of the nice focuses here 00:08:57.370 --> 00:08:59.635 was this was also about skilling people. 00:08:59.635 --> 00:09:00.910 Not just deskilling. 00:09:00.910 --> 00:09:04.975 But how can I build skills by giving people new tools? 00:09:04.975 --> 00:09:07.240 The major change in the process 00:09:07.240 --> 00:09:08.890 was to go out and observe people, 00:09:08.890 --> 00:09:10.820 to watch people do work, 00:09:10.820 --> 00:09:12.330 and to bring that into 00:09:12.330 --> 00:09:14.445 the design of computational systems, 00:09:14.445 --> 00:09:16.290 to synthesize the information 00:09:16.290 --> 00:09:17.849 that's collected with models, 00:09:17.849 --> 00:09:21.185 that then drove the design process. 00:09:21.185 --> 00:09:25.840 Participatory design, contextual design, personas, 00:09:25.840 --> 00:09:27.340 these are all design methods that 00:09:27.340 --> 00:09:29.170 came out of this shift from 00:09:29.170 --> 00:09:31.300 a cognitive usability focus to 00:09:31.300 --> 00:09:34.135 a much more social work focus. 00:09:34.135 --> 00:09:35.890 One of my favorite pieces of 00:09:35.890 --> 00:09:38.800 work at this time isn't a computer at all, 00:09:38.800 --> 00:09:43.900 it's reprographics machine from Xerox. 00:09:43.900 --> 00:09:47.620 This is work done by Fits Richardson Smith. 00:09:47.620 --> 00:09:49.240 I mean, they did a bunch 00:09:49.240 --> 00:09:51.505 of immersive ethnographic research, 00:09:51.505 --> 00:09:55.465 watching people in offices use copiers. 00:09:55.465 --> 00:09:57.340 They found this problem, 00:09:57.340 --> 00:10:01.600 that the workers really needed to own their documents, 00:10:01.600 --> 00:10:04.075 but there was a person called a key operator. 00:10:04.075 --> 00:10:07.190 Anybody, remember the time of key operators? 00:10:07.190 --> 00:10:10.740 The old folks. The key operator 00:10:10.740 --> 00:10:12.405 was the person with the key, 00:10:12.405 --> 00:10:14.660 who could operate the copier. 00:10:14.660 --> 00:10:18.715 They control the order that jobs got done. 00:10:18.715 --> 00:10:20.320 People would try to get 00:10:20.320 --> 00:10:22.690 their job ahead of somebody else's job, 00:10:22.690 --> 00:10:24.790 because this was a big bottleneck. 00:10:24.790 --> 00:10:27.760 If vice president's job needed to get done, 00:10:27.760 --> 00:10:31.405 suddenly your work was falling by the wayside. 00:10:31.405 --> 00:10:34.030 At the time copiers were quite dangerous, 00:10:34.030 --> 00:10:36.685 they were full of sharp metal pieces 00:10:36.685 --> 00:10:39.740 and moving parts that could injure you. 00:10:39.930 --> 00:10:44.455 Based on this insight of understanding work, 00:10:44.455 --> 00:10:50.590 the designers came up with a radical idea for that time. 00:10:50.590 --> 00:10:52.630 Not radical at all now, which is, 00:10:52.630 --> 00:10:54.415 can we make a copier that 00:10:54.415 --> 00:10:57.595 anyone can simply walk up and use? 00:10:57.595 --> 00:11:01.150 Can we make a copier that if it has a problem, 00:11:01.150 --> 00:11:03.490 anyone can fix it? 00:11:03.490 --> 00:11:05.739 This has a bunch of advantages. 00:11:05.739 --> 00:11:08.095 One is simply engineering at the time, 00:11:08.095 --> 00:11:11.154 Xerox was spending hundreds of millions of dollars, 00:11:11.154 --> 00:11:14.620 to make copiers that were very fast that never failed, 00:11:14.620 --> 00:11:17.110 because the amount of time it took to 00:11:17.110 --> 00:11:19.885 fix a problem was huge and that was like, 00:11:19.885 --> 00:11:24.115 the determining factor for why you would buy a copier. 00:11:24.115 --> 00:11:26.020 What it was doing was 00:11:26.020 --> 00:11:27.790 creating this different way of thinking 00:11:27.790 --> 00:11:29.290 about a copy that actually made 00:11:29.290 --> 00:11:30.955 the engineering less important. 00:11:30.955 --> 00:11:35.755 We could actually make copiers that worked worse. 00:11:35.755 --> 00:11:37.630 They could have much worse performance, 00:11:37.630 --> 00:11:40.270 because the end-user could fix the problem. 00:11:40.270 --> 00:11:43.225 Xerox actually had to develop a demo, 00:11:43.225 --> 00:11:45.910 where their product failed, 00:11:45.910 --> 00:11:49.720 in order to show how easy it was to fix the product, 00:11:49.720 --> 00:11:52.015 which is a crazy idea. 00:11:52.015 --> 00:11:54.490 There's a Xerox vice president, "Oh, 00:11:54.490 --> 00:11:59.450 my copier is broken," demoing how simple this was. 00:12:00.090 --> 00:12:03.805 When they created this device, 00:12:03.805 --> 00:12:07.090 they thought very strongly about extensibility. 00:12:07.090 --> 00:12:11.470 This is earlier '80s when they're doing this work. 00:12:11.470 --> 00:12:14.155 I don't know how many of you are familiar with 00:12:14.155 --> 00:12:16.840 Chris Alexander's ideas of design patterns, 00:12:16.840 --> 00:12:18.820 which are conventions that 00:12:18.820 --> 00:12:21.985 grow over time as things are repeatedly built. 00:12:21.985 --> 00:12:23.890 They turned that around to say, 00:12:23.890 --> 00:12:26.140 designers can actually create this, 00:12:26.140 --> 00:12:28.495 they can create a design language, 00:12:28.495 --> 00:12:30.490 which is an intentional set of 00:12:30.490 --> 00:12:34.135 patterns that make a product extensible. 00:12:34.135 --> 00:12:36.670 The idea that the paper goes on the right, 00:12:36.670 --> 00:12:38.065 and feeds to the left, 00:12:38.065 --> 00:12:41.245 that where the paper go is marked in blue. 00:12:41.245 --> 00:12:43.060 A lot of these patterns, 00:12:43.060 --> 00:12:45.205 the way the toner cartridge goes in, 00:12:45.205 --> 00:12:48.699 a lot of those patterns are still alive today, 00:12:48.699 --> 00:12:51.130 even though it's 30 plus years 00:12:51.130 --> 00:12:55.180 since these design aspects were put into copiers. 00:12:55.180 --> 00:12:58.930 This style worked really well, 00:12:58.930 --> 00:13:01.430 and then things changed again. 00:13:03.990 --> 00:13:07.060 Mobile phones become really popular. 00:13:07.060 --> 00:13:09.370 Lots more games going into 00:13:09.370 --> 00:13:13.510 people's homes that have more complicated interfaces. 00:13:13.510 --> 00:13:15.990 Personal music players. 00:13:15.990 --> 00:13:18.525 But also the biggest one, the web. 00:13:18.525 --> 00:13:23.530 The web created a bunch of interactive things. 00:13:23.870 --> 00:13:27.420 They complicated the sense of purposefulness, 00:13:27.420 --> 00:13:30.540 and this is an example of work by 00:13:30.540 --> 00:13:33.700 Alban and Ferris Design from where they 00:13:33.700 --> 00:13:37.405 had to design a website for the Monterey Bay Aquarium. 00:13:37.405 --> 00:13:38.800 What does that even mean? 00:13:38.800 --> 00:13:40.419 Well, you were supposed to have an aquarium 00:13:40.419 --> 00:13:43.660 experience on the web, 00:13:43.660 --> 00:13:46.735 but that didn't make you not want to go to the aquarium. 00:13:46.735 --> 00:13:49.840 It wasn't replacing going into the aquarium experience, 00:13:49.840 --> 00:13:52.075 but it was supposed to be its own thing. 00:13:52.075 --> 00:13:54.190 The methods and processes of 00:13:54.190 --> 00:13:57.040 user-centered design don't really map to this space, 00:13:57.040 --> 00:13:58.180 because it's not about tasks, 00:13:58.180 --> 00:13:59.530 it's not about work, 00:13:59.530 --> 00:14:01.165 it's about something else. 00:14:01.165 --> 00:14:05.605 This is where the user experience lens gets added. 00:14:05.605 --> 00:14:07.330 Again, it's not like usability 00:14:07.330 --> 00:14:08.980 and user-centered design go away, 00:14:08.980 --> 00:14:11.755 simply the definition of HCI expands, 00:14:11.755 --> 00:14:13.450 to incorporate this new set of 00:14:13.450 --> 00:14:17.275 challenges because of the new things that are being made. 00:14:17.275 --> 00:14:19.900 It's still about understanding people, 00:14:19.900 --> 00:14:21.565 but in addition to their needs, 00:14:21.565 --> 00:14:24.595 it's really trying to understand their desires. 00:14:24.595 --> 00:14:26.470 It's looking at how people 00:14:26.470 --> 00:14:28.570 construct meaning in their lives. 00:14:28.570 --> 00:14:31.690 It's envisioning technology with the goal of how can I 00:14:31.690 --> 00:14:34.090 enhance the user's experience 00:14:34.090 --> 00:14:36.025 through their interaction with the product? 00:14:36.025 --> 00:14:38.530 But from a commercial point of view, it's sort of, 00:14:38.530 --> 00:14:40.630 how can I make a consumer take 00:14:40.630 --> 00:14:42.130 this product off the shelf 00:14:42.130 --> 00:14:44.155 and put it into their shopping cart? 00:14:44.155 --> 00:14:47.635 It's a consumptive design process. 00:14:47.635 --> 00:14:49.210 The process is okay, 00:14:49.210 --> 00:14:50.740 we're still going to observe users 00:14:50.740 --> 00:14:52.390 but now we're going to observe how they live, 00:14:52.390 --> 00:14:54.040 not just how they work. 00:14:54.040 --> 00:14:56.020 Experience prototyping. 00:14:56.020 --> 00:14:58.509 Can we simulate this possible experience 00:14:58.509 --> 00:15:00.220 to understand it? 00:15:00.220 --> 00:15:03.400 Cultural probes and dream kits to get at 00:15:03.400 --> 00:15:06.580 people's more hidden aspirations. 00:15:06.580 --> 00:15:08.470 Then ideas like mood boards, 00:15:08.470 --> 00:15:12.160 which is a super traditional design technique 00:15:12.160 --> 00:15:13.630 from industrial design. 00:15:13.630 --> 00:15:15.550 This was a shift, 00:15:15.550 --> 00:15:18.010 but it was actually work industrial designers 00:15:18.010 --> 00:15:20.650 and graphic designers and people in multimedia, 00:15:20.650 --> 00:15:22.570 had always done, which was 00:15:22.570 --> 00:15:24.775 making things that people would choose. 00:15:24.775 --> 00:15:27.925 It's bringing a lot of very traditional design ideas, 00:15:27.925 --> 00:15:30.745 into an HCI practice. 00:15:30.745 --> 00:15:35.050 An example just to draw this idea out that I really like, 00:15:35.050 --> 00:15:36.700 I used to work at Philips. 00:15:36.700 --> 00:15:38.770 I'll promote them a little bit. 00:15:38.770 --> 00:15:41.300 Although they did lay me off. 00:15:41.430 --> 00:15:44.050 This is the living memory table. 00:15:44.050 --> 00:15:48.220 It was done as part of a Y2K project, 00:15:48.220 --> 00:15:50.050 right at the turn of the millennium. 00:15:50.050 --> 00:15:52.810 The design team spent month 00:15:52.810 --> 00:15:55.465 hanging out in a small town in Scotland, 00:15:55.465 --> 00:15:58.090 trying to understand how people kept 00:15:58.090 --> 00:16:01.510 alive the oral history of this town. 00:16:01.510 --> 00:16:05.725 They hung out in a lot of cafes and pubs, 00:16:05.725 --> 00:16:10.195 watching how people would share stories with each other. 00:16:10.195 --> 00:16:13.150 They created this table, 00:16:13.150 --> 00:16:16.030 which was a place for sharing stories. 00:16:16.030 --> 00:16:18.025 It's a cafe table. 00:16:18.025 --> 00:16:21.025 It's got a circular touchscreen. 00:16:21.025 --> 00:16:24.130 You can share a story with it or you can also 00:16:24.130 --> 00:16:27.580 listen to stories other town's people have had. 00:16:27.580 --> 00:16:30.460 Then there are these radiofrequency tokens 00:16:30.460 --> 00:16:33.655 that you can use to take a story with you. 00:16:33.655 --> 00:16:36.040 It's a really lovely piece of work, 00:16:36.040 --> 00:16:38.830 but it's really pushing the boundaries of what we 00:16:38.830 --> 00:16:41.860 think of interactive technology. 00:16:41.860 --> 00:16:48.050 It was transformative of this idea of the cafe table, 00:16:48.180 --> 00:16:51.640 from a witness of the oral history to 00:16:51.640 --> 00:16:54.280 an active participant in capturing, 00:16:54.280 --> 00:16:56.965 and recording, and sharing oral history. 00:16:56.965 --> 00:17:02.150 With each lens, HCI is growing. 00:17:05.460 --> 00:17:08.200 I'm just trying to push this idea that 00:17:08.200 --> 00:17:10.465 identity is constantly unfolding. 00:17:10.465 --> 00:17:11.500 Are we ready for 00:17:11.500 --> 00:17:16.315 another step forward and expansion of identity? 00:17:16.315 --> 00:17:20.560 Some claims to ground the perspective I'm coming from, 00:17:20.560 --> 00:17:24.790 I will claim and try to argue that user-centered design 00:17:24.790 --> 00:17:26.080 and user experience design 00:17:26.080 --> 00:17:28.795 follow a product-centric process. 00:17:28.795 --> 00:17:30.850 It's about making a thing 00:17:30.850 --> 00:17:32.680 that you want someone to purchase, 00:17:32.680 --> 00:17:35.830 or making things that are so attractive, 00:17:35.830 --> 00:17:38.860 lots and lots of people will purchase them. 00:17:38.860 --> 00:17:41.710 It's very much focused on 00:17:41.710 --> 00:17:46.405 user singular or users plural, and their needs. 00:17:46.405 --> 00:17:51.260 If you talk to user experience design teams in industry, 00:17:51.260 --> 00:17:53.360 they often refer to themselves as 00:17:53.360 --> 00:17:57.670 worthy user advocate in the development process. 00:17:57.670 --> 00:18:00.500 There's a very much strong focus 00:18:00.500 --> 00:18:03.240 on making one application, 00:18:03.240 --> 00:18:06.740 or one device, or one computer system. 00:18:06.740 --> 00:18:09.200 With that, it's like the solution to 00:18:09.200 --> 00:18:12.305 every problem is one computer. 00:18:12.305 --> 00:18:16.310 It's a narrowness in definition. 00:18:16.710 --> 00:18:19.270 HCI as it's currently 00:18:19.270 --> 00:18:21.320 practicing and as it's currently defines itself, 00:18:21.320 --> 00:18:23.180 never really talks about money. 00:18:23.180 --> 00:18:25.205 It never talks about payments, 00:18:25.205 --> 00:18:27.440 because that's marketing and business, 00:18:27.440 --> 00:18:30.440 and so there's a definitional boundary that puts 00:18:30.440 --> 00:18:36.440 that part of the work outside of HCI practice. 00:18:37.200 --> 00:18:40.470 Now I want to talk about, where is this not working? 00:18:40.470 --> 00:18:42.860 This is what HCI is and how we got there. 00:18:42.860 --> 00:18:48.180 How are these new things we're making, creating problems? 00:18:50.220 --> 00:18:52.750 So just a quick review of the things 00:18:52.750 --> 00:18:54.790 I'm talking about: services, 00:18:54.790 --> 00:18:57.490 social computing systems, and 00:18:57.490 --> 00:19:00.685 systems with the intention of driving social change. 00:19:00.685 --> 00:19:03.280 I just want to get 00:19:03.280 --> 00:19:05.440 an agreement on what 00:19:05.440 --> 00:19:08.695 products and services are and how they're different. 00:19:08.695 --> 00:19:14.140 A cup of coffee product, Starbucks service. 00:19:14.140 --> 00:19:17.990 They can be strongly connected. 00:19:18.210 --> 00:19:23.155 Companies make products for consumers, 00:19:23.155 --> 00:19:26.380 but they do for a customer. 00:19:26.380 --> 00:19:31.380 Service becomes the do not make and 00:19:31.380 --> 00:19:34.250 an exchange between a customer 00:19:34.250 --> 00:19:37.959 and a product company, ownership is exchanged. 00:19:37.959 --> 00:19:39.370 So when I buy this thing, 00:19:39.370 --> 00:19:42.760 I own it and possess it and I've exchanged money for it. 00:19:42.760 --> 00:19:44.320 But with a service, there's never 00:19:44.320 --> 00:19:46.540 an exchange of ownership. 00:19:46.540 --> 00:19:49.120 I may go into a Starbucks, 00:19:49.120 --> 00:19:50.350 but I do not own 00:19:50.350 --> 00:19:52.180 the Starbucks and I do 00:19:52.180 --> 00:19:54.430 not own the employees that are there, 00:19:54.430 --> 00:19:56.590 but I can still purchase a cup of coffee 00:19:56.590 --> 00:19:59.650 and have ownership from it. 00:19:59.650 --> 00:20:03.385 In a product sense, value is exchanged. 00:20:03.385 --> 00:20:05.185 It's a money for a thing. 00:20:05.185 --> 00:20:07.960 But in a service values experience is 00:20:07.960 --> 00:20:12.085 the performance that's just in that moment. 00:20:12.085 --> 00:20:15.490 It's not exchanged in the same way. 00:20:15.490 --> 00:20:18.805 Another way to think of products is they're tangible. 00:20:18.805 --> 00:20:20.650 It makes less sense with the cup of coffee. 00:20:20.650 --> 00:20:21.970 I mean it's a tangible thing. 00:20:21.970 --> 00:20:23.890 But the idea is you could warehouse it. 00:20:23.890 --> 00:20:26.260 You could create a whole bunch. 00:20:26.260 --> 00:20:28.570 But with a service, it can't. It's intangible. 00:20:28.570 --> 00:20:30.280 It's this thing that is a performance 00:20:30.280 --> 00:20:32.275 and has an ephemeral quality. 00:20:32.275 --> 00:20:33.640 When we're thinking about what 00:20:33.640 --> 00:20:34.900 is it we're trying to make, 00:20:34.900 --> 00:20:37.180 it's actually really important to 00:20:37.180 --> 00:20:40.375 know if you're making a product or a service. 00:20:40.375 --> 00:20:43.510 Also, understand that products and 00:20:43.510 --> 00:20:47.035 services work together in product-service systems. 00:20:47.035 --> 00:20:49.900 But in general, the service is the glue that 00:20:49.900 --> 00:20:51.520 holds the relationship between 00:20:51.520 --> 00:20:54.200 the various products together. 00:20:55.890 --> 00:21:00.880 In HCI, one of the ways service has really come in is in 00:21:00.880 --> 00:21:05.650 the transition that's taking place from COTS. 00:21:05.650 --> 00:21:06.850 I don't know if that's a familiar term, 00:21:06.850 --> 00:21:09.430 but that's conventional off-the-shelf software 00:21:09.430 --> 00:21:11.710 to software as a service. 00:21:11.710 --> 00:21:15.130 There used to be a time you could go to a store and buy 00:21:15.130 --> 00:21:18.670 software in a shrink-wrapped box and you owned it. 00:21:18.670 --> 00:21:20.530 We don't live in that time 00:21:20.530 --> 00:21:24.505 anymore and there's a lot of problems with that. 00:21:24.505 --> 00:21:26.510 Software was very expensive. 00:21:26.510 --> 00:21:28.605 But you have versioning issue. 00:21:28.605 --> 00:21:29.820 You bought it and you own it, 00:21:29.820 --> 00:21:31.485 but when a new version came out, 00:21:31.485 --> 00:21:33.660 you had to buy the new version or at least 00:21:33.660 --> 00:21:36.525 pay something for a new upgrade. 00:21:36.525 --> 00:21:39.860 Particularly for industrial purchasers, 00:21:39.860 --> 00:21:42.160 if I'm buying 200 computers and 00:21:42.160 --> 00:21:44.395 copies of software for my employees, 00:21:44.395 --> 00:21:46.600 I don't necessarily want to pay 00:21:46.600 --> 00:21:49.450 for 200 upgrades all at the same time. 00:21:49.450 --> 00:21:51.655 But if I add a new computer in, 00:21:51.655 --> 00:21:53.170 I can't get the old software, 00:21:53.170 --> 00:21:55.780 so I have to get the new software and suddenly I've got 00:21:55.780 --> 00:21:57.010 a versioning problem because 00:21:57.010 --> 00:21:58.390 some people have this version, 00:21:58.390 --> 00:22:00.145 some people have that problem. 00:22:00.145 --> 00:22:02.845 That version they don't go together. 00:22:02.845 --> 00:22:05.050 Software as a service really changes 00:22:05.050 --> 00:22:07.000 that because you move that ownership 00:22:07.000 --> 00:22:08.680 back to the service provider and you're 00:22:08.680 --> 00:22:12.025 just using the software. 00:22:12.025 --> 00:22:15.730 It's about a pay-as-you-go model or 00:22:15.730 --> 00:22:19.495 even a free model where you're using it for free. 00:22:19.495 --> 00:22:23.710 Microsoft Office versus Google Docs is 00:22:23.710 --> 00:22:25.420 a great example of 00:22:25.420 --> 00:22:28.480 conventional off-the-shelf versus software as a service. 00:22:28.480 --> 00:22:31.885 But you can see it complicates this in a sense. 00:22:31.885 --> 00:22:35.020 Here the user is the customer. 00:22:35.020 --> 00:22:36.880 Here I understand who the user is, 00:22:36.880 --> 00:22:38.500 but who's the customer? 00:22:38.500 --> 00:22:39.535 Because the customer is 00:22:39.535 --> 00:22:41.990 the person that's paying the money. 00:22:42.270 --> 00:22:45.265 There's something else going on there. 00:22:45.265 --> 00:22:47.860 The other place that HCI is really getting into 00:22:47.860 --> 00:22:50.230 services is they are 00:22:50.230 --> 00:22:53.020 making digital tools that are 00:22:53.020 --> 00:22:56.380 used in the performance of service delivery. 00:22:56.380 --> 00:22:58.780 This is Cartwheel. 00:22:58.780 --> 00:23:00.715 I don't know if any of you have used this. 00:23:00.715 --> 00:23:03.220 It's a mobile app recently created by 00:23:03.220 --> 00:23:07.270 Target for customers to use when they're in the store, 00:23:07.270 --> 00:23:09.175 when they're having an in-store experience. 00:23:09.175 --> 00:23:10.480 It's this crossing of 00:23:10.480 --> 00:23:13.330 a digital-physical divide in 00:23:13.330 --> 00:23:16.345 the way that people are interacting with things. 00:23:16.345 --> 00:23:20.210 Services raise a bunch of challenges. 00:23:21.200 --> 00:23:24.510 When you have free mobile apps or 99 cents 00:23:24.510 --> 00:23:27.105 mobile apps or free online services, 00:23:27.105 --> 00:23:29.814 you're really complicating this idea 00:23:29.814 --> 00:23:31.750 that the user is the customer, 00:23:31.750 --> 00:23:32.920 which is fundamentally where 00:23:32.920 --> 00:23:34.300 user-centered design and user 00:23:34.300 --> 00:23:36.565 experience design come from. 00:23:36.565 --> 00:23:39.970 The design goal shifts from this desire to 00:23:39.970 --> 00:23:43.465 buy to either the idea of, 00:23:43.465 --> 00:23:46.690 I want to maximize your use or minimize your use. 00:23:46.690 --> 00:23:49.450 With Facebook, they always want you to use Facebook 00:23:49.450 --> 00:23:50.560 more because you are consuming 00:23:50.560 --> 00:23:52.520 more ads, that's more revenue. 00:23:52.520 --> 00:23:54.480 With Netflix, they would really 00:23:54.480 --> 00:23:56.490 rather that you kept paying but watch 00:23:56.490 --> 00:24:01.335 fewer movies because that helps their bottom line. 00:24:01.335 --> 00:24:06.125 Those become determinants of how you set up a service. 00:24:06.125 --> 00:24:08.980 Notice here, with all of these, 00:24:08.980 --> 00:24:13.120 the financial model is fundamental to the design. 00:24:13.120 --> 00:24:15.040 It can't be done in isolation. 00:24:15.040 --> 00:24:16.300 You don't make the thing and then 00:24:16.300 --> 00:24:18.550 figure out how to make money out of it. 00:24:18.550 --> 00:24:22.120 The design of the financial model 00:24:22.120 --> 00:24:25.495 has to be in the way that people use it. 00:24:25.495 --> 00:24:29.050 Finally, in thinking particularly of 00:24:29.050 --> 00:24:30.910 the products meant to work in 00:24:30.910 --> 00:24:33.025 a brick and mortar service environment, 00:24:33.025 --> 00:24:37.810 they're crossing a digital-physical divide 00:24:37.810 --> 00:24:41.275 but often the teams are working separately. 00:24:41.275 --> 00:24:44.065 I was just working with some folks from Walgreens, 00:24:44.065 --> 00:24:48.415 their digital team was doing the website, 00:24:48.415 --> 00:24:51.010 which is this entirely separate experience, 00:24:51.010 --> 00:24:52.465 and then they had their in-store team. 00:24:52.465 --> 00:24:54.790 Now they're making an app for 00:24:54.790 --> 00:24:57.760 their in-store experience but it's being 00:24:57.760 --> 00:24:59.260 done by the digital team who never 00:24:59.260 --> 00:25:01.090 thinks about the store. 00:25:01.090 --> 00:25:04.060 They're running into all kinds of problems because they 00:25:04.060 --> 00:25:07.420 work in entirely different divisions. 00:25:07.420 --> 00:25:10.390 It's that's problematic. 00:25:10.390 --> 00:25:12.550 But in the definitional boundary 00:25:12.550 --> 00:25:13.900 of human-computer interaction, 00:25:13.900 --> 00:25:17.320 we stop as soon as people quit using a computer. 00:25:17.320 --> 00:25:19.915 That's the boundary that we run into. 00:25:19.915 --> 00:25:23.635 I just want to pull out the example of Waze. 00:25:23.635 --> 00:25:27.700 Think about this user customer model. 00:25:27.700 --> 00:25:29.440 Are people familiar with Waze? 00:25:29.440 --> 00:25:31.075 They've been in the news lately. 00:25:31.075 --> 00:25:32.950 They were in the news a while ago when Google 00:25:32.950 --> 00:25:35.230 paid a billion dollars for them. 00:25:35.230 --> 00:25:42.745 This is a navigation service that's free. 00:25:42.745 --> 00:25:44.920 What happens is when you use 00:25:44.920 --> 00:25:49.360 this navigation service with your mobile phone, 00:25:49.360 --> 00:25:52.090 you're uploading your location data 00:25:52.090 --> 00:25:54.340 which is used to model traffic. 00:25:54.340 --> 00:25:56.110 So if I have enough people using this, 00:25:56.110 --> 00:25:59.425 I can see what the flow of traffic is. 00:25:59.425 --> 00:26:00.970 That's what Waze does, 00:26:00.970 --> 00:26:03.205 and then Waze resells it. 00:26:03.205 --> 00:26:06.310 The users are doing labor, 00:26:06.310 --> 00:26:08.800 that's value for this service company who 00:26:08.800 --> 00:26:11.410 can then resell it to make the revenue. 00:26:11.410 --> 00:26:14.080 But it's a more complicated model 00:26:14.080 --> 00:26:17.185 than just saying what the users want. 00:26:17.185 --> 00:26:21.025 You can't get there from a user-centered point of view. 00:26:21.025 --> 00:26:27.895 So services create a lot of challenges. 00:26:27.895 --> 00:26:31.510 They need design teams to really 00:26:31.510 --> 00:26:35.035 consider a whole bunch of stakeholders, not just users. 00:26:35.035 --> 00:26:38.875 They often follow a disruptive innovation model. 00:26:38.875 --> 00:26:41.890 It's about making something that sells for a lot 00:26:41.890 --> 00:26:45.115 less but then delivers less too, 00:26:45.115 --> 00:26:47.170 which is great from a business point of view. 00:26:47.170 --> 00:26:50.260 You're stealing some conventional industry 00:26:50.260 --> 00:26:53.110 sort of worst customers they're not going to fight for, 00:26:53.110 --> 00:26:56.410 but from a user experience design point of view, 00:26:56.410 --> 00:26:59.650 HCI people are taught to always make something more. 00:26:59.650 --> 00:27:02.500 Keep adding, don't start removing, 00:27:02.500 --> 00:27:04.940 don't make things worse. 00:27:05.280 --> 00:27:07.510 User experience design and 00:27:07.510 --> 00:27:09.040 user-centered design never considered 00:27:09.040 --> 00:27:10.840 the financial model but it's completely 00:27:10.840 --> 00:27:13.615 critical to the design of these things. 00:27:13.615 --> 00:27:15.520 When you start to look at 00:27:15.520 --> 00:27:19.210 particularly the physical service 00:27:19.210 --> 00:27:21.070 and the digital in support of that, 00:27:21.070 --> 00:27:22.450 there's a real need to cross 00:27:22.450 --> 00:27:25.885 this digital-physical divide to do it well. 00:27:25.885 --> 00:27:29.320 Social computing, it's been around forever. 00:27:29.320 --> 00:27:31.120 It's a strong tradition 00:27:31.120 --> 00:27:33.070 of computer-supported cooperative work, 00:27:33.070 --> 00:27:34.960 so it's like how can this not fit HCI? 00:27:34.960 --> 00:27:37.915 It's been around since almost the beginning of HCI? 00:27:37.915 --> 00:27:40.360 But it's transforming. At first, 00:27:40.360 --> 00:27:43.000 it was very much collaboration tools for workers, 00:27:43.000 --> 00:27:44.875 looking at things like bulletin boards. 00:27:44.875 --> 00:27:47.950 But then as it moved into a consumer space, 00:27:47.950 --> 00:27:48.970 it started to change. 00:27:48.970 --> 00:27:51.790 So rating systems and reviewing 00:27:51.790 --> 00:27:54.850 systems, collective information systems, 00:27:54.850 --> 00:27:58.450 crowdworking systems, peer-economy systems all began to 00:27:58.450 --> 00:28:00.520 complicate and conflict with 00:28:00.520 --> 00:28:03.475 user-centered design and user experience design. 00:28:03.475 --> 00:28:06.550 The goal with all of these systems is 00:28:06.550 --> 00:28:09.430 really not to make something that lots 00:28:09.430 --> 00:28:11.440 of people want to buy but to make 00:28:11.440 --> 00:28:15.670 a sustainable social ecology of exchange. 00:28:15.670 --> 00:28:18.025 We're trying to make a little life form 00:28:18.025 --> 00:28:22.400 of exchange among different entities. 00:28:23.190 --> 00:28:26.425 Social computing starts to break down that. 00:28:26.425 --> 00:28:28.060 Just like with services, 00:28:28.060 --> 00:28:30.160 you have many stakeholders now, 00:28:30.160 --> 00:28:32.335 not just a singular user. 00:28:32.335 --> 00:28:37.105 You have customers and 00:28:37.105 --> 00:28:39.430 you have to think about how does that 00:28:39.430 --> 00:28:42.385 fit with this user or provider. 00:28:42.385 --> 00:28:45.355 You have disruptive innovation. 00:28:45.355 --> 00:28:47.200 It works the same way. 00:28:47.200 --> 00:28:50.740 Systems like Waze, like Uber, 00:28:50.740 --> 00:28:52.060 are very much in 00:28:52.060 --> 00:28:53.620 this social computing sense 00:28:53.620 --> 00:28:56.305 but they're disruptive innovation. 00:28:56.305 --> 00:28:59.320 You still have to address the financial model because 00:28:59.320 --> 00:29:01.750 that's fundamentally how many of these systems work. 00:29:01.750 --> 00:29:03.415 It's not necessarily money. 00:29:03.415 --> 00:29:06.520 But gamification, which is big in social computing, 00:29:06.520 --> 00:29:09.220 is fundamentally about a type of reward to drive 00:29:09.220 --> 00:29:13.495 behavior which pushes us towards economics. 00:29:13.495 --> 00:29:15.610 There's definitely an intention 00:29:15.610 --> 00:29:17.695 to shape people's behavior. 00:29:17.695 --> 00:29:19.150 I want them to behave like this 00:29:19.150 --> 00:29:21.115 because that makes my ecology work. 00:29:21.115 --> 00:29:23.500 That's not really about fundamentally meeting 00:29:23.500 --> 00:29:27.950 the needs and desires of users. 00:29:28.290 --> 00:29:30.970 The problem gets even murkier when 00:29:30.970 --> 00:29:33.280 you add social innovation, so this is, 00:29:33.280 --> 00:29:35.920 I think one of the most exciting areas 00:29:35.920 --> 00:29:38.065 of current HCI research. 00:29:38.065 --> 00:29:40.450 There's not a lot of practice there, 00:29:40.450 --> 00:29:41.680 and so there's a tradition of 00:29:41.680 --> 00:29:43.690 that mostly around accessibility, 00:29:43.690 --> 00:29:45.400 making it easier for people with 00:29:45.400 --> 00:29:47.425 disabilities to use computers. 00:29:47.425 --> 00:29:49.660 There's a tradition of using computers in 00:29:49.660 --> 00:29:52.000 education that still works 00:29:52.000 --> 00:29:54.715 very much with a user-centered design focus. 00:29:54.715 --> 00:29:56.080 But when you start to take on 00:29:56.080 --> 00:29:58.390 problems like sustainability, 00:29:58.390 --> 00:30:02.710 or getting people to eat healthy to reduce rates of 00:30:02.710 --> 00:30:07.210 obesity or HCI for the developing world, 00:30:07.210 --> 00:30:08.230 user-centered design, 00:30:08.230 --> 00:30:11.360 user experience design start to break down. 00:30:11.520 --> 00:30:13.840 This whole thing is triggered by 00:30:13.840 --> 00:30:17.290 the perception that social computing can't create 00:30:17.290 --> 00:30:22.510 a new collective action that can drive change. 00:30:22.510 --> 00:30:24.489 Many of the same breakdowns, 00:30:24.489 --> 00:30:26.980 but the big one here is that, 00:30:26.980 --> 00:30:28.630 now you're actually trying to 00:30:28.630 --> 00:30:30.760 radically change people's behavior, 00:30:30.760 --> 00:30:33.580 not meet their needs and desires. 00:30:33.580 --> 00:30:36.835 In general, when you're looking at social innovation, 00:30:36.835 --> 00:30:39.100 you're saying that collective good is 00:30:39.100 --> 00:30:42.040 much more important than the individual good, 00:30:42.040 --> 00:30:44.380 which is not at all an idea that comes from 00:30:44.380 --> 00:30:48.020 user-centered design and user experience design. 00:30:49.050 --> 00:30:51.310 That's about what's not working. 00:30:51.310 --> 00:30:53.140 Let me give you 00:30:53.140 --> 00:30:54.760 a brief breakdown of what 00:30:54.760 --> 00:30:57.070 service is and why I think it can 00:30:57.070 --> 00:31:03.385 help to span this identity breakdown, 00:31:03.385 --> 00:31:06.505 and then talk about some of the problems with it as well. 00:31:06.505 --> 00:31:09.670 Here's a nice graphic from New York Times 00:31:09.670 --> 00:31:12.580 just showing over the last 40 years how 00:31:12.580 --> 00:31:14.935 our country has shifted 00:31:14.935 --> 00:31:21.080 from a products country to a services country. 00:31:21.210 --> 00:31:24.130 The yellow ones are product focus, 00:31:24.130 --> 00:31:25.840 blue are services you're just seeing 00:31:25.840 --> 00:31:29.480 this giant shift in time. 00:31:30.690 --> 00:31:34.150 Service design is really emerges out of 00:31:34.150 --> 00:31:36.895 this shift and it comes from two places, 00:31:36.895 --> 00:31:39.580 operations research, and we were just talking about this 00:31:39.580 --> 00:31:41.320 earlier like do those people 00:31:41.320 --> 00:31:43.165 even exist that do this work? 00:31:43.165 --> 00:31:45.685 What happen to those operations research? 00:31:45.685 --> 00:31:47.680 They had to change because the idea 00:31:47.680 --> 00:31:49.480 of making production of things more 00:31:49.480 --> 00:31:53.080 efficient becomes less meaningful 00:31:53.080 --> 00:31:56.095 as there's less production of things, 00:31:56.095 --> 00:31:57.790 and so they switched to 00:31:57.790 --> 00:32:00.010 starting to look at service delivery, 00:32:00.010 --> 00:32:01.750 and then suddenly realized, "Oh, 00:32:01.750 --> 00:32:03.550 there's this thing called customers", 00:32:03.550 --> 00:32:05.140 and they need to have an experience. 00:32:05.140 --> 00:32:06.910 Because operations research never 00:32:06.910 --> 00:32:08.755 worried about the consumer, 00:32:08.755 --> 00:32:11.440 they were only worried about making efficiencies and 00:32:11.440 --> 00:32:14.605 making the things that then consumers would buy. 00:32:14.605 --> 00:32:17.305 Was also a huge change in marketing, 00:32:17.305 --> 00:32:21.040 you're not selling things you're selling experiences. 00:32:21.040 --> 00:32:24.790 These two groups generally found in business schools, 00:32:24.790 --> 00:32:27.340 started service designed to address 00:32:27.340 --> 00:32:30.460 the idea that the things they're working on, 00:32:30.460 --> 00:32:32.050 the ground that's underneath 00:32:32.050 --> 00:32:33.745 them is fundamentally changing, 00:32:33.745 --> 00:32:36.730 and what's fascinating is this is driven by 00:32:36.730 --> 00:32:40.255 the fact that communication technology changed. 00:32:40.255 --> 00:32:43.690 To some degree, HCI is largely responsible for 00:32:43.690 --> 00:32:47.860 this change in the world that now it needs to react to. 00:32:47.860 --> 00:32:51.280 There's four critical service concepts 00:32:51.280 --> 00:32:52.720 that I think make it particularly 00:32:52.720 --> 00:32:58.040 valuable for these three types of things HCI makes. 00:32:59.280 --> 00:33:04.420 A service concept is the goal of making a new service. 00:33:04.420 --> 00:33:06.640 You're looking for the intersection 00:33:06.640 --> 00:33:09.355 of a customer's need and desire, 00:33:09.355 --> 00:33:12.700 and of a services strategic intent. 00:33:12.700 --> 00:33:13.930 You're looking both ways, 00:33:13.930 --> 00:33:15.850 where can I get value for 00:33:15.850 --> 00:33:18.325 both sides not just value for one side. 00:33:18.325 --> 00:33:21.535 There's the concept of customer competence, 00:33:21.535 --> 00:33:23.620 which means instead of just saying, 00:33:23.620 --> 00:33:25.465 "What can I do for users? 00:33:25.465 --> 00:33:28.315 What can users do that's valuable for me?" 00:33:28.315 --> 00:33:30.790 It's a different way of thinking, 00:33:30.790 --> 00:33:33.850 I just use the example of recapture here. 00:33:33.850 --> 00:33:35.575 I'm guessing most of you know this. 00:33:35.575 --> 00:33:38.920 It's like people have a competency to recognize 00:33:38.920 --> 00:33:40.960 words that weren't scanned well by 00:33:40.960 --> 00:33:43.405 an OCR machine and can't be identified. 00:33:43.405 --> 00:33:46.420 People need a Turing test or people don't need it, 00:33:46.420 --> 00:33:50.050 services need a Turing test to authenticate new accounts. 00:33:50.050 --> 00:33:51.280 We can put these two things 00:33:51.280 --> 00:33:53.530 together and the labor that people are 00:33:53.530 --> 00:33:58.420 already doing can be recaptured to create new value. 00:33:58.420 --> 00:33:59.620 It's a different type of 00:33:59.620 --> 00:34:01.300 discovery that's not going to come out of 00:34:01.300 --> 00:34:05.620 a user-centered design or user experience design process. 00:34:05.620 --> 00:34:08.050 The idea that customers have 00:34:08.050 --> 00:34:10.090 their own customers and that needs to 00:34:10.090 --> 00:34:12.655 be part of the design consideration, 00:34:12.655 --> 00:34:15.880 and then service design is fundamentally systemic. 00:34:15.880 --> 00:34:18.340 It thinks of everything as a system, 00:34:18.340 --> 00:34:20.395 that's how it operates. 00:34:20.395 --> 00:34:22.990 The work is to see a system, 00:34:22.990 --> 00:34:24.475 envision a new system, 00:34:24.475 --> 00:34:26.650 and develop a strategy to get 00:34:26.650 --> 00:34:29.185 from where we are to where we want to be. 00:34:29.185 --> 00:34:31.975 You don't do this strategy in a product design, 00:34:31.975 --> 00:34:33.610 you simply make the new product and 00:34:33.610 --> 00:34:36.040 put it on the shelves for people to buy. 00:34:36.040 --> 00:34:37.840 I just want to break down 00:34:37.840 --> 00:34:41.050 this idea of customers having customers, 00:34:41.050 --> 00:34:43.360 and talk about Google Search, 00:34:43.360 --> 00:34:47.770 I'm guessing everybody's has searched on Google. 00:34:47.770 --> 00:34:49.600 A traditional user-centered design, 00:34:49.600 --> 00:34:51.805 user experience design looks like this. 00:34:51.805 --> 00:34:53.890 You have Google, they're service provider, 00:34:53.890 --> 00:34:55.525 you have the searcher, they're customer, 00:34:55.525 --> 00:34:57.310 and you have the thing the results, 00:34:57.310 --> 00:34:59.214 and it's a one-way flow, 00:34:59.214 --> 00:35:02.485 and it works and everybody knows what their role is. 00:35:02.485 --> 00:35:04.570 From a service design point of view, 00:35:04.570 --> 00:35:06.850 it looks much more complicated. 00:35:06.850 --> 00:35:08.680 You have Google and they're 00:35:08.680 --> 00:35:10.900 definitely a service provider to the searcher, 00:35:10.900 --> 00:35:13.090 they're making those search results which are consumed, 00:35:13.090 --> 00:35:15.190 and that's a very clear relationship. 00:35:15.190 --> 00:35:17.260 But you have ad firms, 00:35:17.260 --> 00:35:21.385 and Google is a service provider to the ad firm also, 00:35:21.385 --> 00:35:27.220 and Google's providing a space 00:35:27.220 --> 00:35:30.250 for the ad firm to put ads. 00:35:30.250 --> 00:35:32.380 The ad firm is a service provider 00:35:32.380 --> 00:35:34.870 to product and service companies. 00:35:34.870 --> 00:35:38.080 The product and service companies hope to be 00:35:38.080 --> 00:35:39.460 a service provider to 00:35:39.460 --> 00:35:43.460 the searchers to have a whole ecology works. 00:35:43.890 --> 00:35:46.855 One of the invisible parts is, 00:35:46.855 --> 00:35:50.140 the searchers is also a service provider to Google, 00:35:50.140 --> 00:35:51.250 because when they search, 00:35:51.250 --> 00:35:53.800 they're providing terms that 00:35:53.800 --> 00:35:55.675 make the search engine better, 00:35:55.675 --> 00:35:57.565 that value can be recaptured. 00:35:57.565 --> 00:35:59.200 What you're seeing is the flow of 00:35:59.200 --> 00:36:01.210 value among all the stakeholders, 00:36:01.210 --> 00:36:04.255 to understand how can I just wrapped it and change it. 00:36:04.255 --> 00:36:06.625 It's going to take you to a really different place, 00:36:06.625 --> 00:36:10.105 but it's not so different from what we already do. 00:36:10.105 --> 00:36:12.760 The work in service design is really 00:36:12.760 --> 00:36:15.460 to represent the service. 00:36:15.460 --> 00:36:20.185 You want to see it, record it, 00:36:20.185 --> 00:36:22.360 and communicate it, and you're capturing 00:36:22.360 --> 00:36:25.195 the relationship between the elements and a service. 00:36:25.195 --> 00:36:28.495 You're capturing how services unfold and change, 00:36:28.495 --> 00:36:30.415 and you're capturing how services 00:36:30.415 --> 00:36:32.500 interact with other services, 00:36:32.500 --> 00:36:35.770 that's flow of value among the stakeholders. 00:36:35.770 --> 00:36:38.649 This is generally expressed in models, 00:36:38.649 --> 00:36:41.095 or shared representations and some of the common one, 00:36:41.095 --> 00:36:43.960 stakeholder model, service blueprints, 00:36:43.960 --> 00:36:47.980 experience journey, and value flow model. 00:36:47.980 --> 00:36:50.050 They're making models, they're working in 00:36:50.050 --> 00:36:52.990 the space of systems design. 00:36:52.990 --> 00:36:56.200 I just took some screen snapshots of 00:36:56.200 --> 00:36:59.605 a website called the Service Design Tools, 00:36:59.605 --> 00:37:01.150 and what I want to point out is, 00:37:01.150 --> 00:37:03.130 they're using HCI tools. 00:37:03.130 --> 00:37:05.619 You will see things like affinity diagram, 00:37:05.619 --> 00:37:08.905 storytelling, role playing, rough prototyping. 00:37:08.905 --> 00:37:11.290 The service design community is 00:37:11.290 --> 00:37:14.515 simply cherry picking methods from HCI, 00:37:14.515 --> 00:37:17.410 ignoring any revenue that might go to 00:37:17.410 --> 00:37:20.725 people that are doing HCI and moving into this space. 00:37:20.725 --> 00:37:22.720 They're envisioning space, they're 00:37:22.720 --> 00:37:25.180 using tools like story-boarding, 00:37:25.180 --> 00:37:28.960 mood boards, mock-up, experience, prototype, 00:37:28.960 --> 00:37:31.210 personas, things that are super familiar 00:37:31.210 --> 00:37:34.269 with common HCI practice everyday, 00:37:34.269 --> 00:37:39.490 but then adding to it things like activity maps, 00:37:39.490 --> 00:37:42.790 service blueprints, customer journey maps. 00:37:42.790 --> 00:37:46.780 They are building on an HCI knowledge set, 00:37:46.780 --> 00:37:48.430 but extending it to really work in 00:37:48.430 --> 00:37:50.170 the space of services and not giving 00:37:50.170 --> 00:37:51.850 any credit to the people 00:37:51.850 --> 00:37:54.770 that did the work to develop these things. 00:37:54.840 --> 00:37:59.680 Service design provides a number of benefits. 00:37:59.680 --> 00:38:02.410 One of the most important is that it's systemic, 00:38:02.410 --> 00:38:05.350 and it lets a design team really think 00:38:05.350 --> 00:38:09.145 about the needs of multiple stakeholders, 00:38:09.145 --> 00:38:11.920 and the role switching. 00:38:11.920 --> 00:38:14.365 How you go from being a customer 00:38:14.365 --> 00:38:17.275 at one to being a service provider and another, 00:38:17.275 --> 00:38:19.240 that's easy to capture so that 00:38:19.240 --> 00:38:24.700 your design solution can address that role switch. 00:38:24.700 --> 00:38:27.940 It investigates what customers can do 00:38:27.940 --> 00:38:30.280 that are valuable for a service, 00:38:30.280 --> 00:38:31.900 allowing an entirely new way 00:38:31.900 --> 00:38:33.400 to think of value production, 00:38:33.400 --> 00:38:34.660 that's not simply selling 00:38:34.660 --> 00:38:35.920 more and more things. 00:38:35.920 --> 00:38:43.105 It designs what the service agent does, 00:38:43.105 --> 00:38:44.815 not just the tools. 00:38:44.815 --> 00:38:46.810 This is a benefit, 00:38:46.810 --> 00:38:50.325 but it's a really hard one to cross for HCI. 00:38:50.325 --> 00:38:52.830 It's designing a job and it's 00:38:52.830 --> 00:38:56.445 designing the steps someone does in their job, 00:38:56.445 --> 00:38:59.055 not just the screens that they use, 00:38:59.055 --> 00:39:03.170 which starts to feel very much not like HCI. 00:39:03.170 --> 00:39:06.805 It definitely crosses the physical digital divide, 00:39:06.805 --> 00:39:08.965 so service design has to do with 00:39:08.965 --> 00:39:11.890 choosing the drapes in a hotel room. 00:39:11.890 --> 00:39:13.780 That seems very far from 00:39:13.780 --> 00:39:16.210 interacting with a computer, but holistically, 00:39:16.210 --> 00:39:18.910 it's all part of a customer's experience and the way that 00:39:18.910 --> 00:39:22.180 they're engaging with that service. 00:39:22.180 --> 00:39:26.620 It plays with business models as a design material. 00:39:26.620 --> 00:39:28.690 It's just part of the consideration. 00:39:28.690 --> 00:39:29.770 Don't make the thing and then 00:39:29.770 --> 00:39:31.495 figure out how to pay for it. 00:39:31.495 --> 00:39:34.960 Figuring out how it's paid for is simply a part of 00:39:34.960 --> 00:39:36.250 the making and it's part of 00:39:36.250 --> 00:39:39.620 the dialogue to figure out the better future. 00:39:41.700 --> 00:39:44.860 Finally, one of the best values 00:39:44.860 --> 00:39:47.190 that supplies this idea of coming 00:39:47.190 --> 00:39:49.380 up with a plan of getting 00:39:49.380 --> 00:39:51.960 from the current state to the future state, 00:39:51.960 --> 00:39:55.545 not just envisioning the thing that exists in the future. 00:39:55.545 --> 00:40:00.415 The plan really helps make it operationalizable. 00:40:00.415 --> 00:40:03.890 There are some limitations to this. 00:40:05.220 --> 00:40:09.219 Service design is really oriented strongly, 00:40:09.219 --> 00:40:13.240 oriented towards redesigning an innovation 00:40:13.240 --> 00:40:18.325 of existing services and very traditional services. 00:40:18.325 --> 00:40:20.980 It is a process that 00:40:20.980 --> 00:40:24.010 developed around looking at banking and finance, 00:40:24.010 --> 00:40:25.750 looking at health care, 00:40:25.750 --> 00:40:28.420 looking at hospitality like hotels, 00:40:28.420 --> 00:40:31.525 very big traditional services. 00:40:31.525 --> 00:40:35.590 It has a very slow plotting pace, 00:40:35.590 --> 00:40:36.940 because of the scale of 00:40:36.940 --> 00:40:38.845 things that it generally works on. 00:40:38.845 --> 00:40:42.129 It has no sense of creating new services 00:40:42.129 --> 00:40:46.910 of radical innovation and invention. 00:40:47.730 --> 00:40:52.045 HCI employs a really simplistic idea of user, 00:40:52.045 --> 00:40:54.340 slightly better with service design, 00:40:54.340 --> 00:40:56.530 but it's still fairly simplistic to think only of 00:40:56.530 --> 00:40:59.275 customers and service providers. 00:40:59.275 --> 00:41:01.960 There's an opportunity for 00:41:01.960 --> 00:41:04.390 work that can begin to address it and come up 00:41:04.390 --> 00:41:09.835 with a richer set of rules for entities to play. 00:41:09.835 --> 00:41:13.165 Service design, if you look at how it's really done, 00:41:13.165 --> 00:41:17.815 there's very much a timidness with using new technology. 00:41:17.815 --> 00:41:20.275 They just want to use stuff they know works, 00:41:20.275 --> 00:41:23.440 and there's no sense of technology as invention. 00:41:23.440 --> 00:41:25.960 With HCI moving, if we 00:41:25.960 --> 00:41:28.390 decide to move into service design, 00:41:28.390 --> 00:41:30.175 I think this is a real benefit. 00:41:30.175 --> 00:41:32.620 The sense of radical invention 00:41:32.620 --> 00:41:34.195 and the creation of the new, 00:41:34.195 --> 00:41:38.300 would push service design in an entirely new direction. 00:41:38.760 --> 00:41:42.385 What I see is these current limitations 00:41:42.385 --> 00:41:45.370 are benefits for HCI. 00:41:45.370 --> 00:41:48.160 It's a chance to take a step towards 00:41:48.160 --> 00:41:51.475 service design and actually own a big chunk of it, 00:41:51.475 --> 00:41:53.230 because it's the space that 00:41:53.230 --> 00:41:55.390 HCI is very comfortable working in that 00:41:55.390 --> 00:41:58.180 current services and service design 00:41:58.180 --> 00:42:01.780 are in no way ready to take on. 00:42:01.780 --> 00:42:04.750 It's great for me to stand here and say, yes, 00:42:04.750 --> 00:42:05.890 go do service design, 00:42:05.890 --> 00:42:08.380 but that's not really giving a plan. 00:42:08.380 --> 00:42:09.550 If I want to say, I'm doing 00:42:09.550 --> 00:42:10.780 service design, I need a plan, 00:42:10.780 --> 00:42:14.680 how do we get from here to there? 00:42:14.680 --> 00:42:18.565 I'm going to suggest a first small step 00:42:18.565 --> 00:42:22.405 is to simply teach service designed to HCI students. 00:42:22.405 --> 00:42:25.975 Replace current user-centered design 00:42:25.975 --> 00:42:27.775 and user experience design 00:42:27.775 --> 00:42:30.550 that's being taught with the service design. 00:42:30.550 --> 00:42:32.500 Which doesn't actually get rid of those things, 00:42:32.500 --> 00:42:33.610 there's still a core part, 00:42:33.610 --> 00:42:37.960 but it adds a few additional dimensions on top of that. 00:42:37.960 --> 00:42:41.650 Add simple things like making a stakeholder map, 00:42:41.650 --> 00:42:45.760 doing a service blueprint as a process of both 00:42:45.760 --> 00:42:47.710 processing the current state of the world 00:42:47.710 --> 00:42:51.080 and envisioning a future. 00:42:51.810 --> 00:42:56.710 By just teaching service design as a part of HCI, 00:42:56.710 --> 00:42:59.980 the researchers doing HCI will naturally begin to see 00:42:59.980 --> 00:43:04.000 opportunities where HCI research can have a benefit. 00:43:04.000 --> 00:43:05.980 I don't have a big research agenda, 00:43:05.980 --> 00:43:07.690 but I know it will emerge in 00:43:07.690 --> 00:43:10.015 the same way it did for usability, 00:43:10.015 --> 00:43:12.040 the same way it did for user-centered design and 00:43:12.040 --> 00:43:15.310 the same way that it did for user experience design. 00:43:15.310 --> 00:43:18.010 I want to stress it's not at all 00:43:18.010 --> 00:43:20.350 about abandoning the current lenses, 00:43:20.350 --> 00:43:23.200 the current thing we consider core to HCI. 00:43:23.200 --> 00:43:24.700 Right now we teach 00:43:24.700 --> 00:43:27.819 students that are training to be practitioners, 00:43:27.819 --> 00:43:30.565 to focus on the interaction, 00:43:30.565 --> 00:43:33.400 focus on the context and understand those 00:43:33.400 --> 00:43:37.180 together and you can propose solutions that will work. 00:43:37.180 --> 00:43:40.450 We're just adding an additional dimension of saying, 00:43:40.450 --> 00:43:41.860 and you also need to consider 00:43:41.860 --> 00:43:44.365 the larger system this exist in. 00:43:44.365 --> 00:43:47.590 The easiest way that I do this with my students, 00:43:47.590 --> 00:43:50.605 is I say, how does this make money? 00:43:50.605 --> 00:43:52.435 How is it profitable? 00:43:52.435 --> 00:43:56.545 But how does your idea simply make enough money to exist? 00:43:56.545 --> 00:43:58.450 It's interesting like the feedback 00:43:58.450 --> 00:44:00.055 I often get back from my students. 00:44:00.055 --> 00:44:03.235 You mean it's not a good idea if it doesn't make money? 00:44:03.235 --> 00:44:06.160 It's like, yeah, why would you want to spend 00:44:06.160 --> 00:44:09.685 your time designing things that couldn't actually exist? 00:44:09.685 --> 00:44:12.010 They're not actually solving the problem, 00:44:12.010 --> 00:44:15.190 but it's a hard step for them to get over. 00:44:15.190 --> 00:44:18.070 Most of them are either coming from engineering school or 00:44:18.070 --> 00:44:19.930 design school and this 00:44:19.930 --> 00:44:21.340 conflicts with what they 00:44:21.340 --> 00:44:23.870 think it is they should be doing. 00:44:24.000 --> 00:44:26.920 Another simple example of how I think 00:44:26.920 --> 00:44:28.870 this can be integrated into teaching. 00:44:28.870 --> 00:44:33.475 [NOISE] A very common teaching experience 00:44:33.475 --> 00:44:35.530 in teaching HCI is asking 00:44:35.530 --> 00:44:38.650 students to design a mobile app. 00:44:38.650 --> 00:44:41.260 Well, a mobile app is for someone, 00:44:41.260 --> 00:44:43.690 there's a service that's providing it. 00:44:43.690 --> 00:44:47.350 Can you also design a dashboard that the service 00:44:47.350 --> 00:44:51.220 is using to understand how their app is being used. 00:44:51.220 --> 00:44:52.840 Can you think of it from 00:44:52.840 --> 00:44:54.940 the other perspective and 00:44:54.940 --> 00:44:56.815 does that change the way this works? 00:44:56.815 --> 00:44:58.945 It's capturing the information 00:44:58.945 --> 00:45:01.090 needed to produce the dashboard, 00:45:01.090 --> 00:45:04.000 so you're beginning to think more of a system, 00:45:04.000 --> 00:45:08.920 more of an ecology and it's a slight change. 00:45:08.920 --> 00:45:10.150 I'm just going to alter 00:45:10.150 --> 00:45:12.070 this assignment a little bit by saying, 00:45:12.070 --> 00:45:14.545 in addition, make a dashboard. 00:45:14.545 --> 00:45:16.180 Small steps, I think, 00:45:16.180 --> 00:45:17.950 move us in this direction. 00:45:17.950 --> 00:45:20.005 So why change? 00:45:20.005 --> 00:45:22.780 I don't think there's a major harm if we don't switch. 00:45:22.780 --> 00:45:24.340 I want to stress, I'm not saying it's 00:45:24.340 --> 00:45:26.170 going to be the end of HCI at all. 00:45:26.170 --> 00:45:29.020 We can be just fine if we don't embrace service design, 00:45:29.020 --> 00:45:31.495 but I think we can be better if we do. 00:45:31.495 --> 00:45:33.865 I think it increases 00:45:33.865 --> 00:45:37.630 the influence of HCI in the enterprise. 00:45:37.630 --> 00:45:40.270 Right now, our students go out and they practice 00:45:40.270 --> 00:45:41.710 HCI in companies and they don't 00:45:41.710 --> 00:45:43.390 get listened to very much, 00:45:43.390 --> 00:45:45.880 because it's not clear how what they 00:45:45.880 --> 00:45:49.120 do adds value to the company. 00:45:49.120 --> 00:45:52.885 By taking over the financial model, 00:45:52.885 --> 00:45:55.690 you move the HCI work much more 00:45:55.690 --> 00:45:58.810 centrally into what companies do. 00:45:58.810 --> 00:46:02.080 It opens up new spaces for research 00:46:02.080 --> 00:46:03.520 without getting rid of any of 00:46:03.520 --> 00:46:06.325 the old ones so it expands it. 00:46:06.325 --> 00:46:10.150 It helps better integrate current threads of research, 00:46:10.150 --> 00:46:12.040 particularly in social computing, 00:46:12.040 --> 00:46:13.840 which are often around, 00:46:13.840 --> 00:46:16.360 how do we motivate people to do something 00:46:16.360 --> 00:46:20.300 with rewards or intrinsic motivation? 00:46:20.760 --> 00:46:24.490 It stops a potential territorial war 00:46:24.490 --> 00:46:26.620 that's going to happen between 00:46:26.620 --> 00:46:28.735 service design in business schools and 00:46:28.735 --> 00:46:32.290 HCI and information and computer science programs, 00:46:32.290 --> 00:46:34.510 as the territory that they're working on 00:46:34.510 --> 00:46:37.555 has an increasing overlap. 00:46:37.555 --> 00:46:40.180 There are definitely risks. 00:46:40.180 --> 00:46:42.160 It means going well beyond 00:46:42.160 --> 00:46:45.475 the interaction with computers and obviously, 00:46:45.475 --> 00:46:48.280 human computer interaction that computer is 00:46:48.280 --> 00:46:51.580 core to the name of HCI. 00:46:51.580 --> 00:46:53.500 It's not without risks to say, 00:46:53.500 --> 00:46:56.200 do we want to go past just thinking 00:46:56.200 --> 00:46:59.980 about the thing and the person at the thing, 00:46:59.980 --> 00:47:03.055 to think of all of the other forces that then 00:47:03.055 --> 00:47:07.210 play at how this interaction takes place. 00:47:07.210 --> 00:47:09.250 It means thinking about things like 00:47:09.250 --> 00:47:11.830 print design, media campaigns, 00:47:11.830 --> 00:47:14.335 the environments people are in and is that 00:47:14.335 --> 00:47:17.950 a step too far for our field? 00:47:17.950 --> 00:47:22.630 It means designing work and workers and jobs, 00:47:22.630 --> 00:47:25.075 not just the tools that they use. 00:47:25.075 --> 00:47:28.120 It means thinking about money and 00:47:28.120 --> 00:47:32.830 motivation as a core way that we work. 00:47:32.830 --> 00:47:36.220 That's all for me. Go service design, 00:47:36.220 --> 00:47:37.360 I'm very curious to hear 00:47:37.360 --> 00:47:39.220 your reaction and if 00:47:39.220 --> 00:47:41.110 you think this is something HCI should do. 00:47:41.110 --> 00:47:51.010 [APPLAUSE] 00:47:51.010 --> 00:47:53.290 Is there any particular reason you didn't slide 00:47:53.290 --> 00:47:55.480 the name system design as 00:47:55.480 --> 00:47:57.190 opposed or why you 00:47:57.190 --> 00:47:59.080 chose service design as opposed to system. 00:47:59.080 --> 00:47:59.445 Yes. 00:47:59.445 --> 00:48:02.575 Yes. There are many reasons. 00:48:02.575 --> 00:48:06.805 System design has a terrible history, 00:48:06.805 --> 00:48:10.930 and I'm going to make a strong link of system design, 00:48:10.930 --> 00:48:14.920 particularly to the design 00:48:14.920 --> 00:48:21.265 of inner-city housing towers for low-income people. 00:48:21.265 --> 00:48:24.580 It's a very top-down view 00:48:24.580 --> 00:48:28.645 that repeatedly failed and did a lot of societal harm. 00:48:28.645 --> 00:48:32.050 So I don't want to grab on to that legacy. 00:48:32.050 --> 00:48:34.900 Systems design and systems theory, if you read it, 00:48:34.900 --> 00:48:37.090 is really abstract and 00:48:37.090 --> 00:48:39.790 complex and it's hard to make it actionable, 00:48:39.790 --> 00:48:41.995 particularly to my students. 00:48:41.995 --> 00:48:45.250 We are already designing services even though we're 00:48:45.250 --> 00:48:48.310 not following a service design process. 00:48:48.310 --> 00:48:49.510 It's like, to me, 00:48:49.510 --> 00:48:51.640 it's the next step, 00:48:51.640 --> 00:48:55.270 but I'm not saying that that isn't the end goal. 00:48:55.270 --> 00:48:56.710 I'm just trying to say what's 00:48:56.710 --> 00:48:59.620 our next reasonable step and how far can 00:48:59.620 --> 00:49:02.260 I take the students 00:49:02.260 --> 00:49:05.590 that I'm working with. How far will they go? 00:49:05.590 --> 00:49:08.140 It feels like it's approaching everything 00:49:08.140 --> 00:49:10.150 as a holistic [inaudible]. 00:49:10.150 --> 00:49:10.720 Yes. 00:49:10.720 --> 00:49:14.815 So systems are supposed to service. 00:49:14.815 --> 00:49:19.870 Service is a very small subset of systems, 00:49:19.870 --> 00:49:21.280 but it gives us a framing, 00:49:21.280 --> 00:49:22.450 there's already a method. 00:49:22.450 --> 00:49:23.845 People already know how to do it. 00:49:23.845 --> 00:49:26.740 In the same way like user experience design seem to 00:49:26.740 --> 00:49:32.110 radical change from the work focus of HCI, 00:49:32.110 --> 00:49:35.110 that style of design already existed. 00:49:35.110 --> 00:49:36.760 It happened in multimedia and film, 00:49:36.760 --> 00:49:38.230 it happened in industrial design, 00:49:38.230 --> 00:49:39.820 it happened in graphic design. 00:49:39.820 --> 00:49:41.965 So they were, in the same way, 00:49:41.965 --> 00:49:44.380 pulling from those and creating something like how 00:49:44.380 --> 00:49:46.885 do we integrate that to the design of technology. 00:49:46.885 --> 00:49:48.700 I'm trying to say I do the same thing. 00:49:48.700 --> 00:49:50.065 Service design already exists. 00:49:50.065 --> 00:49:51.310 People don't really know how to do 00:49:51.310 --> 00:49:54.310 systems design effectively, 00:49:54.310 --> 00:49:56.800 but they know how to design services effectively. 00:49:56.800 --> 00:49:59.664 So let's pull in what we know works, 00:49:59.664 --> 00:50:00.880 we can make actionable, 00:50:00.880 --> 00:50:02.305 we can figure out how it fits, 00:50:02.305 --> 00:50:04.705 and then bring our invention skills, 00:50:04.705 --> 00:50:08.065 our radical technology transformation skills 00:50:08.065 --> 00:50:11.905 to take service design in a different direction. 00:50:11.905 --> 00:50:15.850 Systems design, it just seems too big and dangerous, 00:50:15.850 --> 00:50:18.920 but I might just be being too timid. 00:50:21.330 --> 00:50:25.930 As you said, UCD and HCI, 00:50:25.930 --> 00:50:29.900 people often end up [inaudible] company. 00:50:31.020 --> 00:50:34.285 This design is more wholistic, 00:50:34.285 --> 00:50:37.059 multi-stakeholder approach. 00:50:37.059 --> 00:50:39.175 How do you protect 00:50:39.175 --> 00:50:42.610 the smallest stakeholders when there's going to 00:50:42.610 --> 00:50:45.145 be a lot of weight on the side of let's 00:50:45.145 --> 00:50:48.310 look at your search engine model, your Google model? 00:50:48.310 --> 00:50:52.315 The advertisers, the sellers, Google itself, 00:50:52.315 --> 00:50:53.650 they've got a lot more weight than 00:50:53.650 --> 00:50:56.530 any individual user in that equation. 00:50:56.530 --> 00:51:04.120 So how do you protect the end-user in 00:51:04.120 --> 00:51:08.830 that equation by prioritizing their needs without just 00:51:08.830 --> 00:51:10.570 getting swamped by the needs 00:51:10.570 --> 00:51:14.395 of the deep-pocket stakeholders? 00:51:14.395 --> 00:51:18.040 The whole ecology doesn't work if 00:51:18.040 --> 00:51:20.169 every entity isn't receiving 00:51:20.169 --> 00:51:23.395 more value than they feel they're giving up. 00:51:23.395 --> 00:51:28.990 Right now, if I'm a UX designer at Google, 00:51:28.990 --> 00:51:31.285 I'm really focused on 00:51:31.285 --> 00:51:34.090 saying we need to focus on user experience because that's 00:51:34.090 --> 00:51:36.700 going to be good for Google without actually being able 00:51:36.700 --> 00:51:40.585 to demonstrate how is it good for Google. 00:51:40.585 --> 00:51:44.380 When you talk to design teams, 00:51:44.380 --> 00:51:47.275 and I've talked to a lot of UX design teams, 00:51:47.275 --> 00:51:48.820 the first question I always ask them is, 00:51:48.820 --> 00:51:50.815 how does your company make money? 00:51:50.815 --> 00:51:54.205 Most of them have no idea how their company makes money. 00:51:54.205 --> 00:51:56.950 Then it becomes very hard for them to 00:51:56.950 --> 00:51:59.530 fight for what users needs because they 00:51:59.530 --> 00:52:01.180 can't put it into a frame of 00:52:01.180 --> 00:52:04.930 reference that makes sense for the customer, 00:52:04.930 --> 00:52:06.805 in a sense, that they're surveying. 00:52:06.805 --> 00:52:10.690 I actually think we will do more for users by 00:52:10.690 --> 00:52:12.970 understanding the value that's produced by 00:52:12.970 --> 00:52:15.490 all of the entities and simply saying, 00:52:15.490 --> 00:52:17.770 but I just want to help these people, 00:52:17.770 --> 00:52:21.010 without being able to articulate why that might be 00:52:21.010 --> 00:52:25.000 good other than just for altruistic reasons; 00:52:25.000 --> 00:52:27.670 isn't effective and it's been shown to 00:52:27.670 --> 00:52:30.340 repeatedly not be effective. 00:52:30.340 --> 00:52:32.755 Google is a really interesting example 00:52:32.755 --> 00:52:34.510 of why user-centered design, 00:52:34.510 --> 00:52:36.850 user experience design doesn't work. 00:52:36.850 --> 00:52:40.060 The design team's goal is to discover 00:52:40.060 --> 00:52:41.740 something 100 million people 00:52:41.740 --> 00:52:44.020 want to do every day for free. 00:52:44.020 --> 00:52:46.270 Doing ethnographic fieldwork is 00:52:46.270 --> 00:52:47.545 not going to get me there. 00:52:47.545 --> 00:52:51.315 I can't go watch 100 million people and say, 00:52:51.315 --> 00:52:53.590 ''Oh, they all want to do this.'' 00:52:53.790 --> 00:52:56.650 We need different methods if we're going to 00:52:56.650 --> 00:52:58.660 try to design those kinds of things 00:52:58.660 --> 00:53:02.785 and not just have them randomly appear by fluke. 00:53:02.785 --> 00:53:06.770 Can we more intentionally create those kinds of things? 00:53:07.860 --> 00:53:11.155 Other questions? Yeah. 00:53:11.155 --> 00:53:13.750 IT service management has been 00:53:13.750 --> 00:53:16.600 promoting this service design for a little while, 00:53:16.600 --> 00:53:17.830 how do you think about those? 00:53:17.830 --> 00:53:24.960 You are beyond the scope of what I know. 00:53:24.960 --> 00:53:26.775 I'm not familiar with 00:53:26.775 --> 00:53:29.534 IT service management and how they'd been engaging 00:53:29.534 --> 00:53:33.730 in service design, but to me, 00:53:33.730 --> 00:53:35.995 it's not surprising that anybody that's 00:53:35.995 --> 00:53:38.575 at the intersection of service and 00:53:38.575 --> 00:53:41.620 electronic tools isn't running into this because of 00:53:41.620 --> 00:53:45.265 this transition to software as a service. 00:53:45.265 --> 00:53:48.070 We just need to think in a different way. 00:53:48.070 --> 00:53:51.880 It makes sense to me thinking more like IT within 00:53:51.880 --> 00:53:54.010 a company generally needed 00:53:54.010 --> 00:53:55.900 to do it as soon as people started to say, 00:53:55.900 --> 00:53:58.825 well, how does the IT group generate revenue? 00:53:58.825 --> 00:54:01.000 What's the value that you're adding? 00:54:01.000 --> 00:54:02.380 To some degree, I think they've been 00:54:02.380 --> 00:54:03.700 thinking about it longer, 00:54:03.700 --> 00:54:09.295 but I'm not familiar with the literature at all. Yeah. 00:54:09.295 --> 00:54:16.000 How does a silo UX team advocate for service design? 00:54:16.000 --> 00:54:20.110 Does this have to come [inaudible]? 00:54:20.110 --> 00:54:21.175 Excellent question. 00:54:21.175 --> 00:54:23.920 I think they do it in the exact same way that 00:54:23.920 --> 00:54:28.280 user experience design and user-centered design started. 00:54:28.470 --> 00:54:32.740 I have a number of former students at Adaptive Path, 00:54:32.740 --> 00:54:34.480 which is no longer, 00:54:34.480 --> 00:54:37.240 but they just got purchased by Capital One. 00:54:37.240 --> 00:54:40.660 But the way they did it was they would get hired to do 00:54:40.660 --> 00:54:42.820 user experience design and they would 00:54:42.820 --> 00:54:45.850 do service design and not tell their client. 00:54:45.850 --> 00:54:48.160 Then you're producing these additional 00:54:48.160 --> 00:54:49.240 artifacts and people are like, 00:54:49.240 --> 00:54:50.619 wow, that's really valuable, 00:54:50.619 --> 00:54:52.420 and then suddenly you're doing it. 00:54:52.420 --> 00:54:54.385 You don't ask for permission, 00:54:54.385 --> 00:54:56.995 just do it, build it. 00:54:56.995 --> 00:54:59.350 You don't say, can we have permission to 00:54:59.350 --> 00:55:02.125 think about the financial model. Just do it. 00:55:02.125 --> 00:55:05.635 Produce the artifacts, call yourselves whatever, 00:55:05.635 --> 00:55:07.855 the name doesn't matter. 00:55:07.855 --> 00:55:09.610 But by producing the different thing, 00:55:09.610 --> 00:55:11.080 the value that you're producing 00:55:11.080 --> 00:55:13.675 becomes immediately apparent. 00:55:13.675 --> 00:55:16.945 That's been super successful at Adaptive Path 00:55:16.945 --> 00:55:20.995 with companies that weren't ready to go there. 00:55:20.995 --> 00:55:24.910 You ease the people that you're working with into it. 00:55:24.910 --> 00:55:27.280 That's also like IDEO Talks. 00:55:27.280 --> 00:55:30.130 Traditionally, we don't let 00:55:30.130 --> 00:55:33.490 our clients into the back until they've worked with us. 00:55:33.490 --> 00:55:36.310 We stage bringing them 00:55:36.310 --> 00:55:38.620 in because they're not ready to think this is valuable, 00:55:38.620 --> 00:55:40.885 they just think it's craziness. 00:55:40.885 --> 00:55:45.010 To me, it's like let's follow the same pattern that 00:55:45.010 --> 00:55:48.775 we've used traditionally in HCI to expand our definition. 00:55:48.775 --> 00:55:52.010 I'm just pointing let's go this way. 00:55:55.500 --> 00:55:59.545 In academia, we have the fortunate thing of 00:55:59.545 --> 00:56:01.900 department and discipline that 00:56:01.900 --> 00:56:04.660 puts a really big wall between. 00:56:04.660 --> 00:56:07.510 What one or two courses would you 00:56:07.510 --> 00:56:10.600 recommend for a junior/senior in 00:56:10.600 --> 00:56:15.190 computing to open their minds to this so that they 00:56:15.190 --> 00:56:20.260 don't graduate to [inaudible]? 00:56:20.260 --> 00:56:22.855 I teach a course with Jim Morris. 00:56:22.855 --> 00:56:24.910 I don't know if any of you guys know Jim Morris, 00:56:24.910 --> 00:56:28.015 he's an old-timer in transactional computing, 00:56:28.015 --> 00:56:29.515 former dean of 00:56:29.515 --> 00:56:32.350 the School of Computer Science at Carnegie Mellon. 00:56:32.350 --> 00:56:35.425 We teach a course on mobile service innovation, 00:56:35.425 --> 00:56:37.480 and I think it's a nice model 00:56:37.480 --> 00:56:39.820 for getting students just to think about it, 00:56:39.820 --> 00:56:42.070 where we say to the teams, 00:56:42.070 --> 00:56:46.045 you have to come up with a business model, 00:56:46.045 --> 00:56:48.130 you have to come up with a justification 00:56:48.130 --> 00:56:49.990 for desirability. 00:56:49.990 --> 00:56:51.565 Who wants to do this? 00:56:51.565 --> 00:56:52.900 You have to come up with 00:56:52.900 --> 00:56:57.490 a software architecture simultaneously, not sequentially. 00:56:57.490 --> 00:57:01.270 We do mobile service because mobile services are tiny, 00:57:01.270 --> 00:57:03.670 so it's something like a small team could do, 00:57:03.670 --> 00:57:06.850 and say think like a startup. 00:57:06.850 --> 00:57:09.760 What can you make that people want? 00:57:09.760 --> 00:57:12.850 Then we encourage the students to pivot. 00:57:12.850 --> 00:57:15.490 One great example that worked really well, 00:57:15.490 --> 00:57:18.460 we had a team, they wanted to do translation, 00:57:18.460 --> 00:57:21.220 so I'm an international traveler and I see 00:57:21.220 --> 00:57:22.810 a sign and I don't know what it means and I can 00:57:22.810 --> 00:57:24.895 point my phone and translate it. 00:57:24.895 --> 00:57:27.640 So as they're investigating this, 00:57:27.640 --> 00:57:28.810 they quickly learned nobody 00:57:28.810 --> 00:57:30.220 uses their phones when they travel 00:57:30.220 --> 00:57:31.840 internationally because it's ridiculously 00:57:31.840 --> 00:57:33.895 expensive to use international data. 00:57:33.895 --> 00:57:36.295 It's like, boom, idea's dead. 00:57:36.295 --> 00:57:39.550 But it's like follow the money, think differently. 00:57:39.550 --> 00:57:41.755 Don't think of the money later. 00:57:41.755 --> 00:57:43.030 Yes, people want this, 00:57:43.030 --> 00:57:46.060 but people don't want it enough to pay for it. 00:57:46.060 --> 00:57:47.860 So by following the money, 00:57:47.860 --> 00:57:50.020 what they eventually came up with, 00:57:50.020 --> 00:57:52.270 their end design was a vending machine, 00:57:52.270 --> 00:57:53.920 which is a really funny outcome for 00:57:53.920 --> 00:57:55.870 a mobile service design class. 00:57:55.870 --> 00:57:58.450 But it was going to vend Mi-Fis at 00:57:58.450 --> 00:57:59.860 international airports so when you 00:57:59.860 --> 00:58:02.065 landed you could pick up a Mi-Fi. 00:58:02.065 --> 00:58:03.880 The team said we're going to buy 00:58:03.880 --> 00:58:06.955 two-year contracts at about $30 a month, 00:58:06.955 --> 00:58:09.010 and we're going to sell them three dollars a day. 00:58:09.010 --> 00:58:12.880 So we're getting 200 percent markup on our product, 00:58:12.880 --> 00:58:14.455 and for three dollars a day, 00:58:14.455 --> 00:58:16.060 you're getting Internet for all 00:58:16.060 --> 00:58:17.590 of your devices everywhere you go. 00:58:17.590 --> 00:58:19.900 Look at hotels, they're charging $10 or $15, 00:58:19.900 --> 00:58:22.430 and it only works in a hotel. 00:58:23.040 --> 00:58:27.025 It's like the business part of it is not hard, 00:58:27.025 --> 00:58:29.290 and the business people can't do it because they 00:58:29.290 --> 00:58:32.890 don't actually understand the technology, 00:58:32.890 --> 00:58:35.860 and so their idea was like Redbox. 00:58:35.860 --> 00:58:37.855 These devices were just sitting in there, 00:58:37.855 --> 00:58:39.070 you charging and you just get 00:58:39.070 --> 00:58:40.120 one when you get to the airport 00:58:40.120 --> 00:58:42.475 and you drop it off on your flight out. 00:58:42.475 --> 00:58:45.940 It was like that's the thinking that I want 00:58:45.940 --> 00:58:47.560 you to incorporate in 00:58:47.560 --> 00:58:49.900 the way you think about computation. 00:58:49.900 --> 00:58:52.030 We had a lot of really horrible ideas, 00:58:52.030 --> 00:58:52.660 don't get me wrong [LAUGHTER]. 00:58:52.660 --> 00:58:54.820 They don't all get there. 00:58:54.820 --> 00:58:56.770 But sometimes you see these things come 00:58:56.770 --> 00:58:59.005 up and it's like that's a startup idea. 00:58:59.005 --> 00:59:00.730 Not that I need you to go do that, 00:59:00.730 --> 00:59:02.875 but I want you to be able to think like that. 00:59:02.875 --> 00:59:04.330 The students immediately saw 00:59:04.330 --> 00:59:06.250 the value of the service approach 00:59:06.250 --> 00:59:09.100 over pre-select the users 00:59:09.100 --> 00:59:12.230 you think matter and go find out about them. 00:59:14.610 --> 00:59:16.705 Any other questions? 00:59:16.705 --> 00:59:18.010 Concerns? No anger? 00:59:18.010 --> 00:59:20.380 I was looking for pushback. 00:59:20.380 --> 00:59:29.000 [BACKGROUND] [APPLAUSE]