 Anyone new on the design track? Has anyone joined the first session here? Okay, cool. Welcome. It's a cool room. I like to think of it as the decompression space, in case of not going to the other one. So when I talk about psychology of UX and adoption, I'm a psychologist, not a couch type, but I study psychology and interaction design. And when I was doing that, I learned a lot about models, like frameworks of what drives people to adopt new technologies. And in preparing for this talk, like over the last weeks, I've been trying to apply that to our space to see how does it translate all these frameworks. Yeah, so that's what I want to talk to you about, but first, I want to talk to you about porn. And why that's interesting will become a bit more clear. I hope so, later on. This magazine, Time Magazine, in 1992 declared hybrid porn specifically as one of the bigger inventions of the past millennium. Doesn't seem like a big deal. It was a big deal in agriculture, so it was a big deal for food around the world. But it was also a big deal for understanding what makes people adopt new technologies. Side note, but what's interesting, this is 1992, a special fall edition. A lot of the content is sponsored by IBM. There's a lot of talk about network computers. There's no word on predictions of the internet, as we know it now, and all the services that we use on it. Even though this was 1992, so we would expect that to be kind of ubiquitous knowledge already. So the story about porn, we think it's some days here to check. Porn was invented, hybrid porn specifically, not porn, hybrid porn was invented in 1928 by scientists of Iowa State University. And then years past, it was released to farmers, and it had a lot of benefits to speak of. So the production rate would go up by 20%. Farmers could move to mechanical ways of production, because the way that porn grew had changed the size or the equal height of the stem, so it made it easier to cultivate. It was also much more resistant to drop. So there were a lot of functional benefits that would make farmers be like, okay, let's all switch to hybrid porn. And then next to that, there was an alliance of farmers that also adopted this technology from the university and started pushing it and promoting it. So there was a lot going for this new type of cultivating porn. One of the downsides, obviously there were downsides, was that farmers would need to buy their seeds every year. So they needed to change something in their behavior compared to what they were accustomed to. And still, 13 years later, there was a team of sociologists that thought, oh, wow, this porn business seems to be quite interesting. There's a lot of changes happening there. And so they wanted to figure out why, what made some farmers decide to switch to hybrid porn, others not yet, so it was kind of a big deal in that state at the time. So they asked, they sent out a survey to farmers in the state asking, why and when did you switch to this new type of cultivation? And what they found was that within five years, 10% had switched to this new type of production. The three years after that, 40% more people were farmers switched to that new type of production. No Excel yet, but they plotted the data. And then they found this, which if you've looked at, like if you ever Google, it's like technology reduction, you basically always find this S curve. Okay, there's like a first 10%, and then at some point, it kind of grows a little bit more, and then there's this switch, this critical point that we call, like now there's critical mass. And this is what we all strive for, because at this critical mass, we'll get there. That's when mass adoption will happen. And then this is a guy who kind of summarized that research and followed up by doing loads of research after in other fields to understand how does this translate. Is this S curve something that occurs in every new technology that we introduce? His conclusion was yes, like he actually drew up that S curve with some variations of different technologies. Interesting motivation for him to start looking into that was that he always saw his dad as a farmer kind of wait till the last minute when everyone was starting to use something new and his uncle Eddie, another farmer, would already use something new, his dad would wait. And so he's also the guy that kind of coined the terms early adopter after investigating that based on his family story. What he talks about is not necessarily adoption, like the whole concept of adoption is fairly new. What he talks about when he describes like studying new innovations is a diffusion of innovation, which like in his descriptions is slightly different. He actually says diffusion as a paradigm originates a bit more from anthropology because anthropologists look at things from the perspective of humans and they tend to take on that perspective to then blame a system or like credit a system when something good or bad happens, but always from the perspective of users rather than a system. So Everett Rogers defines diffusion as a process by which an innovation, which is one point, is communicated, which is the second. Over time, time is a very critical element according to him and among members of a social system. I started to think that what are then the difference between that type of diffusion and what we now see as adoption and what we talk about so very often of reaching a crypto mass adoption. Some differences if you look at the different models out there are that diffusion usually refers more to a society as a whole whereas adoption is more related to the decision of one specific individual. Diffusion relates to technology, infrastructure, whereas adoption usually relates to a product or service, something built on top of that. In terms of what it does functionally, diffusion relates to something being supportive, not delivering the actual value instantly whereas adoption is interactive, something that brings you value, some effort into it, it's interactive. And diffusion tends to take longer. So in this study, based on a range of different technologies being introduced, it takes about 20 to 30 years whereas adoption, if you think about all the popular web services, they take between five and 10 years. Hybrid porn was a really rapid adoption at the time for new technology, taking around 13 years I believe. And then also there's just a very different stream of research into both, starting with Edward Rogers but then there's this other guy, Earl Mann, who started to investigate adoption more closely. So next to porn, an interesting example is the computer mouse, which is commonly credited as being invented in 1973. Actually it had been around a little bit longer, and it was a really nice post, I may have included the link, I'm not sure, by Bill Buxton, a great guy if you haven't heard of him, look him up, he has a wealth of information about how new technologies grow. He kind of describes the history of the computer mouse that I think is very analogous to how we might be looking at, let's say, Web3 as a new technology. So the mouse credited to be invented in 1973 was actually 1969, and then it still took years and years before it actually reached household adoption. So there were commercially available products, especially by Xerox PARC, but before David's household adoption was in 84, which is pretty much a long time after. The way Bill Buxton describes that difference is that there's actually, before that whole S-curve starts, there's a long nose of the technology being developed where you go through invention, which might happen at multiple points in time by different people. There's a refinement and augmentation phase, and there's the productization where there's actual stuff built on top of the technology. In my mind, we're kind of like in refinement and augmentation, we're not really there yet, and let alone that we are at that place where you can start to think of it as an S-curve. Why the analogy to a computer mouse to me is interesting is because the computer mouse needed an ecosystem, and this is what made it last or take quite long for the technology to be adopted. So it needed this place to use a computer mouse, it needed programs to effectively control a graphical user interface, and it needed all the manufacturing around it that took so long. So looking at these pillars or variables, I would say, that Everett Rogers describes that are required for diffusion of innovation to happen is the innovation itself. It needs to be new and interesting to people, it needs to actually add some value. It also needs communication channels, it needs a network of merchants or people that can sort of sell the technology like the way the farmers have an alliance that would kind of vouch for the hybrid porn to be more effective than the average way of working. So it needs agents is how it describes it. It also needs time, it needs time for people to go through their phases of decision-making so they need to get their knowledge of how this stuff works, they need to have time to decide if it's something for them. They need confirmation that when they start using it it actually works for them. So there's all these stages that people need to go through first. And then there's a social system where it doesn't make much sense for me to have crypto if I can't use it anywhere else. So it definitely needs to have support for people in a closed environment. So let's say the diffusion of innovation but when we then talk about adoption which kind of grew later what can we learn from that work that can apply to us as well? There is several models that nowadays are referred to as adoption models but actually in the early days they were called technology acceptance models mostly because you didn't adopt technology, you accepted it because it was kind of enforced on or forced upon you by an employer. These are, I mean, in themselves, if you like models that might be interesting, there's a bunch of them. What they have in common is basically these two factors. Something needs to be useful or at least perceived as useful and it needs to be easy to use. And especially the first one, it needs to be useful at least from my perspective where I'm standing, I work for status which is very much infrastructure oriented. Useful is something that doesn't necessarily resonate. It's much easier for us to talk about the usability of the applications that we're developing at this point. So we could sort of add those to the model. And then there's obviously a lot of knowledge about UX which is also relatively new. If you kind of look at the process of how UX came to be, initially there was a bit more of a thing called human engineering when Bell Labs wanted to design the rotary phone and they were kind of struggling to decide do the numbers go on the device itself or on the rotary. And they hired someone to test that and to get input from people and they wanted to design it such that it would fit humans and they called it human engineering. Then the personal computer came and the industry started talking about human-computer interaction which to me that emphasis is interesting because it's talking about the computer, the system as a whole at that point in time. But if you look at what happened to the industry after that when web technology started to come up, the emphasis switched more to human interface design so it moved away from the system and kind of moved up to the surface of this is only what you interact with and that's what we'll try and optimize. And then even later, even though the term started to be used in 93, if you look at the actual definitions, user experience happened and that kind of only became popularized when services became more common on the internet. And this is like looking at the definition of user experience, how that kind of shifted over time. There was this concept already, like context abuse from these applications, right? Will something work in a particular factory with given timelines? That was kind of the extent of the usability. Then in 2010 there was a switch and this is only 2010, it's not that long ago. And that user experience was added as in the person's perceptions and responses that result from the use and or anticipated use of the system product or service. These are standards, they're not the most fun read, but it's interesting to see how they change over time. By the way, it kind of looks like context abuse is no longer important. It was just moved to another position in the document so very much they are still. And then another thing that happened in 2010 was that usability kind of got this side note added to it where there was like another emphasis on specified users, the goals, the context abuse referred to a particular combination of users, goals and context abuse for which usability is being considered. This is like reference to context. So last year, and I'm sorry to say that this was already last year, with some folks of status, we went to South Korea for a field research trip where this context became incredibly clear and also it became incredibly clear why we needed to be in the country to understand the context. What we found was obviously there had been a high adoption rate of cryptocurrencies but what we found was that it was especially connected to the housing market, the prices of the housing market, the social pressure for young people to buy a house and then next to that the investment regulation and how difficult it was for young people to find other means of investment than crypto. So this whole context kind of played into why people started to use cryptocurrency, not just to make an extra buck or because it seemed interesting. There was a whole sphere of context around it. And so looking at another model, there's loads of these in psychology. This is also next to Don Orman, one of the followers of UX I would say. There's this continuous reoccurrence of there's something... Sorry guys, that's my time. There's some reoccurrence of something needs to be usable but it also needs to be useful in the first place. And that you see popping up over and over. So one thing that we can add is the importance of context because we see it coming back in all definitions. This is kind of like a side note but a comment by this guy Nielsen that I find interesting is that over time we kind of switched from a you pay first then you experience something to you experience and then you pay with the introduction of service models. This is an interesting thing to keep in mind because we're in a way reverting that. You pay first, you pay guys and only then pay you experience. So from some of the studies that we did there's one more factor that I would kind of add to that model of what are things that are important to at least think of or keep in mind or check to cover our bases if we want something to be adopted. What we found in Iran where we thought privacy is going to be super important because we build a messaging application turned out privacy was, yeah it was important but it wasn't super important people had telegram, it was blogs and that only made it more trustworthy for some reason. What did come back over and over was this it should be common among everyone and my friends should be on there too so this social network and then another one was from I would say three months later when we did this familiar study in Japan and in this study we would ask people about the current messaging applications they use and we would ask to write either a breakup letter or a love letter depending on how they felt about their current application and when kind of looking at what does an application mean we've got through this strategy of write a breakup letter write a love letter make it really apparent we got to things like I could lose so many contacts or I'd be so lonely and kind of explanations for the use of technology and this to me would be reason to like emphasize or add sorry about the collars there this concept of human connection specifically because it's something that we really do not seem to take into account too much as in any transaction I make on the blockchain so far does not allow me to add a note or receiving any value does not allow me to say thank you which is a pretty basic thing in any human interaction and looking at other applications that seems to be one of the defining aspects so that's kind of like the list of factors that I would keep in mind like either looking at like the wider picture of the technology but also at the lower level like do we give people time between switching like trying out an application before they actually have to decide that they'll use it do we allow people to try something without backing up their seed rates so how do we fit into this this concept of decision making do we have enough communication channels or agents in place and there's a bunch of more I kind of see it as a list of triggers like things to think about and that's all I have except for the QR code which I think you know a lot better what to do with that than I do I was just asking about it there thank you a few more coming in thanks for being here my name is Chris Sugg I'm a UX UI designer at Aeroswap and today what I'd love to do is just tell you some stories about user research that we conducted at Aeroswap and specifically some of the challenges of adapting kind of traditional UX research methods to these specific users within Ethereum okay and I guess we're not the bad I also want to say you know this isn't really supposed to be a prescriptive talk per se but more so just sharing a very lean product team's attempt to conduct meaningful research within the space and figure out how to do it well so I hope we can all chat about it afterwards come up to me if you're doing stuff better than we are so off the bat here's the agenda we're going to do why user research because we can't be in a place like this and not advocate for it at large we're going to talk about two kind of buckets of user research and then we're going to focus in on one side and talk about some specific methods that we have dealt with and adapted alright so why user research the goal here is not just to build good products but it's to build the right product and when I say the right product I mean a product that is for the end user and I think this quote by IVM puts it well user research focuses on understanding user expectations behaviors needs and motivations through methodical investigative approaches insights are then used to ensure that all product design decisions do benefit the user and I think this is something for this space that we need to remember the products we're building are for the user they're not for us as the developers the designers they're for the user in the end and so let's work to get to that end so now let's talk about two kinds of user research if you're a designer or a researcher this may be a 101 kind of situation but we're going to go through it anyway we've got quantitative and qualitative so when we're talking about quantitative research this is evaluative research so these findings will come out in measurable data numerical format some examples of this could be heat mapping so how many times is a user clicking on a button how many times are they missing a field eye-tracking, AB testing, etc we're going to be talking about qualitative today which is more exploratory it sometimes gets to bad rap as sort of the softer of the research sciences but the whole point of qualitative research is to observe behaviors and attitudes the findings are non-numerical some examples here would be interviews field studies and focus groups this is the more sort of personal research it's human connection and understanding sort of core motivations of people as opposed to just numbers so just to give you a little bit of a snapshot the point here is not to say that one of these research sides is better than the other but it's about using them together to get the best research you can get so at AirSwap we use a combination of a ton of tools and this is just a few so on the quantitative side we use a couple products we use hot jar and mix panel this is where we do our heat maps recordings we track our conversion funnels KPIs, daily active users, etc these are the ones that we hear a lot about so qualitative research on the other side we have candy feature requests so our users can free form ask us for features telegram and discord feedback channels so again, sort of a free form avenue for our users to talk to us and then the three that we're going to kind of live in for the rest of our time together are surveys, field studies and user interviews so qualitative research we need it Ethereum needs it and the reason we need it is this kind of research is what takes us out of our context and shows us what our users really need so it takes us out of being developers, industry insiders designers that love Ethereum and puts us in the spot of our users to understand what they really need for products but qualitative research is hard in Ethereum and we're going to talk about why and I think the crux of the thing is basically this our users go by wallet addresses and not names so what I mean by this you can imagine on a traditional platform if I'm signing up for Pinterest I'm signing up with my name my email address probably my date of birth if I'm doing that I may have even checked subscribe to email list while I did that and then on top of it maybe they have social media presence that I'm following with an identity attached to it so in normal web 2 consumer web there is identity that is fundamentally connected to the product so when you're looking at doing recruitment or conducting research there are pretty easy avenues to get to those people to find them to know who's using it for us on the other hand we have wallet addresses and for our users this is one of the biggest value propositions of the entire thing their anonymity, their security is huge and we care about that but how do we conduct research when that's the case and it really goes into everything who even are our users where can we find them and then if we can figure out who they are if we can figure out where to find them how can we actually get to know them again in this true sort of human way get to know their needs so we're going to start with who are our users and we're going to talk about UX surveys so when we're talking about fundamental UX surveys in traditional world the goal of a survey is to gather both quantitative and qualitative insights from a lot of users at once so this is a research method that doesn't cost a lot you can send it out to a huge distribution channel a subscription list you can send it to a company and they can do it for you you can then extract trends and understand all kinds of things from a lot of people the second way this is used would be like a screener survey for recruiting users for future research so for us this piece was really interesting we were interested in the first of course but we want to recruit users we want to find these people I personally wanted to bring them into the office we'll talk about that later but so here's what we found the distribution piece was hard and we've done these are just two snapshots of a couple surveys but we've done a ton so when we were first doing a survey we were like okay let's get it to a ton of people and so we thought maybe Craigslist maybe we could put together a survey that would really weed out the real crypto people the real Ethereum people and we got a fair number of responses we got like 80 responses and we actually did bring three people into the office which was a total shit show but and it was interesting they did give us insight but what we realized is that it wasn't quite spot on enough that this distribution channel it wasn't quite representative of the kinds of people that were using our products they needed to be more savvy so then we were like alright this was too broad let's make it smaller let's not expect as many respondents but let's expect some quality so then we went to Telegram we distributed there you can see for this survey we had 11 responses which you know that's not a ton but the thing with these responses is that they were excellent super high quality responses we had people responding in paragraphs and paragraphs about their experiences they were willing to answer all of the questions and it was great but as far as recruiting them for future research they were not in of those 11 respondents responded with their Telegram handles none of which responded when I reached out to them so we realized that our users are absolutely willing to give rich responses to these open-ended questions as well as closed-ended questions in survey form but they still want to remain anonymous and so here we were like alright we got something from it so there you go we're not going to get those things from you at least not right now so where are our users let's talk about adapting UX field studies for a second the goal of field studies is to observe users in their natural environment and to situate oneself as the researcher in user-specific contexts so coming into the Ethereum ecosystem I come from working with internet stuff and consumer tech and when I think of field studies I think of doing a field study for REI going into an REI store for four hours and watching consumers look at the sunglasses look at sweaters this is not like that there's nowhere I can go to find Ethereum users to watch them do their daily things but just because we can't do it in person does not mean that we can't see what people are doing so this is an example of what I'm going to call a field study this is a screenshot from Telegram we were building a product for OTC traders and so this is a trading chat room and what's really interesting here there's a little bit of gibberish WTS hashtag axle 1900 0.1 ETH translates to want to sell 1900 axle at price 0.1 ETH so what's interesting here we were trying to figure out how to create a really intuitive order builder for OTC trades so the main inputs on this thing are selecting a token and amount to send and then selecting a token and amount to receive but what was so interesting here is that the syntax that these traders are using is not sell 1900 axle and receive 190 ETH which was represented in a lot of the products that were already on the market and so what we got to do is implement a price field which for us was a huge differentiator from what was already out there and we heard from our traders very actively that this really helps their workflow that they're thinking about what they're going to send and a price not just what they're sending and receiving so this is one example of how we sort of adapted field studies and applied it so now let's talk about UX interviews we want to get to know them and traditionally UX interviews are all about an informal setting to really get to know someone to get beneath what they think they want and to get at what they really want so you're trying to uncover user motivations, feelings, needs and pain points and it's very personal it's about putting a face to the person using it now this is hard like I said we had some people come in but again they weren't quite representative of the users that we really wanted to be building for and there are a few challenges with the users that we're really trying to build for which is that they are global they're everywhere, they're not going to come into our New York office and they don't exactly want their face to be seen so while we may not be able to see them we can work to create analogous research environments over chat and through forums so what I did was cold message people on telegram and said hi I'm Chris I'd love to just talk to you about OTC and a number of people were incredibly generous and really did talk to me and you know I don't have their name but I was still able to get to a point where I started to understand who they were and what they needed some other ways that we have approached this so that's sort of the individual connection piece to actively foster community around our products and so we do face product releases on our most recent one we start with a closed beta and this is paired with a telegram or discord channel that we've had people sign up for who then have conversation give us active feedback we implement that feedback we do a public beta we invite people into another channel and so just creating a space where we can still get this informal sort of human feedback on the things that we're doing so in the end we may not have a name and I think that what I have learned is that we don't necessarily need a name because we can still get to know our users through creatively thinking about the kinds of ways that they are interacting with the world and the way that they're interacting with one another so that's pretty much it but I would I mean for those of you in the room I don't know is anyone like a UX researcher or has conducted UX research okay yeah tell me what you do later because it's tough it's tough but it's good thank you yeah so we had to review the first I came and I was actually out on the sidewalk having a cigarette and he came up to me and he was clearly drunk like actually drunk and I was like oh geez and I was like running away and he was like oh is this air swap and I was like oh yeah this is air swap and he was like oh I'm here for the user interviews and I was like come on down we had dark web criminal and the third guy was like absolutely useless like I think he just wanted the gift card I mean there were insights to be had as far as just you know the consumer side these were definitely kind of like edge case people that were not exactly our people we were able to get something from it but yeah that was yeah oh well okay yes Google Analytics yes well if you use it I tried it with my team and they some of them objected it so we ended up not using it and I wonder what you are doing I don't think we're using Google Analytics oh we are we use it and but the fact that it's not the actual and that's the reason did you have a question yeah I mean you mentioned that the gift card so I was wondering about like incentives the other thing is like some of this stuff is new to me but you know like people raise the uniform badges to identify the litter bear to try to do anything creative around like badges and ballads to recruit people well we did think about these kind of things and we used a gift card for this particular session and I think that was quite applicable in the door but at the end of the day like the people that we want to be talking about we don't are talking to rather I don't think that necessarily luring them into the door are the people that we're trying to talk to like you know what we found on the surveys is like these people want to give feedback we all want to contribute to making it better it's just a matter of figuring out how to connect to them and find them so what are you doing like industry-wide studies is there any like somebody must be you're doing a label too there are I'll say yeah the Ripple team is doing the UX pattern in general it's a design system we've done a bunch of those and we are doing now on the industry or layer 2 there's a talk here tomorrow on that layer 2 design pattern how are you connecting with the users not actually users but like like a lot of product people are looking for aggregating and to understand what the connections are so like how do we obviously like these individuals companies have relatively small budgets but you know we need deep research from like that yeah it's it's very special if you want we can no no it's fine this is the the best this is the room for the talk so in terms like are you able to identify contact them directly or not so I think that what we found is you know telegram handles are sometimes the extent of the identification but then there are also of course those users that we do have actual relationships with that we know a name of or at least a first name but I think that in my head most of them are their telegram handles yeah yeah yeah yeah and of course we've done that I think that the idea because I think this is always kind of attention here is like do we just try to build for the people that are using it or are we trying to get you know at least a degree further towards a new user or you know so I think that represents also that goal so can you talk a bit about if any learning came from the quantitative side did you learn something interesting just about the users in general yeah or I don't know you design something and you discover something wasn't working somehow I think well what was fun was talking to some of our users and actually sending them designs so there are a few that we're pretty close with now that I've actually sent like active prototypes to and they have like mocked them up and added comments and you know we take it with great assault but it's been fun to see them sort of actively engage with what we're doing and from mixed panel and from that oh the quantitative side I think those and again it being evaluative it showed us like for example on our decks our instant experience like we noticed that people were missing they were clicking something that wasn't supposed to be clicked and they were missing the thing that was supposed to be clicked and it's that thing that you have to see that really improves it so yeah is there resistance to like things like segment and tracking data like if people freak out I don't know if they freak out but I think that in general we can say like ideologically in this space people don't want their data to be tracked or held but I'm not saying from personal experience I've come up against someone you know up and down and add something to that so not my previous company but a company I worked before we had a lot of users and we saw a drop of basically from 170% people actually blocking things like segment privacy measures so that meant that we were missing a lot of traffic because of that and then on top of that because we're European and we had the GDPR which meant we saw another drop of 40% so in the end we looked at other traffic we were really worried that our traffic was actually going down because of the time span of the year it was dropped by almost 50% but then actually looking at profit and other things that was pretty good so we saw this big mismatch between what our data was telling us and then what the reality was happening on our platform so then we went back to server side tracking and we wrote our own plugins to basically circumvent a lot of the tracking that was blocked so and then we made a combination of those to actually see what is actually going on so that we can make better decisions again because at first it was really scary to see all these drops while these drops were only from people blocking it or not allowing us to track those so yeah so I think that was for us a very good insight to realize that you think differently so that's one and the other thing is that from a statistical point of view it's very difficult if you have a small set of data to actually make something that you can drop conclusions from and I see a lot of smaller startups actually really trying to do analytics and things like this which is really not that useful but you have a lot of traffic so yeah so then things like Altjar recording sessions and looking into this that's very useful but also realize that the platform like Altjar is also doing the same so it's not recording every session trying to think okay this is something significant and then they're awesome