Your Customers Are Talking. Are You Listening?

Zeecee Moscow 

All right. Thank you, everyone, for coming to join us today. So today's session, as mentioned, is 'Your Customers Are Talking, Are You Listening?' And it's delivered as part of RMIT. Future Skills Festival. My name is Cece Mosco. I'm a Student Success Manager here at RMIT Online, and I'll be your host today. Before I get started, I'd like to acknowledge the people of the Woiwurrung and Boonwurrung language groups of the Eastern Kulin Nations, on whose unseated lands we conduct our business today. I respectfully acknowledge their ancestors and elders past and present. I'd also like to acknowledge the traditional custodians and their ancestors of the lands and waters across Australia where we conduct our business and from which all of you are joining from today. 

Zeecee Moscow 

 Today's session is all about understanding the difference between simply collecting customer data and truly understanding customer behavior and decision making. We're joined by Lee Barnett from one of our long-time industry partners, Simplicit, who will share practical insights on customer-led design, spotting overlooked signals, and turning customer insight into meaningful action. 

Zeecee Moscow 

 Before I hand over to Lee, just a couple of housekeeping things. 

Zeecee Moscow 

 This session is recorded and will be shared with all of you after the session today. 

Zeecee Moscow 

 We'd also love for this event to be interactive. So you've got access to the chat function. Please drop your questions through there throughout the session and we'll get through them at the end. And when submitting questions, if you could please just select everyone so the full audience can see them. That will help us to guide the Q &A later on. 

Zeecee Moscow 

 Which we'll have at the end. And so over to you, Lee. 

Leigh Barnett 

 Thank you ZC for the introduction, and hi everybody, thanks for joining today. As ZC said, my name is Lee, I'm at Simplicit, a HL Tech company, and I head up product design here. 

Leigh Barnett 

 I'm actually going to start today with a short story. I want to take you to this paddock. It's 4:30 in the morning. It's cold. And you're somewhere in regional Australia. 

Leigh Barnett 

 While most of the world is asleep, a third-generation farmer is sitting in the cabin of a half-million-dollar harvester. 

Leigh Barnett 

 He's got three different weather apps open on his phone, staring at conflicting rain radars, trying to make a high-stakes decision. Do I harvest a million dollars' worth of grain today, or do I risk waiting out the storm for a better yield next week? 

Leigh Barnett 

 He looks at the UI of the most popular weather app on the market, sighs, and says out loud to an empty cabin, 'I really don't care about the humidity percentage in the nearest capital city right now. I just need to know my topsoil is going to turn to mud in the next three hours. 

Leigh Barnett 

 We can go to the next slide. 

Leigh Barnett 

 I tell this story because as product builders, the decisions we make have real-world implications. 

Leigh Barnett 

 And because of that, we have an obligation to listen to the people we're designing for. 

Leigh Barnett 

 As I said, my name's Lee, and I just want to give a quick shout out to the team that worked with me on this engagement with one of our clients, Ben Carew, one of our principal... data analysts, and Nicola Bolognese, one of our senior product designers. 

Leigh Barnett 

 Without a really good team, good work's not possible. So big, big shout out to Ben and Nico. 

Leigh Barnett 

 What I'm going to talk about today is going to concentrate right at the tip of the product development lifecycle. 

Leigh Barnett 

 So the case study I'm going to walk you through is built around the... weeks that we spent in discovery. 

Leigh Barnett 

 The brief that we were given by our client was that they want to build and develop Australia's number one rural weather app. 

Leigh Barnett 

 There's three things that I'm going to cover off with you today. So as ZC said in the introduction, a big part of today is about listening. 

Leigh Barnett 

 I'm going to talk to you about some of the ways... that we listened to farmers on this engagement. 

Leigh Barnett 

 Importantly though, from listening, we need to move to translation and translate what we hear into impactful product design. 

Leigh Barnett 

 And then lastly, because we are making a translation, we need to validate that what we think to be true has actually worked out the way we intended. So going through a validation phase. 

Leigh Barnett 

 So let's jump into the listen and translate. 

Leigh Barnett 

 I'm going to start with why listening is important. And without a doubt, one of the most pervasive cognitive biases that I and my colleagues and my teams come across in our practice is false consensus effect. 

Leigh Barnett 

 I will say this is a cognitive bias that we are all privy to. 

Leigh Barnett 

 It is looking out at the world and believing that the way I see the world is representative of how others see it too. 

Leigh Barnett 

 This is a really dangerous mindset. It's something I try and talk to all of my clients about typically on day one of a new engagement because it gives us permission to let the data talk and kind of de-escalates from people wearing opinions that are unsubstantiated. 

Leigh Barnett 

 I think just, you know, if you can create a general awareness of this cognitive bias, half the battle is won. The risk of this is also that, through false census effect, because often we will have a strong belief that what we see is normal. It can lead to an overconfidence bias in the solutions that we create. And that's really dangerous. 

Leigh Barnett 

 And it's dangerous because the product development lifecycle gets much more expensive the further into it you get. Now, this is some very old modeling. It dates back to 2004. It's Scott Ambler from Agile Modeling. What he looked at is the cost of the software development cycle as you move through the different phases. 

Leigh Barnett 

 I'm not sure whether or not these ratios will ring true, but what I do know to be true is that the further you move through your product development lifecycle, even in this day and age, the more hands— effort go into building and creating that product and the more expensive it becomes. So the essence of what we try to do through listening is that when things are at a cost of one X. We're understanding the needs and the context of our users. 

Leigh Barnett 

 And we're doing that before we get into detailed design and development. And that just means that if we get something wrong, it's much quicker and cheaper for us to pivot. Change at a cost of 1x as opposed to a cost of 10x. And you can obviously imagine how those issues compound once you get something into market. The cost of sales, change management, marketing, and so on can really drive up the cost of change. So it's really important, if you think about the product development lifecycle that I started with, start with listening, gather your insight. 

Leigh Barnett 

 establish a product vision, validate it, and then once you've got your validation, you can move into the next phase of the product development lifecycle. 

Leigh Barnett 

 So a few different ways that you can listen to your customers. 

Leigh Barnett 

 market analysis. So thinking about who are the competitive set. 

Leigh Barnett 

 Thinking about new entrants to market, we'll often visit sites like AngelList or Product Hunt to identify who's getting early seed venture capital, who are the disruptors coming into an established market. 

Leigh Barnett 

 We can look at data and analytics, if you have an existing product, and you can understand how people are currently interacting with that. 

Leigh Barnett 

 Employees are a great source of information and intel, especially early in engagements. 

Leigh Barnett 

 Often they're the interface between a company and the products and services they offer. And there is a lot of really good insight to mine from employees. 

Leigh Barnett 

 User research, which I'll talk a little bit more shortly on. 

Leigh Barnett 

 Obviously, a really key activity, you know, employee feedback is valuable, but they do not replace hearing directly from your users or your customers. 

Leigh Barnett 

 Research is pivotal, and then, lastly, quantitative level studies. And again, I'll go into that shortly. The general question, trying to answer here, as you gather intel from all these different signals, is: am I building the right thing? And that's the first question that we always try to answer in our product development lifecycle when we start a new engagement. Is giving our client, product owners, managers, the senior executives, the confidence to know that they're building the right thing before they invest in building the thing right? And when we go back to the cost of change, it is obvious why we do that. 

Leigh Barnett 

 So what I'm going to jump into now is some very specific examples of a program of work that we delivered recently. 

Leigh Barnett 

 What I'm going to try and do here is link the insights that we gathered through listening to the decisions that we made in design. And at the end, I'll show you the impact that this listening and translation process had on the final product roadmap and product vision that we handed over to the client. 

Leigh Barnett 

 So the first thing we learned about weather is that time and time horizons matter, especially in the context of farmers. For us, it was really helpful that we had a primary audience that we could focus our research on and we didn't need to, you know, try and... embrace a really broad market that might have a very diverse set of needs. So having a very focused audience of a farmer was really helpful for us. What we learned through our research is that the near term is what is most important to farmers most of the time. So through user interviews, we heard this and observed this through their first actions and first clicks, which was seeking out the seven-day forecast. What we learned about that is that it's a critical tool that they use to plan their week. 

Leigh Barnett 

 Through the Top Tasks Survey, and I encourage you to look this up if you're not familiar with it. It's a methodology developed by Gerry McGovern. 

Leigh Barnett 

 The number one task selected across respondents was checking the seven-day forecast to plan activities. 

Leigh Barnett 

 Through analytics, we saw a high concentration of taps on the seven-day forecast feature. 

Leigh Barnett 

 And through the competitor review, we saw that all the competitor apps that we evaluated offered seven-day forecasts, and it was extremely prominent. 

Leigh Barnett 

 Now, with the benefit of hindsight, it obviously seems very obvious that seven-day forecasts should be prominent. But what's happening here is, by listening in different ways, we're building our level of confidence to make sound decisions in our design process. 

Leigh Barnett 

 And so what you can see here is a quick example of how we put that insight into action through the design choices we made in this prototype. So the first thing you can see is when the app opens and the page loads, you are default set to a seven-day time horizon. 

Leigh Barnett 

 And you can also see that there is a rain probability and amount in millimetres available immediately. 

Leigh Barnett 

 Else we learned about farmers is that they describe themselves as gamblers and they said what they're gambling on is the weather and one of the most critical inputs and variables for them to work around is rain. So for us making rain really prominent and giving them that seven-day forecast immediately was really important. 

Leigh Barnett 

 Now, what's interesting, when we observed farmers using this prototype and interacting with it, the next thing they wanted to do was get more detail on today's weather. 

Leigh Barnett 

 See, with a quick tap you can expand that accordion and you can get an hour-by-hour snapshot of what's happening with the weather over that day as well as a much deeper insight into different weather variables that are occurring. 

Leigh Barnett 

 What we also learnt, as I said, was the importance of rain. Typically, the way that farmers look at rain is through a radar view. 

Leigh Barnett 

 And one of the things that we saw in our user testing was that participants, as well as scanning for that seven-day forecast, are looking for rain probability and amount. The other important variable that they're looking at is wind speed and direction. So if you think about crop farmers especially, one of the things they need to do is often spray their crops. 

Leigh Barnett 

 So understanding when there's spray windows that are suitable is really important for them. 

Leigh Barnett 

 What we also observed, and you might recall this from my opening story, farmers are typically using multiple applications. 

Leigh Barnett 

 One thing we've come to learn... about weather apps is that they all have slightly different predictive models. And therefore, they'll have slightly different outputs for a user looking at things like probability of rain, rain amount, wind speeds and directions. 

Leigh Barnett 

 So they all use the same inputs from the various weather stations and hardware that's available, but they... convert them in slightly different ways with the underpinning algorithms. 

Leigh Barnett 

 What was really interesting was observing farmers using these two to three different weather apps and then overlaying their own judgment to build confidence in their decision. 

Leigh Barnett 

 The top task survey that we conducted. 

Leigh Barnett 

 So when we did an analysis of the top 20 tasks, we saw that nine of the top 20 tasks related to rain. 

Leigh Barnett 

 And the number two top task was viewing the live rain radar for their local area. 

Leigh Barnett 

 The rain forecast was one of the most used features in the current application. 

Leigh Barnett 

 And we also consistently heard from farmers throughout the research that there was one app that they would go to, especially for wind-based insight, and it was windy.com. And we heard them talk a lot about how intuitive the UI of that application felt. 

Leigh Barnett 

 So again, we've listened and now the next part of the process is to translate. 

Leigh Barnett 

 So if we go to the next slide, we can have a look at what that translation looks like. 

Leigh Barnett 

 So where the toggle is, you can see currently this is toggled to map view and where you came from on the previous screen that you were shown was the forecast view. So it is really quick for a user to toggle between the two key bits of information they want to see. 

Leigh Barnett 

 It's really interesting, I must say, designing. weather app when the the bomb app was got it so publicly wrong and perhaps they didn't listen to users as well they could have. But understanding what your users need to access and why it's important for them to access it fast helped us make these decisions about what we made most visually prominent, what we put within a one-click range or one-tap range for users to move between features. 

Leigh Barnett 

 It preloads the map to the user's set locations. So rather than defaulting to a large central city, CBD area, it'll default to where they are. So they're getting immediate insight that's relevant to them. 

Leigh Barnett 

 On the second screen in, what you can see is then you can quickly toggle your overlays. 

Leigh Barnett 

 And that helps you visualize different types of weather patterns. 

Leigh Barnett 

 Now, what's also interesting here is we learned through our research that farmers aren't probably going to pay a subscription. Just to remove ads. They need value-add features that will make them pay, sign up for a subscription. So you can see here, within the radar overlays, some of those overlays are classified as premium features and you basically subscribe to get access. 

Leigh Barnett 

 And then, lastly, talking to our client's chief agronomist, and we also heard this from a number of farmers as well. 

Leigh Barnett 

 We learned that farmers use different weather applications. Some of the more advanced farmers are also using and toggling between weather models. 

Leigh Barnett 

 To visit different websites to do that. So one of the things that we've proposed to build into the new application here is the ability to toggle between weather models. 

Leigh Barnett 

 Our hope and our hypothesis is this will make this app more sticky and reduce the amount of time that farmers are defaulting to use other apps as well. 

Leigh Barnett 

 And then the last example is long-range forecasting and observations. So as I said, most of the time... It is weather insights that have a sense of immediacy that are most important to farmers. However, there are also moments in time. In their planning, whether looking back at other weather observations or looking forward to weather patterns and forecasts far into the future, what you need to remember about these farmers, if you're not familiar with farming, is they're often spending tens of thousands, if not hundreds of thousands, and at some of the large, industrial-scale farms, millions of dollars on inputs to keep their crops healthy, their livestock healthy. 

Leigh Barnett 

 And the weather has a huge impact on the inputs that they're going to be buying. 

Leigh Barnett 

 So this is one of the tools that was really important for farmers. To have access to, so that they can start to think about some longer-term planning for spray windows, harvesting, livestock health, their travel plans, storage, post-harvest, et cetera. 

Leigh Barnett 

 In the analytics here, we also heard that the rain charts, whilst not necessarily as popular as some of the other things, were very sticky and so with the weather observation in particular it was 16. 3 times per user that that was visited with an average visit time of six minutes and 16 seconds. 

Leigh Barnett 

 So what we could see from that immediately is, although it wasn't a huge cohort of users, it was really important to some of our users. And I think it's important to visit those edge cases and understand what's happening. 

Leigh Barnett 

 And so with that, again, active listening, building our understanding of the needs of farmers, how they're using this application to inform their day-to-day decision-making, we then had to make some choices about how we designed the interface. 

Leigh Barnett 

 I think the first thing is the rain radar. What we wanted to provide farmers with here was a really quick view. You can see this is on a monthly forecast setting of what they can expect in the next month. 

Leigh Barnett 

 We've used colour and bolding to set a sense of priority. Thank you. We've used clear icons and text to show probability versus amount. 

Leigh Barnett 

 Then, in the next one for the observation history, what we learned is that obviously rain is really important. So the first thing they're scanning for is rain, and we're also then able, through our user testing, to gauge what was the next most important thing. So beyond rain, it was time and temperature, wind, humidity, and dew point. So again, we're building our interface and we're making our decisions based on the intel of what we heard from farmers. 

Leigh Barnett 

 So the last phase, so we've been through now a very expedited overview of, like, six of the eight weeks thereabouts of what we did. 

Leigh Barnett 

 What's really important, though, is we're going through and we're hearing and we're translating and we're making decisions in our design. It is that we go back and validate them because, as I said earlier, there is a process where we have to make a judgment call on how to execute on the insight that we have. 

Leigh Barnett 

 So the way that we do this, the way that we build our confidence, is we go back through retesting. 

Leigh Barnett 

 And it's really important because what follows... the validation phase is the output of the roadmap. And the roadmap is a product vision, and often that product vision is going to extend out multiple years. 

Leigh Barnett 

 And so for us, as I said earlier, we want to make sure that we're providing enough evidence and confidence that a client is building the right thing before they go into a multi-year investment to build the thing right. 

Leigh Barnett 

 The way that we did that on this engagement was through some more moderated user testing and also using some standardized surveys. 

Leigh Barnett 

 This scenario we use the user experience questionnaire, we use the short version here just because we were time constrained in the sessions that we ran, so it's an eight-question version of UEQ Short. Again, I encourage you to just jump in your browser and look up UEQ Short. There is some really good public resources available for the full year. As well as a UEQ short survey, including analysis sheets. So it's a very simple set of questions you can build into your testing. 

Leigh Barnett 

 Research plans and what I'm showcasing here— so interestingly— in this engagement we hit a moment two weeks into the project where we've done some of our listening, some of our discovery-based activities, but we had hit a snag in terms of recruiting farmers. 

Leigh Barnett 

 And so we needed to think about what we were going to do while we waited— sort of a week, a week and a half— to fill our recruitment quotas. 

Leigh Barnett 

 So we decided, even though we had limited information, that we would design an early version of the app. 

Leigh Barnett 

 The information we had was from our landscape review, our SME interviews with staff. 

Leigh Barnett 

 And that was about it. 

Leigh Barnett 

 So as you can see, we went through moderated testing. We learned some really important information. 

Leigh Barnett 

 We also saw that it benchmarked in the bad category on the UEQ short. 

Leigh Barnett 

 This didn't dampen our hopes at all. As a matter of fact, it left us feeling much... more optimistic because the insight that we gained through prototyping early and testing, we just felt very empowered. In terms of what we were going to be able to do with version 0. 2 of the prototype and product vision. And so, on the right of this screen is... The version of the app that we tested at the end of the discovery program. So we'd conducted all of our user research, our surveys, our data and analytics reviews. And as you can see, it benchmarks in the excellent realm. What I will say, if you do use UEQ on prototypes, it's typically going to... to under-index slightly on the hedonic qualities. 

Leigh Barnett 

 Mostly that's due to the fact that it's a clickable prototype. And so you don't get the ability to kind of show. showcase some of those more delightful interactions that you might in a product that's fully built out. But still, for us, it was really good validation. Of where we started and where we got to, and a great way to showcase back to our client, our product owner, the head of digital at our client, that, you know, we've got a strong sense that you're building the right thing here. 

Leigh Barnett 

 So that's me for today. Thank you very much for, yeah, giving me an audience and I think we're going to do some questions now. 

Zeecee Moscow 

 Thank you so much, Lee. And just before we jump into the audience questions, I just want to reflect that that was so fascinating and so important to really go back to the core purpose of what we do. I liked what you said about constantly asking about that user experience throughout that whole development cycle and what do they want out of this product at a constant basis. It's just so good to have that reminder in our day to day. Before we jump to the audience questions, you work with a lot of organisations on this. What's perhaps one of the most common things that they think they're doing that they're actually not when it comes to listening to their customers and building out based on the customer experience? 

Leigh Barnett 

 Yeah, that's a really good question. I think typically what we see is that, you know, there's not a one size fits all when you look at an organisation. A lot of the companies we work with are kind of larger ASX listed companies or government type clients. 

Leigh Barnett 

 So they're very large and diverse organisations. I think what I see, what my team sees most commonly, is that there's different pockets of maturity in terms of their ability to go through a process like this. 

Leigh Barnett 

 I would say, though, as a general trend, when you introduce somebody that's not familiar to this process to the process. And you don't just run the process kind of at them in a black box and then come back at the end and say, 'ta-da.' But you bring them into it and you show them how you establish discussion guides, how you think about screening for recruitment, how you gather hypotheses. 

Leigh Barnett 

 So you bring them into those rituals, you bring them into the process, and I'm not sure that I can recall that I've ever had pushback with somebody saying, like, 'No, I don't like this.' So, yeah, if you can, bring your clients into the process. 

Zeecee Moscow 

 Awesome. Thank you. It's actually ties in really well with another question we received from the audience specifically around what do you do when your analytics seem to contradict what you're hearing from the customers? Have you ever had an example of that happening and how did you approach that? 

Leigh Barnett 

 Yeah, definitely. I think that's a smart question too because, you know, analytics is telling you what's happening today. 

Leigh Barnett 

 Now, the thing is— if the... application, the website, the software that you're evaluating isn't working well today, it's not going to really be that helpful to look at it. Analytics and try and make judgments about what's the most used feature today, where are people clicking most today on the entry point. So I think what you— user testing gives you— is an ability to look into what is the user actually trying to achieve? What are the friction points that they're encountering in doing that? And typically that won't come up. In the analytics, or it may through something like funnel conversion drop-offs, but you don't know why it's happening. So you might see 100% of people come into a certain flow. And at step one, they all drop away. 

Leigh Barnett 

 So you can say, 'Oh, wow, that's happening,' but you have no idea as to why. So user testing, reaching out to your users, gives you the ability to understand the root cause of why so many people are dropping out, for example. 

Zeecee Moscow 

 Yeah, that makes sense. So it's like the analytics are really telling us where things are at today within the constraints of the existing platform that we have, whereas the customer experience and those conversations are helping guide that and provide more clarity. 

Leigh Barnett 

 For sure, for sure. And I would say, like, for high-performing products, analytics is really helpful because you can see what's working today. 

Zeecee Moscow 

 Hmm. 

Zeecee Moscow 

 And how do you see the role of AI in all of this? Is AI supporting us in understanding our customers better and knowing what they need and guiding those decisions? Or perhaps could it be, I mean, a bit of a disruptor and perhaps causing conflict in that understanding? 

Leigh Barnett 

 Hmm. 

Leigh Barnett 

 This is a big question. 

Leigh Barnett 

 I think it would be dishonest if I didn't admit that I still feel slightly conflicted and confused. By what I read day to day and what I see day to day in terms of the capability of AI. I'm sure many people on the call are feeling similar. 

Leigh Barnett 

 What I can tell you, so in this engagement, we actually built our own synthesis agent. 

Leigh Barnett 

 So we used an agent to synthesise the proletariat. research that we've conducted. 

Leigh Barnett 

 The important thing there, though, was that we didn't just do that process post research. We actually thought about it pre-research. And so a lot of decisions we made in the way that we set up for our research. 

Leigh Barnett 

 Hmm. 

Leigh Barnett 

 They were really important to make sure that we got a good output from AI. So for example, the way that we structured tasks in our moderated guides, they're all consistently numbered and labelled. We carried that numbering and labelling through to our spreadsheets that we used to kind of anonymise data. 

Leigh Barnett 

 And we also carried across the labelling of our screens. And so by the time we fed those inputs into the agent that we built to do the research synthesis, there was a really structured model and framework there for the agent to follow. 

Leigh Barnett 

 Now, we still had, so Nico, who you saw in that introduction—slide— he still ran all of the testing. The difference is we've historically run paired testing. So we have one researcher and one note taker. 

Leigh Barnett 

 Instance it was NICO and transcriptions, but I think the benefit of having the person still conduct the research is that, then, when they get to post-synthesis through AI or an agent, which, by the way, sped up a four-day process to a matter of like 30 minutes. 

Leigh Barnett 

 So a lot of... game, but then also he's close to the inside because he's conducted the interviews so he can, as he's doing his QA, identify are there any anomalies here, does that not make sense, etc, etc. So I do think there's areas where AI is going to you know, I'll say too, I was very impressed with the quality of the way the large language model synced this. 

Leigh Barnett 

 It works extremely well with language, obviously. The clue is in the name there. 

Leigh Barnett 

 But the quality of the output was almost ready for us to just transpose straight into our report. 

Leigh Barnett 

 So I was really impressed with that quality. So I think, yeah, it'll make us faster in pockets but also improve quality. 

Leigh Barnett 

 And then for... us, you know, it meant that it freed up time for Nico to be thinking about iteration 0. 3 and he had more time and days to do that rather than spending it in research since. So that's one view. The other view, just quickly, because it is a big topic at the moment, would be we actually, in version one of this application... Built out like a pharma bot, so the pharma could chat to this sort of chatbot-style device. We called it Eldrix. 

Leigh Barnett 

 Our, like, intent and our thought process there was that, okay, well, what if a farmer could just, like, ask a question to this chatbot? Like, what's the... best sowing window this week or, you know, when should I harvest my crop based on commodity prices this month? 

Leigh Barnett 

 Now, what was really interesting was we prototyped that out really quickly. The feedback we got from farmers, it was really low trust. 

Leigh Barnett 

 We won't completely eradicate it from the long-term roadmap, but I think what it gave us was some clarity that for version one, we probably won't invest in building. That functionality out and there's still work to be done in terms of kind of building trust, building a sense of authority with farmers before you deploy something like that. 

Zeecee Moscow 

 Really fascinating. And that model sounds really sophisticated in terms of synthesizing the data that you have. I'm interested to understand, I guess, as someone in the room today, if someone were to walk away from this session with limited skills, limited budget and access to like staff. Try and put this in place, how is how is a way that they could approach this or assimilate that sort of experience, or create something to generate similar insights in an effective way. 

Leigh Barnett 

 Sorry, so Zissi, is the question around how would you go about creating something that does research? 

Zeecee Moscow 

 Especially with limited resources, yeah. 

Leigh Barnett 

 Yeah. I mean, the quick answer is you could get a product off the shelf, like Askable, for example, that has... these tools built in, you know, probably that product's not accessible though for the person on the low budget today. 

Leigh Barnett 

 What I would say is, you know... AI and large language models are really good at writing instructions for large language models. So your first port of call should definitely be setting some really clear context, telling it the goal you want to achieve, setting some parameters in terms of what it can and can't do. 

Leigh Barnett 

 Put that into whatever it is you're using— Copilot, ChatGPT, Gemini. 

Leigh Barnett 

 Give it the goal of building out an agent. Put it on deep thinking mode. So that it can then come back and ask you clarifying questions and just hit go. 

Leigh Barnett 

 It was really eye-opening for us. Like, that's the way we approached this. And it was really... important for us. It generated a markdown file, which was really clear. It articulated a set of guidelines. Guidelines and rules for responsible AI that were really clear. 

Leigh Barnett 

 So I would say just use the tools, experiment, keep yourself in the loop. Do not trust the tools blindly. 

Leigh Barnett 

 So keep yourself in there as a researcher and just experiment with it. 

Zeecee Moscow 

 Yeah, cool. Moving to the false consensus concepts that you mentioned in the presentation, someone in the audience has asked, 'How do you make the case internally for investing in customer research when leadership already thinks they know the answer?' Do you have any insights on that? 

Leigh Barnett 

 Um, yeah, I do. Look it's really hard. Right, because there's a power dynamic at play if you're dealing with a leader that thinks they know the answer and you're a designer trying to push back on that. It can be hard. I think, as good designers and researchers, we're decent at asking questions. So the first thing I'd be saying is, what gives you that level of confidence to think that's a direction? Is there any data that you've seen? Is there any evidence that you've seen that supports it? And maybe there is. Maybe they've seen it. Data that you're not privy to, and they can share that with you to bring you on the journey. 

Leigh Barnett 

 I think the cost of change model is a really important one for us. You know, if you can put the owners. Back on the decision maker to say, get this wrong, it's at a cost of 100x once it's built, versus allowing a little bit of time now to answer the first part of the equation, which is, are we building the right thing? 

Leigh Barnett 

 So if you can maybe just... talk a little bit in the language of business, talk about risk mitigation, the cost benefit of doing research early rather than finding out in deployment that you built the wrong thing. Those types of things will help you. 

Zeecee Moscow 

 Cool. We had one final question, which links to this before we start to wrap up. So I guess it's speaking about like someone in a position where they're driving data decision, data driven decision making. How's that for a tongue turner? 

Zeecee Moscow 

 And if they're in that role of providing those data and insights to the customer-facing team, how do they help drive those insights? So would you say that's really about making that case when you're having the conversations to put the data in the front and centre and to advocate what the impact of not making these changes based on their research is from the start? 

Leigh Barnett 

 Yeah. Yeah, um, so I think, like, you know, it's really easy to say, like, just do these things. Um, you know, the best definition I've heard of a culture is that the culture is all the sort of micro decisions and micro actions that each of us contributes each day. 

Leigh Barnett 

 So I would say don't undervalue the small difference that you can make. If you, and I would say my team, like, ask for permission, ask for forgiveness. Permission, right? So if somebody goes away and and does some research comes back to me with some insight, I'm going to be I'm going to be really happy with that. And that's a really good way that you can influence your own leaders is perhaps just go away and do a bit of a guerrilla project on the side. 

Leigh Barnett 

 And like I said, it's not a switch. So if you've got to change your mentality, it's going to take a long time and a lot of effort. 

Leigh Barnett 

 The last thing I would say, and I'm sorry, but, like, if you're in a culture that just doesn't seem at all like it's data-driven or going to accept data-driven, maybe that's not the place for you. And, you know, ask good questions when you're interviewing at companies. How do you make decisions? What role does data play in building out your roadmap, et cetera? 

Zeecee Moscow 

 Tchau. Thank you, Lee. That was so fascinating and really got me thinking. I'm sure it's gotten everyone thinking in terms of how they can really embed that customer insight and experience into their product creation in the future. Unfortunately, that's all the time we've got today for questions. So thank you, everyone, for putting your questions in the chat and for engaging in the session. And thank you, mostly, to Lee for sharing with us. 

Leigh Barnett 

 Yeah, thanks so much, Cece, thanks for hosting. Thanks to RMIT Online too for making this event possible. 

Zeecee Moscow 

 Thank you. And just before we wrap up, I'd just like to acknowledge a few things that are happening at the moment for RMIT Online. So we've got Future Skills Fest, which is obviously this is a session that's a part of that. The events are not over. We've got some more sessions going on throughout the rest of the week. So if you enjoyed today's session, please look online at rmit. edu. au forward slash Future Skills Fest to find out more. You can also check out RMIT Online's courses. If you're interested in learning something a little bit more structured, it's a really great way to do that. And we're also giving away five free features. Skills short courses to attendees who share their key takeaways from the session on social media. Just make sure to tag RMIT Online in your post so we can see it. But that's a great way to continue the learning, continue the conversation, and share what you got out of our conversation today. 

Zeecee Moscow 

 Yeah, thank you. Thank you again to everyone. We'll share a copy of today's session with you afterwards and enjoy the rest of your day. Thank you again, Lee.