00:00:02:11 - 00:00:27:01
Speaker 1
Hi everyone. Thank you so much for joining our masterclass today on designing personalized customer experiences with AI. For those that are new, welcome. And for those that are returning, welcome back. It's so great to have you all here. My name is Tash Rallios and I'm the Industry Partnerships Specialist at RMIT Online. As the title points out, RMIT Online is the online arm of the university.
00:00:27:03 - 00:00:57:16
Speaker 1
We specialize in building a wide range of courses, from short future skills courses to accredited courses, all for the working professional and for working professionals who want to upskill, reskill or even pivot in their career. We've had the privilege of working with a wide range of partners, including Concentrix Catalyst, one of our longstanding partners who we've collaborated with on CX Strategy and Design Thinking for innovation.
00:00:57:18 - 00:01:25:07
Speaker 1
Now, whether you're new to our masterclass or you've joined us before, we are genuinely excited to keep bringing these to life throughout the year. Every few months, we bring in awesome thought leaders from our incredible partner network to deep dive into topics that are engaging and relevant, and to get people talking and learning. Now, today I'm super excited to have Eric Fitzgerald and Timothy Spooner from Concentrix Catalist join us.
00:01:25:09 - 00:01:51:13
Speaker 1
But before we get started, I'd like to recognize and acknowledge that I'm hosting this webinar from Melbourne, which is the country of the Woi Warrung and Boon Wurrung language groups of the eastern Kulin Nation. I respectfully acknowledge their ancestors, elders, past and present. Today we're diving into two of the hottest buzzwords in business right now. CX and AI.
00:01:51:15 - 00:02:21:00
Speaker 1
Everyone is talking about them and it makes sense. These days, customers want things fast, they want things personal, and they want things seamless. And that's exactly where CX and AI comes in. While tech is key, the human impact is what truly matters. And with that in mind, I'm personally excited to hear from our speakers today on how brands are utilizing and embracing AI and CX and how to really strike that balance between automation and empathy.
00:02:21:02 - 00:02:47:14
Speaker 1
Now, some of you might have heard of Concentrix Catalyst, but if it is new to you, here’s a...really quick intro. They bring together strategy design and engineering all to create smart tech powered experiences. They work across multiple industries, from retail to healthcare, ultimately helping organizations become more adaptive and customer centric. Now let's get to the really exciting part.
00:02:47:20 - 00:03:20:14
Speaker 1
I cannot wait for you to meet our brilliant speakers who will be leading today's conversation. It is my absolute pleasure to welcome Eric Fitzgerald, Regional Design Director, ANZ and Tim Spooner, Design Lead. Eric directs the ANZ design team leading project strategy, hiring, career development and client engagement. A seasoned start up professional, Eric has co-founded multiple startups and has been involved in a wide variety of accelerator programs in Australia and also in Ireland.
00:03:20:16 - 00:03:47:09
Speaker 1
Tim leads...a team of designers on client projects across ANZ, including the Queensland Police Service Next Gen project, aimed at developing a new mobile platform for frontline officers. With a background in graphic design, Tim is also dedicated to supporting the team's career growth. Now, together they bring a powerhouse of experience and innovation and we are absolutely honored to have them join us today.
00:03:47:11 - 00:04:06:14
Speaker 1
Now, you've heard quite a bit from me, and I'm sure that you're sitting in your seat ready for these speakers to get started. So I promise I'm going to be handing over to them very soon. But before I hand things over, we really want to ensure that this session is as interactive as possible. It's been so great to see so many chat messages come through already, so please keep them coming.
00:04:06:16 - 00:04:24:05
Speaker 1
If you have any burning questions throughout the presentation, please pop them in the chat as we'll be monitoring that throughout, and we'll have some time after Eric and Tim present to get through some questions. So I'm super excited. I hope that you can see that on my face, and I hope that you are just as excited as I am.
00:04:24:07 - 00:04:29:20
Speaker 1
I'm going to pass it over to Eric and Tim now to take it from here. Thanks everyone.
00:04:29:22 - 00:04:35:22
Speaker 2
Thank you Tash. That's an amazing intro, thank you very much. Tim, do you want to say a quick hello before we go to the slides?
00:04:36:03 - 00:04:39:00
Speaker 3
Hi everyone. We're very excited to be here.
00:04:39:02 - 00:04:59:09
Speaker 2
Cool. All right. So I'm going to just, share, my deck. I hope everybody is seeing that. Can I get a thumbs up? Tash, you can see it. Okay. All right. We are Concentrix. Okay. And Tash gave you a bit of an intro beforehand. And what we're here today to talk to you about is AI and CX. So.
00:04:59:11 - 00:05:23:00
Speaker 2
Great. We're mixing all these lovely letters together. But what does that actually mean? So as I said, Tim and I are both designers. So we're talking about this from a design perspective. However, a lot of this is applicable regardless of your like what role you're in. You can use this thinking across the board. So, hello, hello. Myself and Tim, if you want to hit us up on LinkedIn, please do.
00:05:23:02 - 00:05:44:23
Speaker 2
So I want to talk a little bit about Concentrix. Now, I this isn't a paid ad. I want to talk about this because it's relevant to what we're going to show later on and how we actually approach our work. So we are primarily customer support. So that's what the background of what we do is. So customer experience, customer support, multiple industries like we do big infrastructural industries.
00:05:44:23 - 00:06:06:23
Speaker 2
So we have to think across like from the absolute weeds up to the up to the 50,000ft view of what these things are. We're a fortune 500 company and globally we are actually 450,000 employees. We're big. And that number shocked me. I only found out recently, but we're huge. We're global. But we're also probably the one company you've never heard of.
00:06:06:23 - 00:06:23:13
Speaker 2
So, it's quite interesting that way. So Catalyst essentially is the digital arm. That's who we are. Okay. So we design, build and run, these projects now we are to design part of it. But if you look at us from left to right, we do like you have to set it up correctly. That's where design comes in.
00:06:23:16 - 00:06:46:18
Speaker 2
We work across the board, but all the people who do all these other things, the build team and the run team. So we get a unique perspective in terms of delivering solutions and seeing if they work as well. So we work very closely with clients across the board from strategic to actually implementation. So what we're going to talk about today is now I'm going to go with five topics here.
00:06:46:18 - 00:07:09:11
Speaker 2
But I'll really want you to look at this from. We want to give you a mental model about how to approach AI. AI, I don't know about you, but it is like a tsunami. Like there's so much there that I need to know. However, and this is what myself and Tim and our teams had been essentially grappling with for the last year and a half, since, like, the floodgates have opened.
00:07:09:13 - 00:07:33:10
Speaker 2
And what we've tried to do is that we've gone back to like, essentially back to basics in terms of, using AI intelligently because it's very easy to use AI and be producing all this incredible stuff, but it doesn't have any value. Like you might have heard the term AI slop, which I'll get to at the very end, but what AI slop is essentially is it's just output, you know, it's...output without intent.
00:07:33:13 - 00:07:53:14
Speaker 2
So we're going to initially talk about and this a little bit about data and how it works. But it's very relevant. So how AI is transforming customer insights customer insights being what a customer needs and what they think. Then I'm going to hand over to Tim and he's going to talk about designing products to align intent, which is a very important word with autonomous agent behavior.
00:07:53:19 - 00:08:23:16
Speaker 2
Now the agent is the AI agent. Then decide fundamentals. So this is a very bad pun. I will explain what that means in a minute. Proactive AI and UX thinking again we’ll break this down to like everyday concepts that you understand like how can you use AI proactively within your UX or design thinking paradigms? And then finally, once we've hit you over the head with all this, with the theory, we're going to show you how it actually works in a case study, a theoretical case study.
00:08:23:16 - 00:08:44:05
Speaker 2
But like, how can we actually implement this? Okay. So let's start with how is AI transforming customer insights. So if you look at traditional analytics right now, it's this is if you look back over the last year, it's rule based, it’s keyword matching. It's kind of hard to scale to be quite honest, because it's all about manual tuning.
00:08:44:07 - 00:09:06:19
Speaker 2
And if you're actually getting inputs from a client, you can miss stuff like sarcasm, slang, nuance, and it's slow processing and accuracy. So that was up till recently. That's early AI and text mining or what you call machine learning. Now, now we've gone from reactive, which was the past, to proactive. So AI doesn't just analyze your customer, it understands them.
00:09:06:21 - 00:09:31:12
Speaker 2
Now Tim will talk about intent. But when you hear things like about agenetic AI, what does that mean? Like it's really creating agents to work on your behalf. Like, they pursue goals through reasoning and planning and it adapts to the context dynamically. So, as I've talked about with the previous, deep learning and machine learning...the modern AI understands meaning, intent and emotion.
00:09:31:12 - 00:09:56:14
Speaker 2
It learns from huge data sets automatically. Like data is actually the gold mine of all this, and clean data even more so. And it does stuff like the text tone, context, and even implied meaning. So you're almost getting a human interaction and it's fast, it's highly accurate, and time...gives you time, real time, insights. So agentic AI down on the other side of it.
00:09:56:14 - 00:10:16:12
Speaker 2
And actually, Tim, we'll get into it a little bit more kind of acts on your behalf. Now, the key advantages for understanding customer behavior true data is you get real time processing. It instantly adapts to customer behavior, not after the fact. It gives you predictive insights, which is this is the kind of magic that we all see.
00:10:16:14 - 00:10:37:21
Speaker 2
He anticipates your customer needs before they surface. And then the final one. I know it's not very sexy. It is not very designed, but unstructured data is actually key to a lot of this. It's a very important pillar. So it makes sense of things like reviews, charts, social posts, and not just numbers. But that's all unstructured. It's unrelated to each other.
00:10:37:21 - 00:10:48:21
Speaker 2
So it brings that all together to create this, contextual, interaction. So I'll hand over to Tim now.
00:10:48:23 - 00:11:07:24
Speaker 3
Thank you. Eric. So yeah, let's talk about how do we actually do this, how do we go around and create these products or experiences that align that human intent with the autonomous agent behavior? There's going to be kind of three aspects I want to break down here. I'm going to talk about agentic systems and what they are.
00:11:08:00 - 00:11:27:16
Speaker 3
It's kind of the the frontier at the moment in terms of talking about AI. We sort of moved from that generative AI space into the more agenetic systems and what they look like. And then I'll talk about designing for those systems. So how do we actually start to break this down and start designing with AI to actually deliver on these customer experiences?
00:11:27:18 - 00:11:49:08
Speaker 3
And then I might touch a little bit on designing with these systems as well. So how we actually incorporate AI into design practices and use them to enable us as designers as well. Next slide. So here's a quote from Ezra Klein. It's a really great way to just start thinking about agentic systems. So he says here he has this example in his head of he says to his AI, my son is turning five.
00:11:49:08 - 00:12:10:08
Speaker 3
He loves dragons. We live in Brooklyn. Can you give me some examples for planning his birthday party? Then you can choose from those and it'll actually go away and order the cake. Reserve the room and do whatever else might be needed. And there's a couple of things going on. And what Ezra's talking about here. He's he's is using natural language, first of all.
00:12:10:08 - 00:12:31:19
Speaker 3
So with generative AI, we've seen a lot of that. And that's actually an amazing thing. And just being able to talk to a system using natural language is huge. Now he's actually only given it some information as well. And so agents are able to work with partial information and work through that. So he hasn't given it a budget, but maybe when it gives him the options it'll actually think ahead and go, I'm going to give you a variety of options and that kind of thing as well.
00:12:31:21 - 00:12:50:24
Speaker 3
And then it will actually do the things. It'll go and book that event. It'll invite his friends to it, it'll put it in his calendar, do all those things in a kind of proactive way as well. So that's at the core, what agentic AI can do. Go to the next slide. So it is moving beyond this generative AI.
00:12:51:01 - 00:13:10:00
Speaker 3
Playing with ChatGPT. In the old days, it was very much single out prompts. And you would work on that and it would be based off this kind of singular flow. It was very human initiated. You kept having to remind it of things and and fine tuning quite a bit as well. And it definitely wasn't proactive. And agentic AI shifts away from that.
00:13:10:00 - 00:13:29:05
Speaker 3
It'll create its own goals. It should have high autonomy, can multitask. Plan and adjust and things like that as well. And we're seeing parts of that in AI systems already where they do rags, but they'll go and search websites for you and collate information and do deeper thinking if needed as well. But it's going to get even more of that as we move forward.
00:13:29:07 - 00:13:31:07
Speaker 3
Next slide.
00:13:31:09 - 00:13:54:08
Speaker 3
So what does this mean for customer experience or digital product design? It means there's new actors in the ecosystem. And we like to we can see it going where there's there'll be the kind of personal AI agents. You're starting to see that with, Siri bringing ChatGPT into their operating system. Gemini, that kind of thing. So that's the your agent that sits on your side as a, as a person.
00:13:54:14 - 00:14:21:06
Speaker 3
But then there'll be also organizational side AI agents as well. And you might engage with them directly. Or if we go to the next slide, it might actually be this, humans talking to agents, talking to other agents, so that this idea of you'll have your personal agent initiate something for you, which would then go and interact with another agent, to, to collate data or to to book a system or to do something like that as well.
00:14:21:06 - 00:14:40:02
Speaker 3
And that agent may actually then go on to book to something else. So if you think about Ezra's example, he wants to get some things, booked which might your personal agent will go into Booking.com, which might then engage with the actual place places agent to book it or something like that, or might call the person on their behalf that kind of thing as well.
00:14:40:02 - 00:14:59:21
Speaker 3
So this sort of ripple effect happening there as well, and it is important to think about agents as being multiple. If generally you're not going to have a single agent system, they won't be common. It will be this idea of multiple agents engaging with each other and being orchestrated, likely by your personal agent or another orchestration layer agent as well.
00:14:59:23 - 00:15:07:06
Speaker 3
So that ripple effect is important to consider when you think about this as well.
00:15:07:08 - 00:15:25:06
Speaker 2
Thank you Tim. So design fund-AI-mentals. Sorry. She laughed at us. It was a pretty good joke in my book, but, so, I just want. So Tim and I have just brought you through some conceptual things. Now, I want to bring you through something else conceptual, so you can start actually thinking. How do I use this?
00:15:25:06 - 00:15:43:04
Speaker 2
Where do I use that? Like bringing those sensibilities of like, it's not just this magic wand of AI that you kind of flutter around your, your project. You actually have to be quite systematic about how you do it. Now, this is an approach. There are many approaches. Find the one that works for you. This is the one that works for us.
00:15:43:09 - 00:16:08:06
Speaker 2
So what is design? Design and design thinking takes the abstract and brings us through to something concrete. Okay, so we look at a project problem, whatever it is. And then you come out with something very, very clear at the end. Now when we look at a successful design solutions, we look at them from inside out. Now, you might be I don't know where, everyone is, but I highly recommend you look up.
00:16:08:06 - 00:16:27:02
Speaker 2
There's the five planes of UX design, which are going to show, you know, and this is a model that helps us understand how to create great experiences. And it's from the big picture right down to the final interface. So each of these layers that I'm going to show you builds on top of each other, and it moves from the abstract through to the concrete.
00:16:27:04 - 00:16:45:09
Speaker 2
And it also Tim will get into it. But also if you start going through these, it starts qualifying what you should use when. And that's not just AI, that's every time. So it's a reminder as well. What the user sees is only as good as what has been...thought through underneath. That's a bit of a tongue twister, but let's start.
00:16:45:09 - 00:17:04:17
Speaker 2
Okay, so the first one is the strategy layer. Okay. This is like the bedrock. Like, what are you doing? Why are you doing it? Understanding the product objectives like this is the fundamental one and one. You have to start here. And this is where, you define everything about what is the project then scope like how is how big is that?
00:17:04:17 - 00:17:27:01
Speaker 2
So, you need to get into like like absolutely laser focus on what things do and why they do them. When you get the structure, then you have to these two, strategy and scope to dictate what the structure of this thing should be. You should have a pretty good idea. And as you can see now we're going from the abstract to a little bit more concrete as we go along.
00:17:27:02 - 00:17:49:19
Speaker 2
Now the skeleton. Okay. So you start thinking about like the user interface design information architecture, like, element layouts, buttons, all of these things start happening, and then finally you get up to the surface, the visual UI design. So user interface UI, sorry if I'm using acronyms. But this is usually the shopfront. This is what people see.
00:17:49:21 - 00:18:13:17
Speaker 2
And this is usually the fun sexy stuff okay. But that will that's just fluff unless you've got the rest of it below. So how does this translate? And I just want to walk you through, again, a few conceptual kind of tools that we use. So I don't know if you're familiar with it, but customer mapping is a incredibly important tool for us.
00:18:13:17 - 00:18:35:05
Speaker 2
So what we do is we we track users from A to Z on whatever their task they're doing, and we slice it up. This is a very simple one here. We can show someone booking a flight and getting on to a flight. So that's customer journey mapping. So it's the timeline, the key steps touch points, emotions and pain points which are super important.
00:18:35:07 - 00:18:57:22
Speaker 2
And we build these through workshops, stakeholder interviews and empathy mapping. So this is stuff that you can’t do with AI okay. You can probably get there some part of the way, but there's some real world things you need to do for good product design. The service blueprint if you think about is more or less the the happening underneath is the other side of the of of the the coin it maps out not what is the.
00:18:57:22 - 00:19:24:15
Speaker 2
frontstage moment for the customer, but also behind the scenes systems. So the roles like the processes and that deliver the experiences of the team behind it. So think about your actual company or brand, delivering. So think of it as an operational twin to the journey map, where empathy meets execution. And then if you put an end to end, solution delivery together, this is where we bring it all together.
00:19:24:15 - 00:19:57:00
Speaker 2
It's the full picture from insight to implementation and the full interaction to final outcome. And it connects dots across teams, tech time and making sure what we design actually gets delivered. Now why I'm showing you this is these are the tools that will really, be impacted by using AI intelligently and accelerate what you're doing. But again, don't just depend on the, I use it as a, a tool in your overall goal.
00:19:57:02 - 00:20:17:02
Speaker 3
To okay, so I'm gonna expand a bit on what Eric was talking about there. If we go to the next slide. So this quote here, I've used it a few times and I'm starting to think it's misconstrued at least. Or a bit unfair. So designing agentic systems requires, a shift from reactive to proactive UX thinking, anticipating user needs, not just responding to them.
00:20:17:07 - 00:20:36:22
Speaker 3
I think most good designers, service designers, product designers would be anticipating user needs. I think what Greg's actually trying to say here is it's about designing for systems that themselves can anticipate user needs. And that's the kind of shift you need to take with your thinking and how you adapt a system that will itself adapt to the user.
00:20:36:24 - 00:21:04:05
Speaker 3
If that makes sense, we'll go to the next slide. So I'm going to keep talking about this, five layers of, of design. And we're going to start talking about the strategy piece again. So here the important thing to think about is how do we take those goals of the user and of the product and make sure that we're incorporating agents or AI in a way that that's meaningful and trustworthy.
00:21:04:05 - 00:21:22:03
Speaker 3
And it really takes, stepping back and looking at what, what why does this product exist? Why does this services exist? Why does this customer experience exist? What are we trying to do with it? And where does AI fit into that? And as Eric showed you, we found journey mapping is still a critical tool as part of that, and service blueprints as well.
00:21:22:03 - 00:21:42:20
Speaker 3
These aren't new ways of thinking about service or design. But you do need to think about those new agents in the system, whether it's the person's personal agent or the agent you're building, potentially for your service. And where does that fit in, particularly thinking about those trust building moments? Where is it going to be autonomous? Where do we want to bring the human in the loop?
00:21:42:22 - 00:21:57:19
Speaker 3
That kind of things really critical at this point as well. And it's also important to start thinking about modality here as well. By which I mean, are they in the car? Is Ezra making his first request in the car? But then we show him options when he goes back to his phone. That kind of thing is really important.
00:21:57:19 - 00:22:12:24
Speaker 3
And you starting to see that more and more even in the interaction space when you interact with your phone, when you talk to Siri or Gemini or whatever device you talk to, it will offer options on a different device. And it's this is kind of using those journey maps to understand that's really valuable. We've also found a lot of value.
00:22:12:24 - 00:22:34:06
Speaker 3
Sorry, I just taken one tech and the jobs to be done framework. This isn't new either. You can go and Google what it is. Essentially, it's, talking about designing or a service or product based on a person hiring it to complete that job. So I hire Uber to take me somewhere. And this is useful because this is actually kind of how we interact with AI already.
00:22:34:06 - 00:22:51:10
Speaker 3
We say, I have this goal. Can you help me complete it? And jobs to be done already works with that kind of, thinking, okay, next one, Eric. Okay. So we we've had the kind of high level goals and now we're getting a little bit more concrete with, well, what are we actually going to do? What are the tasks that need to be completed?
00:22:51:12 - 00:23:06:13
Speaker 3
For a person to, to get through doing what they want to do, what they hired the, the product to do or the service to do. So once again, jobs to be done is useful here. And you can break it down. It breaks it down further into tasks. That's actually part of the framework there to talk about it as well.
00:23:06:15 - 00:23:23:18
Speaker 3
Thinking about agenetic or AI systems. You're thinking about, well, what tasks is the AI going to do versus what the person person's going to do? Some of these tasks are already going to be done because the personal agent already knows where Ezra lives. And everybody knows that his child loves dragons. So it's actually going to complete some of those pieces of information that the system already needs.
00:23:23:22 - 00:23:41:02
Speaker 3
So you need to start thinking about that as well what's what's already known or what can be automated prior to actually and what requires the person to interact with it. Does this have more jobs to be done? There's a thing called job stories, so I don't know. I'm sure people might be familiar with user stories if you're in digital product design.
00:23:41:02 - 00:24:02:18
Speaker 3
So a user story is as this persona, so as a parent, I want to book a party so that I can not have to worry about it or whatever. So that's a user story we've looked at shifting towards job stories, which is scenario based. So when I want to book a party, I want to do this so that I can book it easily.
00:24:02:24 - 00:24:27:22
Speaker 3
So the case, if there is going from persona based to scenario based, and I think given that there's so much personalization is going to happen on those personal agents, it's already going to know a lot about you. This kind of persona based, way of articulating requirements is going to be needed less and less, and it's actually more about the scenario that's needed and what's needed to get through that piece of the scenario.
00:24:27:24 - 00:24:49:06
Speaker 3
And then we get into the actual how does it all hang together? So actually, if we're talking this is about digital product design, but you can see it within service design and things as well. Stepping away from these linear sorts of work and ways of working through things, it's going to be much more about state based thinking, which is to say, has this information been gathered already?
00:24:49:06 - 00:25:06:19
Speaker 3
What information do we need so that we can deliver this service to this person or complete the interaction with them? And then understanding with those given states, when should the human interact? So we need more information from you, Ezra, what is your budget? We don't have a budget. Is the state there? So we ask them a question.
00:25:06:19 - 00:25:33:10
Speaker 3
And what does that look like? When does that come together in the flow. So it actually becomes less about these sort of connected nodes and they become disconnected. But they have a state. And when we need to call that part of the node or that step in the experience to get that additional information. And so it's much more, decoupled from each other and less linear as a way of thinking about how the architecture hangs together.
00:25:33:12 - 00:25:56:05
Speaker 3
The skeleton itself, this is where we're getting even more concrete. So the skeleton is actually what the content we're displaying, more or less. What are we actually going to show them? We really need to think about, confidence and explainability. So as we were saying things like annotation, where did that information come from? Is the system actually 100% confident that that's giving you the correct information with hallucinations and things like that?
00:25:56:05 - 00:26:16:08
Speaker 3
In generative AI? That's going to be a really important aspect of any kind of service design. The ability to interrupt or override choices as well. Very critical. The user should always feel in control. We want to automate as much as possible, but it should feel like they automated it. If that makes sense. It should be something that they initiated and they're comfortable with everything that's going on.
00:26:16:08 - 00:26:51:18
Speaker 3
And if there's some critical moment that you know off within a service, make sure you're allowing for that interruption or override to occur. And a really great one that I love to talk about is this machine readability. So accessibility has been an important part of digital product design for a long time now. And it's actually coming to show that the companies that did this well are able to ease into, AI and agentic systems more easily because the systems can actually read there, what's happening on the screen more easily because you've already put in these Aria labels or this kind of technical back end so that the system knows what it's looking at.
00:26:51:20 - 00:27:13:21
Speaker 3
And this is working well with agentic systems. It also works well with that, modality as well. If you think about that, you set up a scene, a digital product for a person with a screen reader who can't actually see the screen. That means if you're in the car, this is the system can actually use that to help talk to the person rather than display things to them as well.
00:27:13:23 - 00:27:32:06
Speaker 3
And then the actual surface itself. So the visual elements of it, the polish I guess. So there's the visual UI. But there's a, this is the personality as well. So if you think about that agentic ripple effect, there's going to be a series of agents that talk to you. So the personal engine will talk to you in a certain way.
00:27:32:06 - 00:27:49:02
Speaker 3
But when you start to talk to bookings.com, that agent itself is probably going to have a personality. And some of that will come through in visual elements as well. But there's going to need to be much more thought put into. How does that agent talk to you? How does it like guide you through things as well? And like how slow or fast is it?
00:27:49:02 - 00:28:08:10
Speaker 3
How automated or hand-holding is it? How much does it ask you? That personality is going to be something that needs to be thought about it, that kind of sensory level as well, and adapt. So one more thing on the adaptive UI elements. This is for the hardcore design people in the room. Think about your design systems as not being built by you.
00:28:08:16 - 00:28:22:07
Speaker 3
Imagine, an agent actually pulling parts of your design system together to create an interface as needed for that person. In the given context. So adaptable. Design systems as well. Sorry. Very crafty.
00:28:22:09 - 00:28:43:04
Speaker 2
UI design there. Yeah. Thanks, Tim. I was a bit fast on that. So we’ve given you a theory. Let's put it into a case study. So what we've done here is we, as a company, we've worked a lot with airlines over the years, and we've just put together essentially a a mock up of a you taking a flight?
00:28:43:06 - 00:29:03:18
Speaker 2
As you can see here, it's, about trip planning and actually getting to the flight. And this is Meera our persona. So the main point here is travel logistics, not the flight caused the most stress. And also the doing the 50 things before I even board. So we're going to look at how AI can improve, CX.
00:29:03:18 - 00:29:32:20
Speaker 2
But let's first give myself a migraine even thinking about this. So I know everybody has probably done this at some stage or other, but like, the, customers, like Meera face over like a hundred micro, decisions before a trip and 70% say in consistent experience experiences, increase their stress. So as you can see here, like, all of these things happen, I'm sure everyone who's ever gotten a flight has had something like this happen to you.
00:29:32:22 - 00:29:57:06
Speaker 2
And I don't know, looking at this, I get I'm actually getting PTSD. So let's move on and see how that can be cured. So how can AI be the invisible copilot? And this is a key part of it. Like the AI in customer experience is there ultimately to to the ultimate customer experience. Like that first class, experience like everything is done almost in the invisible.
00:29:57:08 - 00:30:28:24
Speaker 2
Like Tim touched on it there a lot like even with the brand voice, like, how do you design that? How do you design, speech? How do you design conversations? This isn't visual, just, how you actually manifest in brand, but using predictive analytics and segmentation, we can do things like smart discovery, like AI analyzes Meera’s, travel history, business travel preferences and sends them, her a targeted email, with a discounted premium, seat upgrade on a route she frequently, flies.
00:30:29:01 - 00:30:56:19
Speaker 2
And then there's effortless booking, like a conversational AI assistant, just like I need, like, very much like booking the, the, Ezra’s birthday cake. So the natural language processing and preference learning is a key part of that. So these virtual agents can handle tasks like searching flights, comparing prices, booking tickets, even suggesting alternative options, and then applying your loyalty points automatically.
00:30:56:19 - 00:31:21:20
Speaker 2
One of my pet peeves, having to do that manually, but this is using particular types of AI. Because you hear AI, there's so many flavors of AI, you need to know what you're using and why you're using it. Go back to our layers. Think about this down at that very strategic level. Then pre-departure, like using an AI powered, virtual assistant and contextual support, we go from reactive to predictive.
00:31:21:22 - 00:31:52:13
Speaker 2
So AI keeps her one step ahead. It's proactive context rule like ten touched on it like where is she? What's she doing? Like that understanding of where you are physically is such a key, aspect of future, AI services. But it's tuned into your needs so you can leave the house, calm and not chaotic on waiting for that day personally, but, next we have AI powered notifications, a proactive customer experience.
00:31:52:15 - 00:32:12:01
Speaker 2
So she'll once she's at the airport, the assistant keeps her effortlessly in control. So there's a gate change. She's alerted to it, but with enough time to adjust. Walking time to the lounge. It's all mapped out in the app, and even boarding groups are updated in real time. So with AI, the airport shifts from a confusing maze to a guide personal journey.
00:32:12:01 - 00:32:39:13
Speaker 2
It's calm, clear and connected. I remember on going back to that like first class experience, where everything is anticipated, then finally predictive analytics and agentic AI even in mid-air. So, Meera stays in control. If her her connection is, missed. AI rebooks her automatically. Her bags tracked without her asking, and her curated in-flight entertainment is all we cued up based on past preferences.
00:32:39:15 - 00:33:02:19
Speaker 2
So after landing, she receives a tailored check in, and not a generic survey. So feedback is so important, but it's very hard to get it. We've all been gotten those forms on something we want to follow, but imagine that it's a natural language conversation. You can get so much more quality insightful feedback that way. So it's predictive AI. Predictive agentic AI in action.
00:33:02:21 - 00:33:26:23
Speaker 2
It's going working quite smoothly in the background for the journey. Even when things go off script. So every touchpoint is deeply personalized yet feels natural. AI anticipates oops, sorry, I jumped on there. AI anticipates her needs and is reacting to issues and the experience again is seamless across devices. This is the actual goal. So and in physical environments.
00:33:26:23 - 00:33:49:11
Speaker 2
So as a customer you feel valued and not tracked. So it's all there for your end benefit. So I know we're at a blinding pace. Natasha, I think we're almost there, but I just want to quickly recap of what we're doing. Those of you who are designers out there, I just want to just highlight two stats.
00:33:49:11 - 00:34:07:17
Speaker 2
And this is a little bit old to this, but McKinsey and company did a business value design design is often the invisible thing in a company. It's wanted, but it's very hard to track what it is. But they found that and this is a few years ago, and I would imagine this will go up, in time. And I'll tell you why in a second.
00:34:07:17 - 00:34:40:21
Speaker 2
But like, companies with strong design capabilities achieve 32% higher revenue growth and 56% higher, total returns to shareholders over five years compared to industry averages. Think of Apple, Nike, any of these big companies, they've got design at their core is their X factor. And with and why I’m saying that this should be something that keeps you, the AI future, positive is because we see lots of tools coming in, like Lovable, Bold and the rest of them.
00:34:40:21 - 00:35:03:01
Speaker 2
And what they're doing is they're able to create, design automatically. However, you need to have design thinking, as we've shown you, to create quality outputs. So as long as you have that design mindset and design thinking, you're golden. So why do I say this? If you look at this here, I took this off LinkedIn recently, but it really sums up what I'm seeing at the moment.
00:35:03:03 - 00:35:23:13
Speaker 2
So if you look at, you're seeing a lot of AI out there and essentially it's just called AI slop, it's just looks great, but it's not really that valuable. However, if you're going to actually create something with value, your interaction should be minimal. It should anticipate what you want and should, facilitate your end goal with the minimal of interaction.
00:35:23:13 - 00:35:44:00
Speaker 2
And that's where strategy comes in. And design thinking. So I think like in a few years, what you're going to see is you're going to see a massive, wash, of we're going to be flooded with AI stuff, and people are actually going to start craving the real. And that's going to be probably the pushback against it.
00:35:44:02 - 00:36:04:20
Speaker 2
And in summary, I just wanted to say is that, what you actually need for all of this, for ultimate customer experience is you do need to think about it strategically. You do need to bring in design thinking, and you can use AI to facilitate your end goals. And that makes for happy customers. The end.
00:36:04:22 - 00:36:21:06
Speaker 1
Thank you so much Eric and Tim. That was amazing and so insightful. I can definitely relate to Meera within that case study as well. Just looking back to the times when I've had to book a lot of flights but we've received so many questions that have come through. Thank you so much to everyone. That's that's shared their questions.
00:36:21:06 - 00:36:45:08
Speaker 1
So we'll try and get through as many as possible. But to kick us off, Eric, I know that you mentioned that customer mapping, that whole process is so important because it's not something that you can do on AI. How do you ensure and how to how do CX teams ensure that, as they're mapping it remains relevant as customer needs evolves that kind of that that's a really important part of that journey mapping process.
00:36:45:10 - 00:36:52:16
Speaker 1
So how do you make sure throughout that whole period that it remains relevant as customers change that?
00:36:52:18 - 00:37:15:22
Speaker 2
That's a fantastic question. And I'll hand over to Tim after my little bit here, because there's a lot of this. But look, customer interviews are so critical and so important for any successful, product delivery. It's so easy to to work in the corner with your design. I'm guilty of it being like, my my skills are our design, you know, visual design.
00:37:15:22 - 00:37:31:16
Speaker 2
And then if you think about there's a sliding scale from UX, which is the very analytical to UI, which is the very visual. I'm kind of in the middle. So some people would be more towards the UX end, which is the analytical, not so much visual, and some are very visual, and I've lived at the very visual end.
00:37:31:18 - 00:37:52:16
Speaker 2
But if you want to be like get a good outcome, it's in the middle and talking to people and validating what they say is, is critical because you can take an opinion. But I wouldn't run with an opinion. But it's also about talking to many, many people. And there's qualitative and quantitative, interviewing and research as well.
00:37:52:18 - 00:37:56:13
Speaker 2
Actually, Tim, I'll hand over to you. I'm blabbing about this, but it's more your specialty.
00:37:56:14 - 00:38:16:15
Speaker 3
So it's really great question because I didn't really get to touch on using AI for design or for customer experience as well. So where there's this, it's a really exciting space, the customer journey map. And it does change. And right now, well over the last however long they've been around, it's to kind of a artifact that you produce.
00:38:16:15 - 00:38:35:05
Speaker 3
And it's at a point in time so that that this is the current journey. Here's all the parts where it's bad, here's what we want to fix. And it kind of is static at a point in time and your point towards what you want to change to improve it. But I think agentic systems are heading towards and you can see this with dovetail, which is a, user analytics system.
00:38:35:07 - 00:38:52:01
Speaker 3
Is this kind of living artifact where you might, in fact have your customer journey there, but it's actually taking in data in real time and telling you what's going on right now in that customer journey at those different points in time, and aggregating the data that's coming through, whether it's from surveys or usability analytics, that kind of thing.
00:38:52:03 - 00:39:05:23
Speaker 3
So this kind of I call it a living artifact is this thing that's going to start coming forward, I think, where it really will start to become more real time, and those artifacts will become even more important because they'll actually be a lens into what's happening now, not just, a point in time.
00:39:06:00 - 00:39:20:05
Speaker 1
Awesome. Thank you so much, Tim. That's that's yeah, good. Good insight. The next question that we had was, how can design UX and CX intertwine together to create personalized and unique experiences for the users?
00:39:20:07 - 00:39:44:04
Speaker 2
Well, that's a big question. No. So yeah, that's huge. I think it's it's a pretty logical follow on from what we've just said, is that you need to you need to map the pain points that are probably the most important thing. So there's a bunch of user methodologies out there. And like I saying is you can pick ones that work for you.
00:39:44:04 - 00:40:20:15
Speaker 2
I use the business model canvas and value proposition design. Those paradigms work for me. You know, there's other ones you can use as well. But if you can identify, genuinely identify a pain point and solve that pain point, that could be a product in itself. But the more pain points you can, identify and create solutions for and smooth the road for them, like in that, that the, the user journey that we showed you that makes for the ultimate customer experience, because ultimately, the ultimate customer experience is that they get what they want in the end without any hassle, you know?
00:40:20:16 - 00:40:29:24
Speaker 2
So it just happens, which is the magic of AI in some ways. And what Tim went into as well. Tim.
00:40:30:01 - 00:40:47:22
Speaker 3
I think as human beings you can't eat the elephant. So that's why we we went through that layered process because it really helped. We were struggling because there's so broad to think about it and putting it into those layers really helped us to actually go, okay, well, here's the important bit. When I'm looking at this bit, which is the surface design, here's the important bits.
00:40:47:22 - 00:41:05:01
Speaker 3
When I'm starting to look at the strategy and the scope, the sort of more conceptual stuff. So yeah, as a person trying to bring all those things together to try and break it down like that, having said that, AI is really good at eating the elephant. You can throw a heap of data at it and then start breaking it, helping you break it down for you.
00:41:05:01 - 00:41:15:10
Speaker 3
So if you have a bunch of data already, chuck it in there. Stop mucking around with it to try and pull it out and break it down. Something you can actually take action on. However that maybe.
00:41:15:12 - 00:41:27:16
Speaker 1
Thank you. And those frameworks are so useful as well. So thank you so much for sharing those and including them in your presentation. The next question that we had was, what tools would you use to create a prototype?
00:41:27:18 - 00:41:48:09
Speaker 2
Brilliant question. Right now, like use everything you can get your hands on. AI design tooling is I think it's it's just beginning, but there's a few obvious ones out there. There’s Lovable, Bolt, which essentially is the same thing. I think it's data...databutton. I think it's called. Sorry. I'll get back to you exactly what that one is.
00:41:48:09 - 00:42:10:08
Speaker 2
But these are, oh and Figma. Figma have just published make as well. And we've been using with, And the key to actually using these is learning how to write prompts correctly. Now, what we showed you is actually the way you can start structuring your prompts as well. So Tim talked about the jobs to be done, user stories, etc..
00:42:10:08 - 00:42:29:15
Speaker 2
Thus, if there is one thing I would recommend to everyone who's using this tool and is really, really learn how to prompt correctly, there's frameworks out there and it's a moving it's a moving space because I'm constantly on LinkedIn where I get all my, my hacks, and I don't pretend that I come up with any of these things.
00:42:29:15 - 00:42:50:13
Speaker 2
I usually see someone else and I grab it. But if you can instruct your AI to do and importantly, what not to do, and then if you have done that through that layered process that we talked about our use, whatever works for you. Again, what that means is that your end delivery from the AI shouldn't be a surprise to you.
00:42:50:19 - 00:43:07:02
Speaker 2
Like there might be a few nice things in there. However, it should do the job it needs to do, like a jobs to be done. So, Lovable, Bold, and above all, play with everything. Like play with everything you can, but play with intent.
00:43:07:04 - 00:43:28:09
Speaker 1
I've heard that prompting as well is is so helpful, especially when you're playing around with with AI and just making sure that you're really specific as well. So some good tips then even just using LinkedIn as well. The other question that we had was, what advice would you give to someone trying to get leadership buy in for investing in AI and CX design?
00:43:28:11 - 00:43:56:16
Speaker 2
Great question. Always base your your I guess your case on business outcomes like its bottom line. Always. So when you do go to them often the and this is where like developers are guilty of it, designers are guilty of it. And we show something that is a cool design and we show something that's a cool piece of code, but we don't see what the actual end business benefit is.
00:43:56:22 - 00:44:24:12
Speaker 2
So you can you can do the same thing, but always, always table it in terms of the ultimate business outcome. Because once you start saying we can do X, Y and Z and it will impact our bottom line or our productivity or our sales responses, this way it makes sense to the person you're talking to. One of the things that we do is we tailor our, I guess, our sales pitch, our benefit pitch to who we're talking to.
00:44:24:12 - 00:44:48:15
Speaker 2
So when I talk to a CEO versus a CTO versus a CMO versus a COO, it's a very different, proposition I'm putting in front of them, because what what the CEO wants to see is completely different to a chief marketing officer. So you need to table it in the business terms of what they understand and in the, in the, in the role that they are fulfilling.
00:44:48:17 - 00:44:53:11
Speaker 2
So put yourself in empathy. Put yourself in their shoes.
00:44:53:13 - 00:45:11:05
Speaker 1
Great. Thank you so much again. Yeah I know that that's super relevant to any industry or any role that you work in as well. So thank you for sharing. Thank you so much everyone for joining today's session. It's been so great to see that we've had so many attendees join and really appreciate all of the questions and interactions that are coming through the chat.
00:45:11:07 - 00:45:30:15
Speaker 1
Again, a big, huge thank you to Eric and Tim. The presentation was amazing, and we really appreciate you putting all of that together for us. If you would like to explore the subject further, please check out our Customer Experience, Strategy and Design course. That's actually commencing next month on July the 14th. And we've also recorded the session as well.
00:45:30:15 - 00:45:41:17
Speaker 1
So please keep an eye out for it, out for it in your emails and you can watch in your own time too. So thank you again. Really appreciate it everyone, and we hope to see you at our next masterclass. Thanks all.
