Jason Maynard, APAC CTO at Zendesk, joins Georgie Healy for a conversation that cuts through the AI automation hype and gets to what is actually happening on the front lines of customer service.
Jason shares the framework every business needs right now for deciding where automation genuinely helps customers and where it quietly destroys trust. He introduces the dolphin problem, the counterintuitive reason why the brands that listen hardest to complaints end up winning the most. And he explains why, despite everything you are hearing about AI replacing jobs, customer service job postings in the United States went up 10% in 2025.
They also get into digital snap and how to design your way out of it, why your AI agent is really just an extension of your brand identity, a brand new role emerging inside service organisations that looks a lot like what happened to marketing in the early 2000s, and three rapid fire scenarios that reveal exactly when you need a human and when you do not.
This is one of the most grounded and practical conversations we have had on the show about what AI in customer service actually looks like when it is done well.
Transcript Synced · click any line to jump ▾
Georgie Healy: Founders scale faster on Deel. Set up payroll for any country in minutes. Hire anyone anywhere. Get visas handled fast and get back to building. Visit deel.com/dayone. That's D-E-E-L.com/dayone.
Jason Maynard: Every time you have someone reach out to your brand, if that's a customer having an issue, I always call it a dolphin problem because when you see a dolphin on the surface, you know, of the water, you you know that there's like 20 dolphins in the pod below that, that you don't see.
Georgie Healy: Yeah.
Jason Maynard: Right? I think customer service is very similar to that.
Georgie Healy: By the time I'm reaching out to customer service, Jason, I'm already mad. I'm so mad. How do you prevent something called digital snap?
Jason Maynard: For those failure modes, you know, automation is probably the worst possible thing you can put between your brand and a frustrated customer. Your agents are a reflection of your brand. You can think about it as like designing your website. What's your brand presence? How do you show up to the world and how do you show up to your customers? You want your agents to show up in that same way.
Georgie Healy: Hello and welcome to In the Blink of AI. This week we have an international guest. We've got Jason Maynard, the CTO of Zendesk. We're talking to Zendesk because Jason is the absolute professional expert on customer service and AI agents or bots. And this is the future, guys. The way we work and live on the internet will be with customer service agents. We'll even have our own in the future talking to other customer agents. This stuff matters and these insights you're going to see play out over the year. We've got a few little tidbits for you here though. What is the direction of customer service jobs going in? What's the worst thing you can do with a furious customer? And what of the 3 buckets of conversation, where should AI never play in? Your agents are a reflection of your brand. So after you listen to this show, the next time you speak to an agent, you will know whether it has been well made or terribly made. And I think that matters for the future and the way you you choose your brands that you work with. Uh, you've got two bogans on the show today. You've got me from Queensland, and you've got Jason, self-proclaimed bogan because he lived in Melbourne at the age of 18, drinking VB, driving a ute, welding stuff— all the things that I do, really. Uh, I can't wait for you to hear the show. Let's dive right in.
Jason Maynard: You're listening to a dayone.fm show.
Georgie Healy: Jason, thank you so much for coming on In the Blink of AI. I've been really excited about this. You're visiting us in Sydney, but before I dive into the interview, is it true you have a craft brewery and winery in California?
Jason Maynard: Do. That's the side gig. I don't get to spend as much time on it as I'd like, but yeah, that's the, you know, it sort of scratches the itch of the non-tech side. Making something tangible, you know, bringing something to Ojai where I grew up. And I think my mom gets to spend more time in the brewery and winery these days than I do. But yeah, it's a really fun sort of like side project, hopefully retirement plan.
Georgie Healy: Yeah, I'd love to know the thinking of like, do we own a brewery or a winery? And you're like, how about both? Like, how does that work?
Jason Maynard: Yeah, so it started with beer.
Georgie Healy: Okay.
Jason Maynard: I have a really good friend from high school who was, you know, grew up, you know, coming through restaurants, was an amazing baker. And then the beer that we make is pretty unique. It's technically Gruet. It's sort of flavored with non-traditional ingredients. So instead of using hops, we use florals, sort of like botanicals that grow in the Ojai Valley. So it's like instead of hops, we use sagebrush and wild sumac. We eat a lot of pixie tangerines grow in Ojai. So we make a wheat ale. It's made with like a pixie tangerine marmalade, pixie tangerine blossoms. So yeah, it's pretty non-traditional, but if you come to Ojai and you sort of love that place, it's very hyper-local. And then we did the same thing with the wine. We've made wine for 2 years and we had this amazing winemaker who had been working in Ojai and we just pick little plots of land that are growing grapes around Ojai Valley and we use those to make a very local wine from the Ojai Valley.
Georgie Healy: It sounds beautiful. Like it sounds lovely.
Jason Maynard: It's a very lovely place. Ojai, California, just north of LA for the Australian audience.
Georgie Healy: Yeah, yeah. Paint a picture for us. Okay, so when you're here, is there any beer or wine that you like or are we behind the times here?
Jason Maynard: No, so I took a gap year, lived in Melbourne for a year. I worked as a welder. I had a ute, I had an apartment. It was good times. And I think the sort of thing I would do on Fridays was I'd drive the Ute, we'd get off work, and I'd get a slab of VB, which I think is what you call it. And so that was sort of the beer of early choice when I was 18. That's definitely stuck with me. My grandpa loved Crown Lager.
Georgie Healy: Yeah.
Jason Maynard: So when I wanted to get fancy, we'd go get a case of Crown Lager and I'd take that home. But so VB has stuck with me. I don't know if it's still popular, but I think you call it a stubby.
Georgie Healy: Yeah.
Jason Maynard: So I'll have a VB stubby. And then last night I was having a Cooper's.
Georgie Healy: Yes.
Jason Maynard: The pale ale.
Georgie Healy: Yes.
Jason Maynard: Which is lovely. So yeah, I've, I've, I think you'd technically call what I, I, I was doing back then, my friends now have been like, well, you were just a bogan.
Georgie Healy: You were a bogan.
Jason Maynard: I was a bogan.
Georgie Healy: I'm sorry, that's definitely bogan behavior. But as a conflict doc, that is a safe space.
Jason Maynard: I might still be a bogan.
Georgie Healy: That's okay.
Jason Maynard: Deep down inside, 'cause I still love my VB. I'd probably still like drive a ute if I could.
Georgie Healy: So these are very iconic. Like if I was to generate an AI image of an Australian white male, like you kind of ticked every box there.
Jason Maynard: Yeah, you know, I started young.
Georgie Healy: Yeah, you nailed it.
Jason Maynard: It's been a while, but it stuck with me. Yeah.
Georgie Healy: Very cool. And then when we talk about where you're at now at Zendesk, for people just listening and not aware of your LinkedIn profile, seems like a jump. Seems like a jump from like living in Australia with a ute welding and very technical, very on the cutting edge of AI and agents. Tell us where you're at now.
Jason Maynard: Yeah, so it's been a few years, believe it or not, since then. The time I had here was very much, you know, how do I get hands-on? How do we build things? I went, you know, my background is in manufacturing engineering and systems engineering. And I think through that, I really developed a love of math and statistics, like sort of how to apply that in the real world. I spent some years in consulting, so I think the original reason I went into consulting was I love this idea of sort of how do we connect, you know, sort of like early internet, how do we connect all these instruments we have in the field and then use that data to be proactive and actually drive sort of like real outcomes for customers by looking at the log data and doing proactive maintenance. And so I think I sort of like moved from, you know, manufacturing to data to software and connectivity. And then that really sort of led me to Zendesk in the early days. You know, Zendesk, I joined in 2012. We were growing like crazy. And, you know, I think back then you didn't have all the amazing tools you have now for product analytics and how to understand user behavior. So the first thing I did when I came to Zendesk was, you know, we had just started our MySQL database. We sort of lost visibility into everything customers were doing on the platform. And so built some of the first tools that we used to bring that data together. And we just had this like amazing opportunity. We had 30,000 customers back then, and customer service is just such a sort of data-driven business and sort of function. And I think we had this insight that, you know, if you were sort of a big enterprise, you could go hire Bain or McKinsey to sort of benchmark how you were doing against competitors.
Georgie Healy: Mm-hmm.
Jason Maynard: But we had this amazing insight with 30,000 customers across different industries, like how are they performing compared to their peers on customer satisfaction, first reply time, resolution times? And so we benchmarked a lot of those things and then just share that with our customers and they could go, oh, this isn't actually an area where I can sort of improve and drive the business forward. You know, and back then it was, it was simpler times. It was like a chart's good, but a chart with a target of like the 90th percentile for your industry, you know, is even better. That sort of gives you a goal to work towards. And so sort of like putting those systems in place was sort of my first job at Zendesk, really.
Georgie Healy: So a lot of human elements you keep bringing up, despite the fact it's obviously a very technical, data-driven process.
Jason Maynard: Yep.
Georgie Healy: Was there any specific data point that customers weren't aware of that you had to keep being like, this is important for you guys?
Jason Maynard: Yeah, I mean, there's a lot of sort of counterintuitive data in service, so, and it's very unique to each business. So like a good example was one of the early data products we launched was something called Satisfaction Prediction. We had all these customer satisfaction surveys, but we also had all this operational data around, you know, what led to that outcome. And, you know, back then we'd build like one logistic regression model per customer. But you could see actually what weighted higher for different customers. So the red thread was time between responses and first reply time highly correlated with positive outcomes. But then there were sort of customers that had counterintuitive data. So we worked a long time with a customer, Upwork, and they work with these freelancers, people that don't necessarily have a team around them. And most of the times, the number of touches on a conversation would be highly correlated with negative satisfaction. But for them, like the more they engaged in conversations with their customers, the higher the likelihood was it would actually be a positive outcome. So, you know, every business was a little bit different. There were sort of like red threads that sort of like patterns that were highly correlated with, you know, positive outcomes, but definitely businesses that were sort of unique in the way that, you know, customers wanted them to engage as sort of like with service as the face of their brand.
Georgie Healy: Super cool. And we'll get into like the evolution of that into the AI space. Space and agents and things like that. You mentioned having 30,000 customers in 2012, is that right?
Jason Maynard: Yeah, 2012. So bit of just like origin story. Zendesk started in 2007, 3 Danish founders started on a literal Zen desk in—
Georgie Healy: What's a Zen desk?
Jason Maynard: Well, so it's, it was a, I think it was a door in, so 3 founders, Mikkel, Morten, and Alexander, in an apartment in Copenhagen, and Morten had literally like a door that had been converted into a desk that they built the first version of Zendesk on top of.
Georgie Healy: That's very chic.
Jason Maynard: And that was, you know, very Copenhagen. Yeah. And, you know, they had all come from consulting backgrounds as well. I think they had just seen, you know, they'd been implementing this software. You wouldn't remember any of the software. You're too young. But so young in the '90s and 2000s, like service software was not very nice. You know, there was, there was sort of— it was built for, you know, an institution to manage your process, manage your people. And the actual customer experience of that was sort of an afterthought. It was like, how do I manage the metrics? How do I manage my people? And, you know, Zendesk was built with sort of a very sort of like simple idea, which is what if we start with the customer experience first? Like what makes a great customer experience with between a customer and a brand? And, you know, how do we make that more human? And then how do we build back from that into sort of like the different parts that make that successful? You know, and it was a lot of the things like what makes a great relationship, right? It's like you need to be transparent, you need to be timely, you need to sort of like build trust. And how do you sort of embody those sort of like human elements in the software and the experience that you deliver through customer care? That was really how the product started.
Georgie Healy: Mm-hmm.
Jason Maynard: And then, you know, I think we fell in love with the other problems that sit around that. You know, I think the, the job of a service agent is just really, really hard, right? Like you're dealing with people at their most frustrated, you know, at their most vulnerable when they need help. And we just saw so many customers where, you know, things in the organization got in the way of just being a great brand and being caring to your customers, you know, whether that was, technology barriers, policy barriers, and really built Zendesk around like, how do we solve those problems, you know, from the customer to the people really supporting those customers and make that a really core part of the brands that we grew up with. And I think a lot of our customers were thinking about service differently. Yeah, we grew up with like great brands like Uber and Groupon and Warby Parker, Zip, MYOB, Canva. Like they're all brands that have sort of changed the way that you think about service. You know, and it's much more at the sort of center of how they think about delivering a great experience. And, um—
Georgie Healy: And I'm jumping ahead here, but you guys clearly care so much about that interaction.
Jason Maynard: Yep.
Georgie Healy: How do you possibly use all those years and experience and care and then let an agent do it? Like, I guess you guys are very, like, are watching very carefully.
Jason Maynard: Yeah.
Georgie Healy: I know I'm jumping ahead, but I think listeners would be screaming if I didn't ask this.
Jason Maynard: No, and I think it's like the, right now it feels like we're, you know, we're in this pendulum swing towards, you know, automate all the things.
Georgie Healy: Yeah.
Jason Maynard: Right? And it's, you know, it's an amazing, we're in an amazing time in technology, but it also feels like we're at a time where, you know, there's one hammer looking for a lot of nails. And I think our most successful customers are the ones that, you know, think about that in a much more nuanced way. And where is the right places to insert automation where it's additive to the experience and where are the right places where you still need to engage with a human touch. We are just seeing that whole evolution happen. And I sort of— I grew up like working a lot of restaurants, like in hospitality. Right. And so, you know, these are industries where you want to engage with the customers. You know, someone walks through your door, you know, and you want that personal touch. You know, it's not like a restaurant is running on larger margins than a lot of the businesses, you know, around the world, the customers we work with. But it does feel like sometimes we're doing everything possible not to actually have a human in the loop in terms of like helping customers when they really need it. But I think the thing that's happening right now is like where if you can be really precise about where you deploy automation, there's a real opportunity to uplevel how you engage at the right times with customers.
Georgie Healy: Mm-hmm.
Jason Maynard: You know, and you can do that in, a more cost-effective way. You can invest in people in a new type of way. You can— we have a lot of our customers that are rethinking the parts of the customer base that they support. And maybe you have a very, very large number of free users on your platform that you've never been able to provide sort of like hands-on support to. But if you can drive, you know, much more efficiency with automation, you can actually engage with your customers in a much broader way. And so we're, you know, I think we're helping a lot of our customers sort of design that experience from day one. And figure out where do we deploy automation where it's really additive.
Georgie Healy: Got some very pointy questions.
Jason Maynard: Yeah.
Georgie Healy: When do you think customers would prefer to speak to a chatbot? And I'll start there, I'll start there. For the most part.
Jason Maynard: Yeah, so I always think of it as there's sort of like 3 types of service interactions broadly. There's a failure mode, so something's gone wrong in the experience and something needs to be fixed. That's sort of category one, you know, from our benchmarking, that ends up being about 30% of all the interactions that come into a sort of a service organization. Then there's sort of, you know, I call them like transactional interactions. Like someone just needs to get something done with your brand, with your company. Think about a return and an exchange and address change. These sort of like common questions where it doesn't necessarily help to have someone in the loop, you know, having someone who's sort of like empathetic and there and accountable. Typically slows down those types of interactions. You're waiting in a queue, you're waiting for it to just get something done. And so that's sort of like the second bucket. And then third bucket are the places where, you know, someone really needs to count on your brand or you can add real value in terms of, you know, being consultative, helping them, you know, solve a problem, helping them identify a new opportunity in their business. And those are really relationship building, you know, helping with retention, helping with, you know, growing the value that a customer is getting from your products or your services. If you take those 3 buckets, like the, that middle bucket of sort of transactional services, like that's the no-brainer, right? It's like you can introduce automation there and enable a customer to self-serve when they just need to get something done and they just need sort of that efficiency. That's sort of the best place, you know, for failure interactions. We always talk about how do you, you know, solve that interaction in real time for the customer. But then go upstream to like, what caused that issue? Is it a bug in the product? Is it a performance issue? Is it confusion on a campaign that you're running? So a lot of those things you sort of want to take upstream and find the root cause and actually usually solve it outside of your customer service organization. A lot of that, that bucket is like, how do we communicate back to the organization? Like where our customers are really struggling with our brand, with our products, with our services. And then that, that sort of like final bucket of, you know, sort of high-value interactions. Like how do we put, you know, the best people on this? How do we really engage customers there? And those are probably the places where you don't want to introduce automation. You know, someone's really looking for, you know, accountability with you as a brand, you know, or looking for you to be empathetic or provide judgment or, you know, provide sort of best practices. And so I like to think of it in those 3 buckets and, you know, be really selective around, you know, where do we want to deploy automation? Mm-hmm. And then where are the places where we still want that human touch? And then where do we have to sort of like work with the rest of the organization cross-functionally to sort of like solve the root cause of customer problems?
Georgie Healy: I mean, I'm being so cheeky, but the whole time I'm thinking, I bet you've seen some hectic conversations with chatbots. Like I read 800 million AI interactions. Any conversation with an agent that comes to mind of being extra funny or interesting or different or unique?
Jason Maynard: I mean, there's definitely unique ones and I think there are definitely changing. I think we, we all have certain interactions that we've had that are, you know, what you never want your brand to put forward as service, right? You know, I think we've all picked up the phone and been stuck in an IVR, you know, forever and waiting on hold. I think we've all had, you know, an agent interaction that feels like it's looping and you're not getting anywhere. You know, the handoff where, you know, there's never, you know, the context about the whole experience you've been through. From trying to self-serve and get information to interacting with an agent to then getting connected to someone, and you're having to repeat yourself at each one of those stages. So we all have like the bad experiences.
Georgie Healy: You're telling me you've had bad experiences even knowing this? Oh, definitely.
Jason Maynard: Have you been mean to a customer? I'm probably the most ruthless. You know, I'm known to, you know, if I'm going out with one of our customers, I'm often, we call it like secret shopping. Like I will actually go through their whole service experience, you know, go buy something, see what it's like, try to return it and sort of see what that's like. And oftentimes that's sort of like slot filling for me. I'm like, oh, you know, this is places we can help you improve. So that's sort of front lines. But then I think we have the other side of it as well. Like I think the ability to handle nontraditional questions has really changed. So I don't know if you have this brand here, but we have a brand in the US that's called Noble. They're like a shoewear company that do They do strength training shoes, running shoes, golf shoes, so sportswear. And they've deployed our AI agent on their website. And it's the out-of-the-box things that you can do are pretty standard, right? You can initiate a return and you can change an address or you can check on shipping status. So all of that's automated within the agent, but then you can go off script. You can just ask it a question. So I'm on the road a lot. I'm always on planes and I always have one pair of dress shoes and then I need one pair of something to work out in. And then I need another pair that I can sort of like pair of shoes that I can wear casually. So I was talking to their AI agent on the website and I was asking it, what shoes could I work out in but also are sort of casual but dressy that I could wear? And it started sort of riffing with me of like, oh, you should try this line. it's sort of like solid colors. And I was like, what shoes can I pack really well, even in a small carry-on bag? And it's actually able to totally go off script and like help with these sort of like other questions that, you know, are sort of like out of what you think of as sort of like common service and really sort of like helping me understand more about their product catalog and, you know, you know, what might be the right fit for a product for me. You know, and that's all happening without interacting with one of their customer care team. And they've sort of designed that experience in a way. And it's, you know, it's this sort of like magical real-time experience that you can have on their website, like chatting with a product expert about their product catalog.
Georgie Healy: Is there any benefit to training it to be silly, like, or being silly? Like, is there any benefit, or is that just the most rogue customers that you don't even need to help? Like, if I'm like a shoe that helps me make friends with the animals, like, is there any benefit to training a model in that way?
Jason Maynard: I mean, Your agents are a reflection of your brand.
Georgie Healy: Okay, so it depends on the brand.
Jason Maynard: It depends on the brand. You know, it's like—
Georgie Healy: It's a playful— If you've—
Jason Maynard: if you're a playful brand and you have, you know, it's sort of like you can think about it as like designing your website, you know, like what's, what's your brand presence? How do you sort of like show up to the world and how do you show up to your customers? You want your agents to show up in that same way, you know, and that could be you know, in a sort of like playful, you know, silly way, you know, and, you know, the amazing thing is, you know, through the agents you can say, you know, we're, you know, we want to be playful. We want to use emojis in responses. We want to sort of be silly and witty in the way that we're responding to customers. That's sort of that thinking holistically about the experience you want to give to your customers. What is our brand? How do we want to help them? You know, being, you know, very contextually aware about, you know, how they're going to use the recommendations that you have. You're going to, you're going to design that agent experience in a very different way.
Georgie Healy: Amazing. And you've got so many customers now, right? Like, especially since you started in 2012.
Jason Maynard: Yep.
Georgie Healy: Are there any trends with these customers and their roles or anything like that that's evolving or changing?
Jason Maynard: Definitely. Yeah, I think the, you know, I'd say like two things. Like, one is I think we're seeing a very new role come up within a customer service organization. It's a, you know, customer service for the longest time was a people business. It was like, how do I find great people who represent my brand? How do I hire them? How do I train them? How do I retain them? And how do I, as that team grows and we scale as a business, how do I keep bringing those people in and sort of like scale through people? I think we're hitting a point where a large portion of those interactions and that design experience is happening through technology. And that's taking a very different role that sits within operations. You know, it's, it's, we have all these like, you know, for years I've worked with these like highly creative operators and admins that are technical enough to be dangerous and they know what they need to do, but they need the right tools to be successful. And I think we're really seeing that progression happen right now. You know, these sort of like, everyone calls it something different. You know, I was just talking to a customer yesterday. They call it like a service designer. I've heard it called like an automation engineer, an AI architect, but it's this sort of like you're part data analyst to sort of like understand where the problems are. You're part, um, sort of experience designers, sort of thinking holistically around, you know, the journeys that you're trying to design for a customer, almost like a product manager type role. And then you're part like system integrator, like how do I connect into the right systems to get things done for my customer?
Georgie Healy: Mm-hmm.
Jason Maynard: And that's just a very new role within sort of a service organization. I often equate it to like what happened in like the 2000s with technical marketers, right?
Georgie Healy: Yeah.
Jason Maynard: You know, marketing used to be this highly sort of relationship-based, you know, you owned a channel, you knew all your provider, your channel providers. And then, you know, when performance marketing and Google became sort of this dominant sort of marketing channel, you had this new role of like a technical marketer, you know, someone that could look at the data could connect the right systems, could sort of, you know, build that funnel experience, build that sort of experience for how someone understood your brand and your use case and then how you attracted them to your site, how you sort of brought them through that, through that experience. And I think that very much that same role is growing in CX organizations.
Georgie Healy: So interesting you say that because I don't know if you've noticed this as well, but marketers to me are the most data-driven people I've ever met.
Jason Maynard: Yeah.
Georgie Healy: It's incredible. And, and to your point, it was only last week because speaking to the CEO of Vercel, and she was talking about the go-to-market engineer. So engineers are kind of getting better at that. And then it's like the technical people are understanding that side. And to your point, like it's kind of happening across the bridge as well.
Jason Maynard: Totally.
Georgie Healy: Different names.
Jason Maynard: Different names. But I, you know, and I think it's this sort of like, like a new job category, right? If you can really build those skills and go into companies, you know, in these operations roles and help them, to really automate a lot of those sort of like back office tasks, to really think about what the experience needs to be for your employees, for your customers. You know, it's just such a, it's going to be such an in-demand skillset right now. I think we're seeing a lot of people sort of gravitate and, you know, take on those roles, build those skillsets. And it's just a very fun time because I think we've always had a lot of customers in our base that, you know, they knew the problems that needed to be solved in the organization, and they just didn't necessarily have the tools at their disposal to go do it. You know, like service has just always been an interesting sort of stepchild in an organization, you know, like from getting IT support, you know, your IT team prioritizes marketing and sales and finance. And if there's enough left over, you know, we're going to go sort of think about our service tech stack. And, you know, in the world that Zendesk grew up in, you know, a lot of our customers self-implemented, you know, was wholly owned within a line of business. And, you know, thinking about like how you really enable the operators that are close to those problems to actually just go and not be beholden to an engineering team or an IT team to go solve a problem, to just go out and solve that problem for customers or employees. So I think that's really—
Georgie Healy: I still love that too. I don't know if as a consultant you saw this, 'cause I saw this. You'd go in, you'd see a client, and you'd ask them all the pain points and try and solve it and get all this data. And you'd speak to the team members, certain team members, and they'd be like, I've told the company a million times this issue. No one's ever listened to me before. I wonder if this is a new age where they will feel less disenfranchised because they can actually—
Jason Maynard: Totally. We're in like a, it's just a time where I think everyone feels more empowered to just go solve problems. I think that's exciting, especially when it's, you know, you can just be the one to sort of like see that problem and solve it for your customers, for your employees. I think that's a very sort of like powerful moment that we're in in technology.
Georgie Healy: So powerful.
Jason Maynard: I wanted to just talk about the other side. So that's sort of like that sort of like operations role, you know, which is sort of a new role that we're seeing grow in CX. And the other really interesting thing that we're seeing, you know, that's sort of like counterintuitive and it's sort of like a statistic that we're digging into is Indeed is a job site in the United States. They just published their their job numbers by sector for 2025. A really interesting thing was product development was flat. Some retail jobs went down 6 or 7% year over year. Customer service jobs in the States, in terms of job postings, went up 10% year over year. Just very, very counterintuitive. But I've been spending a lot of time with our customers in the States, obviously, and One of the things that I think is a common pattern is there's a lot of teams that are bringing their service teams back in-house. I think a lot of service teams use what we call BPOs, business process outsourcers. And so you can think about that as how do I take some of my tier 1 support and actually use a partner to go and serve that use case for our customers? They'll run our contact center or our service for tier 1 support. And I'm actually seeing a lot of our customers in the States actually go, we're actually driving a lot of efficiency into our service organization. We're actually able to automate a lot of the transactional type of tasks. And then what's left are these more complex types of issues where you need judgment, you need to solve a problem for a customer. And they're going like, okay, that's actually something that's a core competency. It's a differentiator for us to go and invest in that. And so we're actually going to like hire, you know, we're going to hire a lot of the people from that BPO. We're going to invest in training. We're going to invest in sort of like upleveling some of those skills. And I think it's just, it's sort of a very interesting counter-narrative to like, you know, a lot of the sort of like AI doomsday— Customer service is dead. Customer service is dead, you know, and it's sort of a counter-trend to that. It'll be really interesting to see where that goes.
Georgie Healy: Fascinating. I have not heard this before, but it kind of makes sense from— I don't know if you're noticing this as well, like personality is a moat, brand is a moat.
Jason Maynard: Totally.
Georgie Healy: Humans are a moat in a way because the technology is less and less, even in industry verticals, less and less of a moat, right?
Jason Maynard: Right. Yeah, if you're automating a lot of those transactional pieces and that, you know, these sort of like complex judgment-oriented problem-solving tasks are left, those people on the front lines are the face of your brand. You know, it's like they're the people that you're going to interact with at a company more than anybody else. And so, really investing in those people, I think is actually like a very positive trend for the industry. You know, thinking about how you make that a real career, you know, and the, you know, I talked about hospitality earlier, but it's like—
Georgie Healy: I was just thinking of that.
Jason Maynard: Yeah. You know, it's like, you know, how do you get to a place where, you know, when someone's reaching out to your company, to your brand, you know, it's actually a very lovely experience. You know, you know that you're going to get a lot of expertise, you're tapping into their knowledge, you know, they're going to be, you know, if it— you do have to get to the place where you pick up the phone, you know, you're actually getting someone who's going to help you out on the other end.
Georgie Healy: Imagine going to the most beautiful restaurant, but the, you know, the waitstaff treats you badly. It's gonna put a bad taste in your mouth, literally, right? Beautifully said. Okay, so by the time I'm reaching out to customer service, Jason, I'm already mad. I'm so mad.
Jason Maynard: Yep.
Georgie Healy: Like, how do you prevent something called digital snap. Maybe explain what digital snap is and then what do you recommend or what do you tell your customers and what do you try and encourage?
Jason Maynard: Yeah, I mean, I think digital snap is like, it's in one of those scenarios where you've fallen off the happy path, right? You know, I think you could get, there's certain times where you're just trying to get something done and you can't get it done, right? So I was just talking to a customer earlier this morning, and they were just talking about how, you know, it's a hotel brand and they were talking about how when their customers need to get an invoice, you know, they have to actually pick up the phone and reach out to their contact center and go, you actually need a copy of my invoice for my hotel stay. Like, it drives a huge amount of our contact volume. And, you know, that's a place where, you know, it's not great for the customer because, you know, their contact center is open 9 to 5. Monday through Friday, if you need an invoice that is out of those times, you can't get it. You have to call back at the right time. That's just a frustrating experience that falls into how can we introduce automation or the capabilities into the customer experience that allow them to do that? So I think that could be one way you get to frustration is just like, I just need to get this done. And then there's times where things just go wrong, right? And for those failure modes, automation is probably the worst possible thing you can put between your brand and a frustrated customer, right? Like if you want someone to take accountability and own an issue, that's actually where you want to design into the experience a way for those people to bypass a lot of the automation, to be able to get in touch with someone who can be empathetic who could go, yo, so sorry. The same way if you got the wrong plate in a restaurant, they would go, oh, I'm really sorry. We're going to fix that right away. It's on me. Let me bring you something while you wait. But I'm going to own this and I'm going to fix this for you. And I think that that's probably one of those places where sort of a nuanced point with automation in customer service where an agent can't be accountable.— it's not an accountable body that someone can go, I've had an issue and I just want to talk to a human and I just want to be able to know that you're going to fix this for me. And as a brand, it's sort of the best moment to sort of turn something around with them, to be empathetic, to connect with them and to solve it as quickly as possible. And so I think that's a lot about the sort of design of the experience. How you solve for digital snap is using the signals that we can get in those early interactions around sentiment. What is the sentiment when someone's speaking on the phone? What sort of cues are you getting as someone's interacting with an agent? And then using that as a way to escalate to a human when you know there's that sort of like, when there's that need and that sort of like need for accountability between a customer and a brand.
Georgie Healy: I was dying to ask you about that actually. What is more important when reading that data? Is it the sentiment or is it the, the hard requests? Like, I think I had an example here of like a tactical thing, like I need an invoice, and versus like clue words that they're really frustrated. Like, which, which, which do the agents still need a bit of practice in, if anything?
Jason Maynard: Yeah, I mean, I think they're both important is sort of the short answer. So there's, there's ways that we can understand sort of duress and stress and sentiment through those interactions and use that as a cueing mechanism to escalate to to a human, you know, through a conversation and the signals that we're seeing through that. And then there's, there's places where, you know, as you're doing the service design, it's really critical that, you know, you're, you're bypassing automation in the right places. When someone's contacting us about trust and safety on our platform, you know, that's not a place for automation.
Georgie Healy: Mm-hmm.
Jason Maynard: That's where we should bypass. We should connect to our trust and safety team. We have, you know, agents that are highly skilled in handling those situations. And, you know, so it's both, it's both sort of like what's the topic that someone's contacting you about? And then, you know, how frustrated is that customer, you know, in that part of the experience?
Georgie Healy: And so see that being so critical. I have a ridiculous question, but I don't know who else to ask.
Jason Maynard: Love those.
Georgie Healy: Yeah, fantastic. I wanted to put a prediction in my newsletter this week, which is like, I have so many friends that are using the apps for dating and they get frustrated by those. Why not let your AI agent date with another AI agent of someone else's and then escalate to an actual in-person date or not? Is this the craziest idea you've ever heard or—
Jason Maynard: Not the craziest idea. I haven't thought about the dating use case as much, but we have our own predictions about like how personal assistants are going to interact with service, right? So, you know, we've all followed like OpenClaw, you know, this sort of like, you know, proliferation of how, how you can use some of these agents in your own personal life and sort of connect with—
Georgie Healy: And those are, you know, they can dive into your WhatsApp and respond to your spouse or your friend or whatever.
Jason Maynard: Yep.
Georgie Healy: Cool.
Jason Maynard: And then we have, you know, Maltbook, which was sort of agents talking to agents in a social media setting. You know, there's a lot of kinks to be worked out, a lot of sort of security to be put in place before I think those are sort of ready for broader primetime. But you got this glimpse into you know, what if we all had a personal agent that was much more connected into what I do in my day-to-day that I can help me, can help me go do tasks. And you, you just think about that in a service setting, right? Is like, you know, if customers are armed with all the same amazing technology that a business is in terms of, you know, automating and, um, getting work done on, on your behalf, you, you get into this scenario where I think there's going to be a place where there's a lot more agents, personal assistants in terms of agents talking to brands' agents and sort of like how they're interacting. Right. And so getting to a place where, you know, instead of me having to be the one that goes through and fills out my information and, you know, gets a return slip and gets an email to me and then, you know, sort of like goes through that whole process, like how do I delegate that to an agent as a, as a customer? Yeah.
Georgie Healy: And I always want like reference numbers and then I'm like, fuck, that is—
Jason Maynard: and like, yes, that would be so good. Connected into your email, you know, it can sort of like parse all that data and sort of like pull what's relevant. And I think we're going to get to a place where there's just much more sort of like automation connected to, you know, personal automation connected to brand automation and, you know, how those things are going to work together. And I think that's actually sort of like a great thing. Like I always describe something called— I call it the dolphin problem. In customer service. So, you know, every time you have someone reach out to your, your brand, you know, if that's a customer having an issue, I always call it a dolphin problem because when you see a dolphin on the surface, you know, of the water, you know that there's like 20 dolphins in the pod below that, that you don't see.
Georgie Healy: Yeah.
Jason Maynard: Right. I think customer service is very similar to that where if someone reaches out to your brand about an issue that they're having, there's actually probably 20 other customers that have had that Exactly.
Georgie Healy: In, in a vacuum, right?
Jason Maynard: Right. It's not a vacuum. And how many times have you just had an issue and you're like, I'm just not going to pick up the phone.
Georgie Healy: Can't be bothered.
Jason Maynard: And I'm not going to, I'm not going to start chat. I just don't have time right now. And I think service is very much like that in that, you know, that one customer is doing you an amazing service by actually reaching out and going, this is the problem I'm having. Because behind them, there's a whole bunch of other customers that had that same problem that are disenfranchised that you can go then solve for. And the thing that's really exciting to me about the personal automation side is like, as friction to sort of like interact with a brand decreases, I think we're going to see actually a lot more interaction and a lot more sort of feedback to brands. So you could imagine as personal automation comes up, like actually the number, number of service interactions being handled by a brand, like doubling, tripling, quadrupling over the next few years because those friction points are really being reduced and it doesn't have that same time commitment to actually sort of like give feedback or deal with that sort of return that, you know, just has been sitting in your closet forever or whatever it is. Right.
Georgie Healy: And I think that's— Like, I would love to ask you, this is going to come on your radar soon because AI is moving so quickly. I would love to ask you, what are you going to do when my personal agent speaks to the customer service agent and all it wants is a discount? I want to 20% discount and it's harassing your agent. That's like, why would I give a discount? Is this a problem you think you're going to have to navigate in future years?
Jason Maynard: I think it's going to look a little bit different than that. You know, I don't think it's going to be as adversarial as that. You know, it's sort of like thinking about that as like a zero-sum game where, you know, there's like sort of like one agent battling another for a goal. You know, I think about it more as, you know, as you're designing these experiences, like as a brand designs experiences, it's going to— you're not going to have it adversarially challenge a customer or another agent in terms of trying to accomplish some goal. You know, you're going to have policies and procedures that you've defined that it's sort of got guardrails to handle. And I think as something falls out of that, I think that's like a perfect time where you'd actually want to engage and escalate and go, you know, is this someone— is this someone who needs this for a certain reason? Can I actually jump in here? And that's, that's like where humans and judgment will come into play.
Georgie Healy: I've been watching too much Celebrity Traders, so I'm always—
Jason Maynard: No, I love it.
Georgie Healy: Gotta be careful. I love it. Enemies everywhere. Okay, we've got a spicy rapid fire, 3 quick scenarios. Are you ready to go?
Jason Maynard: Ready to go.
Georgie Healy: Okay, tell me with each of these, are they better solved via an agent or by a real life human?
Jason Maynard: Yeah.
Georgie Healy: A customer's wedding dress hasn't arrived 2 days before the wedding. Agent or human?
Jason Maynard: Easy, human. You want someone that's account— high stress situation, really important event. You want someone that you're going to talk to that you know is going to be accountable and deliver on what you need. Just no way that talking to an agent is going to give you that sort of warm fuzzy feeling that you're being taken care of.
Georgie Healy: Set my laptop on fire, I think an agent came up. A user wants to cancel their subscription based on a horrible experience, like they've lost their job and can't afford the subscription.
Jason Maynard: So I don't think you ever want to put barriers between a customer and canceling your service. I actually think you have legislation coming in Australia that's going to force the hand on this one. So, you know, I think the days of you can sign up self-service for our product or our subscription, but you have to get on the phone, wait, call us in order to cancel it is an extremely frustrating experience.
Georgie Healy: That's one word for it. It does feel wrong, right?
Jason Maynard: It feels very wrong, right? It's like if I could do it self-service when I was signing up, I think it should be the same level of friction if I need to end this.
Georgie Healy: Last one. A teenager is trying to jailbreak the chatbot for fun. Basically all the things I kept saying I would do.
Jason Maynard: Yeah. I think a place where the capabilities of the agent is really important. I think we've put a lot of work into how you build the guardrails around what the experience is. Perfect place for automation to step in. You could think of it like spam. How do you stay on topic? How do you answer a question? But how do you not let the agent do anything that could be off-brand, something that could be snapped, put on social media? How do you really have those controls to handle those adversarial situations, you know, with grace and in the right sort of way. And yeah, not let it go off topic, but, you know, not something that you also want to sort of like take up your team's time with.
Georgie Healy: They should be having the important conversations, right?
Jason Maynard: Exactly. Yeah.
Georgie Healy: Yeah. This has been such a pleasure, such a joy. Thank you, Jason. Before I let you go, where can people find out more about Zendesk, more about you? How do they get in contact with the company? You're in Sydney, how often?
Jason Maynard: Down here a few times a year. If you want to get in touch with us, reach out to us, zendesk.com is our website. You can reach out to me on LinkedIn if you're interested in what we're doing. If you're in the area, if you're in Sydney or Melbourne, would love to chat with what you're doing, what problems we can help you with. LinkedIn's probably the best way. I won't tell anyone this, but I'm, LinkedIn/CaptainCheeseburger. Well, it's probably a story for another day, but yeah, you can reach out to me there, CaptainCheeseburger@zendesk.com.
Georgie Healy: I did not even notice, and I added you.
Jason Maynard: You did add me. And so it's a— well, we'll have to meet up again. I can give you the deep dive behind the story. But yeah, that's the best way to reach me.
Georgie Healy: Thank you so much for being on the show.
Jason Maynard: Oh, thank you so much. It was great. Really fun. So thank you.
Georgie Healy: Thank you for listening to In the Blink of AI. You can check out the show notes for anything discussed in this week's episode, and we will be back next week. This podcast was produced by Day One with music by Dan Hansen and visual artwork by Sophie Tyrell. If you loved the episode, please tell your mates, and I love AI news. Please share your thoughts questions and suggestions to georginarosehealey@gmail.com.
Jason Maynard: [FOREIGN LANGUAGE]
