Zendesk CEO Tom Eggemeier On Building The Autonomous Service Workforce

[00:00:00] - [Speaker 0]
So a special thank you to Denodo for supporting the Tech Talks Network and helping us keep these conversations going because moving beyond AI pilots all starts with connecting your models to trusted enterprise data. So if you're ready to move beyond AI pilots, Denodo can help you connect your AI models to trusted enterprise data in real time. So you can scale faster and reduce risk. So if you're interested in turning AI into business value, simply visit denodo.com. What does customer service look like when AI moves beyond answering questions and starts orchestrating entire workflows?

[00:00:46] - [Speaker 0]
Well, recorded at Zendesk Relate, my conversation today goes far beyond the usual AI hype cycle. Because, yes, while many companies are still experimenting with copilots and chatbots, Zendesk is a company making a much bigger bet on what it calls an autonomous service workforce, where AI agents, humans, workflows, governance, and enterprise systems all work seamlessly together as part of a connected operational platform. I'm very privileged today because here at the event, I've managed to secure a little time with the CEO of Zendesk. And together, we will unpack why he believes the future of enterprise AI will be measured by outcomes rather than activity, why CIOs are suddenly becoming cool again, and how businesses are struggling to keep up with the pace of organ organizational change that is required to make AgenTik AI actually work. And he'll also share how Zendesk has transformed its own service organization with AI and also talk about why the company made the bold move towards outcome based pricing tied to verified resolutions and why governance orchestration, permissions, and trust, how all these things will ultimately become more important than the AI model themselves.

[00:02:11] - [Speaker 0]
So, yeah, we're gonna look at the bigger picture today and try and understand what happens when most customer interactions become AI agent to AI agent conversations, but also how companies can avoid AI sprawl and what role humans will still play in this world increasingly shaped by automation and intelligent systems. Yeah. There's a lot to unpack here, So buckle up and hold on tight because I'm gonna beam your ears all the way to Denver, Colorado, where you can sit down and be a fly on the wall in this conversation. But enough for me. Let me introduce you to him now.

[00:02:48] - [Speaker 0]
So thank you for sitting down with me today at Zendesk Relay in Denver, where the theme is experience the future of agentic service. Now as the CEO speaking to customers around the world, what does that mean to them? What problems are we solving here?

[00:03:02] - [Speaker 1]
Well, I think we're just really in a unique time where businesses can fundamentally transform how they're interacting with their customers, whether those are consumers, businesses, or in quite frankly, their employees, and we have an opportunity to work with our customers and potential customers to help them on that journey, where it is very eminently reasonable to get to 80% of your interactions are automated 24 by seven, twenty four hours a day, seven days a week, 100 different languages at a higher customer satisfaction than you previously were serving your customers and that just gives you a lot of freed up headcount to go work on what I call service dividend, things that you haven't been able to work on in the past that you've always wanted to work on to increase the loyalty with your customers.

[00:03:48] - [Speaker 0]
And I do attend a lot of tech conferences, and I have to say there's a big focus on all of them on AgenTik AI, but I've also noticed the appetite is changing from shiny demos to outcomes, measurable impact, especially with audiences there. And one of the things that stands out here is you've emphasized outcomes over automation volume, and even introduced outcome based pricing tied to verified solutions, which feels like a big bold move, and a major shift for the enterprise software industry. So why do you believe the industry needs to move away from just measuring activity and towards measuring actual resolution quality?

[00:04:24] - [Speaker 1]
Yeah. So, you know, I think I think business leaders don't wanna buy seats ultimately, they don't wanna buy even consumption because you're just consuming all business leaders are driving an outcome with the businesses they serve, the consumers they serve and the employees and so we think the world of Agenetic AI is gonna lead to outcomes and so we were the first or one of the first people to go with automated resolutions. So we only get paid if we verify verifiably resolved one of our customers issues, customers' issues quite frankly, and we just think it aligns us with our customers better. I can tell you right now, I wake up and the first stat I look up every morning is, how are we doing on automation resolution rate for our customers and if we go down or go up, why and how can we drive

[00:05:10] - [Speaker 0]
that higher? When you look back to when that was first implemented, when that was suggested, there any pushback in the room? Imagine that

[00:05:17] - [Speaker 1]
So would have we were a little worried, because you don't know how to you don't know how to you don't know how to measure it completely, you don't know what how the impact is gonna be on your P and L. This is one of the nice things about being private. We went to our board and said we're gonna move over, we think the future is outcome based pricing and outcome for our services automated resolution. We think it's gonna be great for our customers, which will ultimately be great for us but there could be some, you know, rocks in the way, some boulders in the way and like we're just gonna deal with that. We think over, know, the next three to five years, it's great for our customers and our board was really, really happy about it, our customers have been happy and it's changed how we do some things like, we're putting along with our partners hundreds of people sitting by our customers right now to try to go get their resolution rates up because we only get paid if they're driving resolutions.

[00:06:04] - [Speaker 1]
So it's changed our behavior as well, think in a good way.

[00:06:07] - [Speaker 0]
One of the things that stood out today in one of the press sessions is that many enterprises are still underestimating the amount of workflow design, governance, permissions, and architectural claims changes that a need to make AgenTik AI actually work. Are you finding that the tech is moving faster than maybe the organizational readiness right now?

[00:06:24] - [Speaker 1]
I I do, and I I just think it's just general, you know, human nature, and the tech is solving some of that. Like one of the, you know, work processes is in the past, you create a policy or procedure or a knowledge article about something. Now AI can do that for you, but you need the judgment to go approve it. So I think the big hold up to adoption right now is people and processes that they got to be fundamentally rewired, which is not trivial and the second is you just need to go change the skills and capabilities of some of the people that you have and so this is the blockage right now and that's why we're investing along with our partners to go sit side by side and try to share best practices with our customers, so they can adopt AI quickly.

[00:07:06] - [Speaker 0]
And for people that are listening and unable to attend, what is Zendesk strategic direction and growth priorities right now? What are they missing from this week?

[00:07:14] - [Speaker 1]
So number one is to serve you, our customers, our potential customers the best. It's like I am just more and more convinced that the more value we drive for you, our customers and that better loyalty you create with the consumers or businesses that you serve, the better is it for Zendesk and so that's our big strategic priority is help you drive automation rates up while you drive customer satisfaction up. Hopefully that frees up resources that you can spend getting our root causes of why you have some service issues and we think the world is gonna be more personalized, more agentic, more automated in the future, and we hope we are the platform that our customers, potential customers use for that.

[00:07:51] - [Speaker 0]
And if we zoom out, I do think there's almost a fascinating tension running through enterprise AI right now. Everybody wants autonomy and efficiency, but they also want governance, oversight, and trust, which is a big balancing act. But as Zen Zendesk scales autonomous AI agents across voice, messaging, and external platforms from ChatGPT to Gemini, etcetera, how do you balance that innovation speed with accountability that the enterprise is also demanding? It feels almost an impossible task.

[00:08:20] - [Speaker 1]
Don't know if it's a possible I think this is one of our strategic advantages. We've spent eighteen or nineteen years on security. We have hundreds of security professionals, for instance, thinking about this every day. We've done a little things like our marketplace, we've required everyone to do OAuth to make sure that we are really really secure and not just our applications, the applications that are connecting with us and so security governance, what kind of permissions you're giving to every employee on what data they can see, all these things are becoming more and more important, that's why I think this year is gonna be the rebirth of the CIO. The CIO is cool again.

[00:08:55] - [Speaker 1]
SaaS applications, so it went from when you have premise software CIO decision, SaaS went to business leaders. I think the CIO is gonna have a more and more important role because of privacy, security, data, governance, all these things.

[00:09:08] - [Speaker 0]
CIOs are cool again, I love it and at the event this year, we're hearing inspiring stories, a big focus on Zendesk and the customers that you're serving there, but Zendesk is also on a journey itself. So if we take a peek behind the curtain, tell me more about your own AI driven service transformation at Zendesk from classic SaaS to outcome based AI companies. Hell of a journey.

[00:09:32] - [Speaker 1]
It's a big journey, and I'm gonna give you these numbers just because I think it's a reflection of how quickly we are moving. We'll do about 2,100,000,000 in AI and ARR this year and we'll do about $450,000,000 of AI ARR in our AI agents co pilot and a little bit of AI quality assurance this year, just like zero two years ago or two and a half years ago and so we've retooled our service operations, just to give you one example, level one and level two support, now have 60% automated resolutions and that's a business to business use case, which is usually a little lower than business to consumer, about a 20 increase in customer service for those interactions and we've been able to take those, you know, those roles that used to do that and put them in with our customers to go drive automation and so we focus first externally, but we've got a big big push right now on not just Zen for Zen, that's what we call our internal, you know, service use case, but all AI, know, efficiency and customer satisfaction use cases.

[00:10:34] - [Speaker 0]
And there's the ROI right there, and looking at Outlay as well, the Forethought acquisition seems strategically important because it's introducing self improving AI agents into the Zendesk ecosystem. What was it that you saw in Forethought that convinced you that the future of service would maybe require a different kind of AI architecture, than simply layering AI into existing workflows?

[00:10:56] - [Speaker 1]
Yes. Forethought is just really really, when we looked at the customer satisfaction they were driving and things like G2, it's an independent platform to for peak customer reviews and they were number one in AI agents. We thought it was because some of these things like self improving AI agents that were really consistent, some of their thought leadership on how they took actions, you know, from an agentic perspective. We've done seven acquisitions in the past three years, with a deliberate strategy along with a lot of organic innovation. We just think that overall customer, the market is moving some clock quickly, you're gonna make a lot of right decisions as a company but you're gonna miss some things and we miss some things, we owe it to our customers to go acquire companies to make sure that the word, the, you know, leading part of innovation.

[00:11:46] - [Speaker 0]
And talking of your customers there, what are your customers and indeed your attendees? What is everybody talking about right now? Are there any trends that people are talking about regardless of industry that are they all looking at the same things, asking the same questions or is it some are they all different?

[00:12:00] - [Speaker 1]
I I think they're generally they're generally looking how deep I can go in my vertical with the, you know, use cases, whether that's ecommerce, a tech company, a, you know, traditional retailer, manufacturing company, like what can I go do to a genify, I don't even know what the right word is, my AI workflows there? I do think we were going back to the point that we made earlier Neil, that you know, service, the security governance, data privacy, all that is coming up more and more to the forefront and then finally, what does this mean for my employees? Like reskilling, how do I get people that to get AI knowledge, particularly that aren't like AI or digital natives, questions like that, but it's all about the transition. I think a couple years ago, people were like, is AI real or not? I think everyone says AI is real.

[00:12:44] - [Speaker 1]
It's kinda like, how do I move quicker, quickly while benefiting my employees and customers?

[00:12:49] - [Speaker 0]
And there's so many big announcements coming out of this week, and I think it's easy to lose track, because any event, there's always five, six, seven, eight, nine, ten different press releases, all getting bombarded out there. What excites you about the announcements this week?

[00:13:03] - [Speaker 1]
So, there's just probably the breadth and depth. We are gonna launch about twice as many new products and twice as many features as we did a year ago and I give a lot of credit for, know, AI to go help us in that that that journey. So what excites me, everything from our Zendesk contact center, that way you don't have to have Zendesk Suite, you can just go to the contact center first, some cool stuff on from employee service that we're really, you know, revolutionizing how employees get served from the companies they work for. I do think some are co pilot announcements, where I'm big on AI agents but co pilot admin, where I can see all the AI agent interactions, the human interactions, what's going on and I can you know change with procedures and workflows, like get recommended how you should change your systems to go do that. I just think it's just absolutely amazing what you can do with a couple clicks of you know, your cursor now.

[00:13:54] - [Speaker 0]
What kind of feedback have you had from your customers so far? Know it's early days,

[00:13:58] - [Speaker 1]
feedback, but it's kind of what we talked about at the beginning, they all want us to go faster, deliver more value and continue to be in you know a platform. I think they've seen our transformation with these seven acquisitions and we're spending over $400,000,000 a year in research and development. We've upped our percentage of research and development spend compared to you know revenue. They see how quickly we're going on but they say they want more faster and then second, how can they go take advantage of all the features? We've got to do a better job as a company on the new products to get them adopted because it's hard to keep up for our customers with all the innovation.

[00:14:35] - [Speaker 1]
And so I would not have had that problem two or three years ago, and now that's one of the things our customer says. I love all the innovation. How do I keep up and how do I adopt it?

[00:14:43] - [Speaker 0]
A quick thank you to NordLayer for supporting the podcast and helping me make these daily conversations possible. And if you are listening and you're responsible for security or IT, you will know the reality. The reality that most of your risk now sits inside SaaS apps and browser activity. That gap is exactly what NordLayer is addressing with its new business browser. So instead of bolting security on from the outside, it builds it directly into the browser itself.

[00:15:17] - [Speaker 0]
This means you can control access, monitor activity, enforce policies, and reduce shadow IT all from one single place. And most importantly, it does it without adding deployment headaches or complex onboarding. You get things like browser based data loss prevention, SaaS access control, and zero trust browsing, but delivered in a way that your team can actually use. So if you've been trying to simplify your stack while improving visibility, please check it out at nordlayer.com/browser. And, of course, with all the noise that surrounds AI and Agenty, one of the things I keep hearing is that many enterprises are becoming overwhelmed by fragmented AI tools, Copilot's workflows, so many disconnected systems.

[00:16:04] - [Speaker 0]
Is Zendesk maybe positioning itself as a customer service platform, or are you increasingly trying to become maybe an orchestration layer for how AI agents, humans, and enterprise systems all work together? Is that is that in my Definitely.

[00:16:17] - [Speaker 1]
We've really gone through all these channels that you mentioned from voice to email, to messaging, to you know, AI agents, we've now got AI agents, we got CoPilot, we got our core suite, we've done all this to quality assurance, workforce management, we've done all this to create platform. We've really focused on our orchestration level for service, anything that touches service and we've invested a lot in connectors, so that we can connect to different data sources within your within someone's company in an easy way because it is getting increasingly fragmented. We particularly our small medium sized customers, we have enterprise customers also small medium, they love buying one platform and we think that's the way of the future is you're gonna have a couple orchestration platforms, you're gonna have one deep probably, or two deep service platforms, and so we've transformed ourselves to kinda listen to our customer requests.

[00:17:09] - [Speaker 0]
And what would you say makes you stand out in the market like this, because the world of CX is such a crowded marketplace, what makes you guys stand out, do you think?

[00:17:16] - [Speaker 1]
I think we first have this whole overall platform, where there's very few companies that can do everything from voice to messaging, to have a 100% of the interactions measured by a quality assurance platform, and have that all integrated with AI agents, a great knowledge base and so the whole platform, I think we have the depth and breadth that others do not have. Second, I do think are really focus on openness with those 800,000,000,000 AI or API calls a year headed towards a trillion this year and MCP server that you can bring your own AI agents, you can have a Zendesk AI agent or you can create your own on top of things like action builder and app builder, and some of the other things that we have is a really flexible platform, but that has a good orchestration layer.

[00:18:03] - [Speaker 0]
As an overcautious ex IT guy, the the idea of of launching hundreds, thousands of AgenTik AI agents out there just make me a bit uncomfortable. Do you do you think that that we will see any problems? If he's not with Zendesk, but in general, are we underestimating the scale of of Yeah. Without the proper guardrails?

[00:18:19] - [Speaker 1]
And that's why I said I think CIOs are cool again, because I do think if the business leaders just launch a bunch of AI agents without orchestration layer, without thinking of governance, without thinking of what permission each employee has to go get data because a lot of when you create your own AI agents about what data you're pulling in for the system and do you give all the data to an employee to create their own AI agent or something to do a task. So I'm not playing to my audience here, don't think Neil, but I really do, I've said it before, I do think the CIO becomes more important because you are trying to put all this together across your whole enterprise in a rational way. I give the example and I think I gave it this morning, we had like 17 different people you know, some of these tools to bring a you know, pitch decks, an AI pitch deck and it makes more sense to go buy something off the shelf than it does to create, know, 17 little applications. Even one of them, full time job, was to create a pitch deck AI agent, and so I think these are the decisions that CIOs are gonna be making, and business leaders are gonna be making, but I do worry about sprawl.

[00:19:20] - [Speaker 0]
And for that CIO that's listening thinking, man, I'm gonna be cool again. Anything they should be doing to prepare, do you think? I think

[00:19:26] - [Speaker 1]
they should be using their own tools as well because they should be using the tools. I'm probably of the same age vintage a lot of the CIOs are and you know, she or he has the opportunity to go really use them and I know that's one of the good things for myself. I'm not a software engineer but I pushed some code two weeks ago, that's actually out in production right now and I saw what our software engineers are going through. I also hit that service dividend, it was a it was a four year old request from 46 of our customers that we never got to. I was able to knock it off in an hour.

[00:20:02] - [Speaker 1]
Now our processes were pretty heavy to go get that approved in production code, which makes sense but this was just a tooltip hover thing that it should not have been such a laborious process to get it approved and so I've learned a lot from that. I did an application with my son, you know, the the tools like Codex, Cursor and Claude and so I would encourage CIOs go use them massively, personally and professionally, and even if you're gonna fail a lot, I think you're gonna have much more understanding of what these tools entail.

[00:20:33] - [Speaker 0]
I I think it's just staggering the pace of technological change that we're seeing right now. In the last three years alone, since the arrival of generative AI to now where we are with AgenTik, etcetera, and I suspect that questions that you are being asked now were not asked even six months ago, you know? So where where are we heading with all this? How do you think the next two or three years will pan out? I know it's almost impossible to make a prediction like that.

[00:20:57] - [Speaker 1]
Yeah, think you're gonna see some interesting things. One, think is it three years or four years? I don't know, but I think the majority of consumers, I'll use consumers or businesses that are interacting with the company are gonna do that through their own AI agents, so you're gonna have the majority of interactions being AI agent to AI agent is one of the trends I think we'll see over the next three to four years. I think we are gonna make more progress to the future of work from a technology point of view than people think but I think people are underestimating not just the work, the process reengineering and the skills and capabilities but the regulatory environment. For instance, there's a lot of predictions that 50% of attorney work for first and second, third year attorneys are gonna be gone in the next twelve to eighteen months.

[00:21:40] - [Speaker 1]
I don't think the legal or regulatory framework is gonna allow that, so even though the technology might be there and so I think we're gonna be grappling with a lot of legal and regulatory points that are smart. I also think we're gonna be regulating with data privacy even more, know, used to be okay with having a European Union footprint for AWS instances or others, maybe Germany with the German privacy laws, we want a little more specific. I'm now hearing from customers in Europe and elsewhere, I want my my data to stay in Belgium. I don't want to go beyond the Belgium borders and so I think we're gonna be a lot of these kind of security privacy data questions are gonna kinda regulatory questions are gonna rear themselves again the next couple years.

[00:22:23] - [Speaker 0]
Wow, and another huge topic, I know you're an AI optimist, and AI is not about replacing people, it's about training people, bringing people along for the ride. From your own personal journey at Zendesk, what do you do to continue that and and help people along the way there?

[00:22:38] - [Speaker 1]
I think we try first be transparent, and I tell everyone if someone, like, there's some people out there saying like, I know how this is gonna go evolve, maybe there's five ten people in the world who really have a, you know, know how it's gonna go but we try to tell people there's uncertainty. Second, try to be empathetic. I've got a 23 year old daughter and a 21 year old son and they're, you know, shortly entering the workforce and you know, they're excited but they're fearful about what's gonna happen in AI and I try to put, you know, I try to think about, you know, fellow employees that are in that same situation and talk about the skills and capabilities they need to get, that are gonna benefit them no matter what the direction of AI goes, and I think that's an important thing. That's why I'm big on trying to lead by example, by using it myself in a big way. I think everyone needs to do that.

[00:23:24] - [Speaker 0]
And for everybody listening here, you've been you've been what, 12 interviews today. You're on stage, press briefing, so talking to so many different people. When you do get that a little downtime, maybe on that long flight, or not so long flight home, but on that flight home, are you gonna be thinking about and reflecting when you when Zendesk Relate is over?

[00:23:43] - [Speaker 1]
Yeah. I'll probably learn a couple things from from customers, press, analysts, you know, not even a couple things, a lot of things. I think that's one of them the the business was somewhat static in the SaaS area for about ten years and now I've filled a couple notebooks through full of like golden nuggets from different people and so I'll probably take my thoughts and try to distill it down to the top five or 10 things I learned from our customers, the press, analysts, partners, over the past couple weeks, and then try to transition to spending a little time with the family.

[00:24:14] - [Speaker 0]
Absolutely love it. And as I said, I can only appreciate how busy you are today, so thank you for sparing a little time to sit down with me.

[00:24:19] - [Speaker 1]
Thanks for having me, Neil. I appreciate

[00:24:22] - [Speaker 0]
Wow. One of the many things that stood out to me in this conversation with Tom today was just how much the conversation around AI has matured in such a short space of time. Because here at the conference, it's no longer about shiny demos or experimental copilot sitting on the edge of existing systems. The discussion has actually shift towards orchestration, governance, operational redesign, and measurable outcomes, and perhaps most importantly, how organizations are preparing their people for what comes next, investing in people, training people, ensuring nobody gets left behind. And I think Tom painted a fascinating picture here of a future where AI agents increasingly interact with other AI agents and where automation rates climb dramatically and where customer satisfaction can improve at the same time.

[00:25:18] - [Speaker 0]
And maybe that's why the company made such a bold move towards outcome based pricing tied to verified resolutions. But we did also acknowledge today the very real challenges around this, around trust, permissions, privacy, workforce, transformation, and organizational readiness, all big topics. And, honestly, if we zoom out from everything we've talked about here, I think that balance between optimism and realism is for me why this conversation felt so valuable. So if you enjoyed this episode, I'd love to hear your thoughts. Are businesses truly prepared for the shift from AI assistance to autonomous operational systems?

[00:25:56] - [Speaker 0]
And as AI becomes more deeply embedded into enterprise workflows, what role do you see humans playing in the next chapter of work? As always, keep the conversation going with me on LinkedIn, x, Instagram, just at Neil C Hughes, and pop bytechtalksnetwork.com. Over 4,000 interviews there. But that's it for today. So please keep your comments coming in.

[00:26:19] - [Speaker 0]
I'll be back again tomorrow with another guest. But thank you for listening as always. Bye for now.