Latest / CXOTalk / Zendesk CEO on AI for Customer Experience | CXOTalk #866
Transcript
- 0:00Welcome to CXO talk #886 I'm Michael Krigsman and we are
- 0:07cutting through the hype around artificial intelligence and
- 0:10customer experience with Tom Eggmeyer, CEO of Zendesk.
- 0:16Tom will share honest, real world lessons from Zendesk's AI
- 0:21journey, what's worked, what hasn't, and how AI is reshaping
- 0:26customer expectations. Tom, welcome to CXO Talk.
- 0:32I'm delighted to chat with you. Great to be here Michael, and
- 0:35I'm delighted to be #886. Everybody has received messages
- 0:41from Zendesk. We submit a support request and
- 0:46we get something back from Zendesk.
- 0:48So just briefly give us some background.
- 0:51Tell us about Zendesk. Zendesk is AAI first customer
- 0:55support software company. We're all about helping
- 0:58companies resolve their customers, their employees, or
- 1:01their fellow business issues. What's going on with AI and
- 1:07customer experience today? When we think about AI from a
- 1:11customer support perspective, we think there's just huge, huge
- 1:15gains for companies, employees and consumers.
- 1:18From an AI perspective, the it's not hype, it's actually real.
- 1:22So when you say that AI is disrupting, can you elaborate on
- 1:28that? AI is actually solving customers
- 1:31problems. The hype, I'm old enough to know
- 1:3420 years ago we talked about machine learning or predictive
- 1:37analytics. We're going to be able to solve
- 1:39customers problems, quite frankly, without human
- 1:42intervention. And it's been a little bit of a
- 1:45hype cycle for the past 20 years, but we're actually seeing
- 1:48that happen right now. We have one product called AI
- 1:53Agents Advanced where you put a bot into one of our customer
- 1:57environments and with three clicks of a mouse, taking some
- 2:01knowledge from the company, we can solve almost instantaneously
- 2:05more than 30% of customers interactions.
- 2:07And what's interesting is the customers are more satisfied, it
- 2:12costs less, it's more accurate and you can do it with three
- 2:15clicks. And that's kind of the promise
- 2:17of AI is always on fantastic accuracy, lower cost for
- 2:22companies so they can go put their finite resources into
- 2:26maybe solving the the root causes that are causing them
- 2:29having problems with their customers.
- 2:30There's so much discussion around AI and you've just
- 2:35presented. Can I call it a a kind of
- 2:38idealized version or is it totally realistic?
- 2:43Or where's the gap? The gap is a couple things.
- 2:46Number one, you know, I see some people in the marketplace, you
- 2:50know, guaranteeing 65 to 85% of interactions can be resolved
- 2:55with AI within a matter of days. And I think that's hyping a
- 3:00little bit. There are some of our customers
- 3:02that are getting 85 or 90% resolutions, but there's some
- 3:05customers with more complex use cases that are getting 20 or
- 3:0830%. So I think part of the hype is
- 3:10it kind of depends on the customer type, what kind of
- 3:14interactions having with their employees or with their
- 3:16customers on what kind of resolution rate you can get.
- 3:19The second thing there's a little bit of hype is right now
- 3:24you need to have something called a knowledge base or a
- 3:26bunch of information or other knowledge, you know, structured
- 3:30data so that you can get resolution rates quickly.
- 3:33And that's not always the case. And so we got to be honest with
- 3:36customers and honest with consumers about that.
- 3:38And then third, there are still mistakes that AI makes just like
- 3:42humans, and we don't want to underestimate that.
- 3:45And we want to make sure that we get as close to 100% accuracy as
- 3:48possible. What's interesting is we've done
- 3:51some AB testing and we see less hallucinations from AI than we
- 3:57see mistakes from humans. But it's interesting that we use
- 4:00different nomenclature. We call AI agent, you know,
- 4:02mistakes, hallucinations and human agent mistakes, mistakes.
- 4:06When things are set up in the right way, that the data is in
- 4:10the, is organized properly and so forth, then the AI does a
- 4:16great job, is essentially what you're saying.
- 4:18Exactly, I was with a customer about a month ago in Australia,
- 4:21for instance. And what they, what we found out
- 4:25working with them is they had five different policies, for
- 4:28instance, on from different parts of the organization on
- 4:31what constituted a return. It was an e-commerce company.
- 4:34And when you have data like that, of course the AI is not
- 4:38going to get it perfect because it's getting input at 5
- 4:40different, you know, policies that are slightly different from
- 4:43each other. And so really having good
- 4:44policies, having good knowledge, having good information is
- 4:49really, really key. Now we can help and you know,
- 4:52other players in the market can help create that knowledge.
- 4:55Like one of the things we do now is we see a bunch of the
- 4:58customer interactions and we see you have a gap right here.
- 5:01In the past, we would have said you have a gap, you should go
- 5:04solve it. Now we say we, you have a gap,
- 5:06here's a recommended. We write it for you, OK?
- 5:10The AI writes it for you. A recommended article or
- 5:13recommend a policy, Do you want to accept?
- 5:16Do you want to decline? Do you want to edit it?
- 5:17So the world is kind of flipping on itself right now.
- 5:21But like you said, data is really structured and
- 5:24unstructured. Data that's consistent, that's
- 5:28publishable, and that's readable is really important in this AI
- 5:31journey. You mentioned this phrase AI
- 5:33journey. I was really interested.
- 5:35At what point did you recognize that AI and customer support was
- 5:42so important that you had to? Would it be?
- 5:47Is it too strong a term to say, transform the company around
- 5:51this? Absolutely transform the company
- 5:53because the way we look at it as Zendesk, we can be the disruptor
- 5:57and AI customer support or we're going to be disrupted.
- 6:01And we were the disruptor 17 years ago with customer support
- 6:04helps the help desk, and we want to do that again.
- 6:07I remember when Open AI launched, I think it was October
- 6:102022, and there was an amazing amount of buzz about it.
- 6:15And just thinking about the implications of it in the summer
- 6:19and the fall of 2022 and I was over in our Lisbon, Portugal
- 6:23office. I remember this vividly in
- 6:24December 2022 with some of our lead, we had a small group of AI
- 6:30data scientists and engineers, about 10 of them, and I asked
- 6:33for our road map. I just started that month as CEO
- 6:37and we had a bunch of great ideas, but we didn't have a road
- 6:40map. And I think talking to industry
- 6:43experts in the summer and the fall of 2022, you know, I came
- 6:46to the conclusion that just like Internet, just like mobile, just
- 6:50like premise to cloud, there was going to be a major, major
- 6:53disruption. I think some people came to that
- 6:54maybe 6 or 12 months ahead of that and we did not have a road
- 6:58map. We did not have enough people on
- 6:59it. We did not, we're not thinking
- 7:01about it. And so we spent a couple days
- 7:03writing on a whiteboard. I wish I would have taken
- 7:05pictures of it in Lisbon, Portugal, about what kind of
- 7:09first product that we wanted to launch.
- 7:11And so that was kind of the aha moment for me.
- 7:14You know, I would have thought it would have been the Bay Area,
- 7:16but was actually in Lisbon, Portugal.
- 7:18You mentioned disrupting yourself.
- 7:20That's a hard thing to do. Most of the time businesses get
- 7:25disrupted because there's a competitor, there's something
- 7:27going on out there and they have no choice.
- 7:30But in this instance, you saw the tidal wave coming and so you
- 7:36took proactive steps and said, hey, we need to change.
- 7:39It seems like that's what you were just describing.
- 7:42I don't think we had an option, honestly, Michael, you know, if
- 7:44I'm really honest, because the customer support, like we have a
- 7:49stat, we think 80% of customer interactions with companies are
- 7:53going to be through AI within five years and a 100% are going
- 7:58to involve AI in some way within five years.
- 8:00Many humans will use AI. So we didn't have a choice, but
- 8:03we have a lot of great Zandeskians employees that saw
- 8:07this wave coming and saw that, hey, we were the disruptor 17
- 8:11years ago. Mikkel and the whole Zendesk
- 8:14team disrupted the help desk and customer service and we wanted
- 8:17to be, we wanted to do it again. So I would say we're fortuitous
- 8:23that this was coming, that we had some of the building blocks
- 8:26and pieces in the foundation. But also it's just an imperative
- 8:30that if we didn't do it. And I had seen in the premise to
- 8:32cloud contact center world, I work for a company called
- 8:35Genesis that was 100% Prem. It's now one of the leading
- 8:39cloud providers of contact center.
- 8:41They went, we went through that metamorphosis where the other
- 8:44two leaders of Contact Center Prem didn't go through it.
- 8:47And so it's easier to do as a private company.
- 8:50But there's definitely an imperative to do it because
- 8:52there's going to be winners and losers here.
- 8:54And I really feel responsible first for our customers to give
- 8:57them value, second to our employees and finally to our
- 9:00shareholders that we need to disrupt ourselves.
- 9:03I want to remind everybody that there is a tweet chat taking
- 9:07place right now on Twitter X and you can ask your questions using
- 9:12the hashtag CXO Talk. And if you're watching on
- 9:17LinkedIn, just pop your questions into the LinkedIn
- 9:20chat. And we have some questions that
- 9:22are coming up. But I'm I'm still really
- 9:26interested in this transition. So as as you were moving from
- 9:32the pre AI world into AI, did you get any resistance?
- 9:38Did you from from stakeholders, from your board, from employees,
- 9:42from your customers? Because it's a big change, and
- 9:45change is always hard I. Don't know if there's
- 9:47resistance. I'll be honest with our board,
- 9:49it's actually been how can you go bolder, How can you go
- 9:52bigger? Because they see the
- 9:53transformation. So if anything, I, we have a
- 9:55great board and I say they do strategic agitation and their
- 9:58agitation again is go faster, think bolder, think bigger on
- 10:02how you can really give value to your customers.
- 10:05I think there's different levels of maturity within our customer
- 10:07base. Some people are like, want to be
- 10:09cutting edge. Some people were over the last
- 10:12couple years saying, hey, is this going to be real?
- 10:14Is this really going to disrupt? Are we really going to be having
- 10:17AI agents copilot generative search for knowledge bases,
- 10:21really transform customer support?
- 10:23I think the vast majority of customers are saying right now,
- 10:27you know, after these two 2 1/2 years that AI is real, AI is
- 10:30disrupting customer support. So I think it's gone from a
- 10:33question mark 2 1/2 years ago with our customer to a
- 10:36consensus. The board has already been there
- 10:38and employees I think that ask real good questions is like, how
- 10:42are we going to be the disruptor and not just be disrupted?
- 10:44How are we going to do all the core things that we need to do
- 10:46on our core customer service Omni channel platform while
- 10:50adding AI? Am I going to be left behind if
- 10:53I'm not part of the AI revolution?
- 10:55And so we have, you know, coffee chats, we have Amas where we
- 11:01talked to the employee base. But if anything, I think the
- 11:04employee base is really, really excited about being the
- 11:07disruptor, being number one in customer service and really
- 11:10serving our customers and getting real, true, tangible
- 11:13value to our customers with AI. Now would be an excellent time.
- 11:16To subscribe to the CXO Talk newsletter, go to cxotalk.com.
- 11:20We'll keep you updated on upcoming shows.
- 11:22Tell us about the changes that you drove into the product.
- 11:28So how is customer support rooted in AI different from what
- 11:36came before? There's a couple of things where
- 11:39we are now and where we're going in the future.
- 11:40Where we're going right now is we think we can automate a lot
- 11:44of customer interactions. So for the customer, again,
- 11:47that's what it's all our customers.
- 11:49Customer it's all about. It's quicker, it's more
- 11:51accurate, and they're able to solve their problem in a really
- 11:55good way. For companies, they're spending
- 11:57less on it. And So what has happened right
- 12:00now in AI is a little more reactive.
- 12:02And I'll give you a couple of use cases.
- 12:03In the future, it's going to be more proactive.
- 12:05But in the past we have about 30,000, 1000 knowledge based
- 12:08customers. So think you go into a company's
- 12:12website and you say how you type in how do I return these pair of
- 12:15shoes? A knowledge base in the past
- 12:17would come up with 5 or 10 blue links and you'd say and you'd go
- 12:22pick one on how you return shoes.
- 12:24What happens now is we've launched something called
- 12:26generative search, and this is free part of our product for
- 12:29knowledge bases where like a Google Gemini search, you
- 12:32actually get the answer, you know on how to return the shoes.
- 12:36And it's getting even better with a gentic AI where you might
- 12:39be able to do that whole process, including going back to
- 12:42an ERP, for instance, all you know the actions all through an
- 12:47AI agent. So that is kind of reactive
- 12:49still. You then go to an AI agent,
- 12:51maybe from the degenerative search and an AI agent can act
- 12:54like a human agent and do all the things a human agent can do.
- 12:57Sometimes generative search and an AI agents cannot going to
- 13:00solve the problem. And so you go to a human agent,
- 13:03but that's all reactive right now.
- 13:05We're kind of flipping that right now in customer service
- 13:08where it's proactive, where we're taking all this
- 13:11information from the searches you make, Michael, the
- 13:13interactions that you have with an AI agent and human agent.
- 13:16And we're saying these are 7 root cause issues that are going
- 13:20on. We keep sending out black shoes
- 13:22when Michael won it, red shoes. And then we have customer
- 13:25interactions. What do we need to go do and
- 13:27like have an insight into the supply chain?
- 13:29Why do we have a problem there through AI?
- 13:32And so it's getting being flipped on its head right now,
- 13:35customer support from reactive to more proactive, from reactive
- 13:38to more root cause analysis and from reactive to actually
- 13:42actions of solving root cause problems that a company is
- 13:45having interacting with its employees or customers.
- 13:48And again, I want to come back to this, this comment you've
- 13:51made a couple of times on the accuracy.
- 13:54Can you talk about that? Because of course this is a key
- 13:58dimension of everything here. We've done a lot of AB testing
- 14:02and on let's call it the top 50 or 100 use cases.
- 14:05Usually an AI agent is as accurate or more accurate than a
- 14:09human agent. When you get to really complex,
- 14:13maybe a customer asks 3 different things in a query,
- 14:16they say in an e-mail or a chat or a phone interaction where
- 14:20there's voice AI. Now you say how do I return my
- 14:23shoes? I want a credit, I want to order
- 14:26a new pair of red shoes instead of my black shoes, and I want to
- 14:30go order a sweater. When you get 3 or 4 queries in a
- 14:34row, like in one message, sometimes AI does not nail that.
- 14:38I think a human does better still, but it's getting better.
- 14:41The other place where I think AI is not quite there is a really,
- 14:45really complex corner case that the, again, the data does not
- 14:49solve where a human needs to get involved.
- 14:52On the other hand, I think AI, it's interesting, some of the
- 14:54data that we've had some research recently, data is doing
- 14:58just as good or better. The data says on empathy, which
- 15:01is surprising. I would have always thought
- 15:03humans are going to be more empathetic, but there's some
- 15:06interesting research piece I'm forgetting where I read it
- 15:08recently where doctors, AI agent, doctor and AI or a human
- 15:13doctor. AI agent doctors are considered
- 15:17more empathetic and so we're starting to see more empathy,
- 15:21more caring from AI agents, which is something I never
- 15:24thought we would have had. We have come a long way from the
- 15:30chat bots, which were simply prescriptive and frankly for a
- 15:34lot of users, honestly a waste of time.
- 15:38Things of There's a very different universe now.
- 15:41Exactly. So those chat bots that you
- 15:43probably have in your mind that would drive me, you know, we're
- 15:47they'll be frustrating for me, frustrating for you.
- 15:50We're kind of decisions tree. You know, if then you go yes,
- 15:53and you kind of go down the decision tree and a lot of times
- 15:56again, the inquiries were more complex and you get stuck at
- 15:59part of the decision tree and then you get handed over a
- 16:01human. The human wouldn't agent
- 16:03wouldn't have the context of that chat bot.
- 16:05What's different now with the gentic AI is there are these AI
- 16:08agents are basically autonomous and can reason and that's the
- 16:12fundamental difference from the chat bot experience everyone
- 16:16had. And So what I I tell everyone, I
- 16:18know you've probably dealt with chat bots in the past, but if
- 16:20you drew you deal with a true agentic AI agent right now, I
- 16:24think you're going to be wowed by the experience.
- 16:25It's not perfect, just like human being AI.
- 16:28Human agents are not perfect, but that's the fundamental
- 16:31difference. You're going from a kind of a
- 16:32decision tree pre baked to something that's agentic.
- 16:36That's something that's reasoning.
- 16:37What we think is really important.
- 16:39And as part of this is that you need to be white box, not black
- 16:42box. And if I were a company or
- 16:43consumer, like how did the especially a company, how did
- 16:47the autonomous AI agent, the agentic AI agent make decisions?
- 16:52What was their thought process? One of the things we do is we
- 16:55lay that all out, you know, for every single interaction an AI
- 16:59agent or a bot has with a customer so that companies can
- 17:03go refine that by giving natural language instructions to it to
- 17:06keep getting better and better. I think that's really, really
- 17:09important because if you're black boxing it as a company,
- 17:12you don't know why the AI agent is doing something.
- 17:15So it's harder to improve. I use large language models,
- 17:19multiple LLMS every day all the time and for the reasoning
- 17:25models, when they expose their logic, it is so useful.
- 17:30I agree. I'll give you an example of my
- 17:31personal life. My wife's grandfather, he was
- 17:36born in the US, but he grew up in Ireland and my wife wanted to
- 17:40know can she get Irish citizenship?
- 17:42I took ourancestry.com, gave the password to a large language
- 17:46model, connect it to an e-mail account that I made-up and
- 17:51create it. Made-up create it and said can
- 17:53you go research and grab records to support the claim that my
- 17:57wife can go get Irish citizenship.
- 18:00It was amazing what it was able to do.
- 18:02It requested parochial grade school records from an Irish
- 18:07Parish School. It got me all this information,
- 18:10then wrote a legal memo on why it thought my wife was able to
- 18:13get Irish citizenship. The only problem kind of to your
- 18:16point, Michael, is it got the legal conclusion wrong.
- 18:20It looks I it said the legal conclusion was, yes, it's
- 18:23probably a 5050 corner case that she got it, but it exposed some
- 18:28of the logic and some of the reasoning and that's how I
- 18:31caught it might have been wrong on its conclusion.
- 18:34So that's why I just think it's first of all, cool business
- 18:36case, right? You can take your tree from
- 18:38ancestry.com, Create an e-mail account, get a legal memo,
- 18:41document supporting it by just giving some prompts.
- 18:44But on the other hand, that's why it's so important to give
- 18:47that white box treatment to really understand and the logic
- 18:51behind it, so you can still see flaws because there's flaws and
- 18:53mistakes just like human beings would make.
- 18:55When the AI makes a mistake and a customer service agent is
- 19:03actually working with a customer, do you have ways of
- 19:07trying to trap that or how how is that get managed?
- 19:12Because the mistakes can be insidious, right?
- 19:15The LLMS speak with the full confidence of quote UN quote,
- 19:19someone who knows they're absolutely right all the time.
- 19:24So one of the things we think we differentiate and we think it's
- 19:26important for any player in customer service is we have
- 19:29something called quality assurance.
- 19:31And quality assurance is using a different algorithm than a large
- 19:35language model. And we actually score every
- 19:39single AI agent or human agent interaction for accuracy, for
- 19:43tone, for resolution. And we find through that where
- 19:49humans are making mistakes or AI agents are making mistakes and
- 19:52we surface that. And so again, the white box and,
- 19:56and you're totally right, one of the flaws right now is AAI agent
- 20:00will say with absolute certainty that my wife can get Irish
- 20:04citizenship where it's a little murky, but be 100% confidence.
- 20:07And, and so we have this real strong point of view.
- 20:10You should be as a company doing quality assurance, not on a
- 20:14sample size. Right now, most companies are
- 20:16doing like 1% of interactions. They're analyzing, we say to
- 20:21analyze 100% of interactions and you can get summaries of that
- 20:25now based upon AI. And that's how you're going to
- 20:27spot issues, whether humans or AI agent.
- 20:30And what's really key about that is sometimes you're going to see
- 20:32the AI agent is handling a question or a query better than
- 20:36the human agent. But sometimes you're going to
- 20:38want to have with like my 80 year old parents, my mom will be
- 20:41upset. She's not quite 80 yet, but my
- 20:43father, my father's 80. You're going to want to have
- 20:46them talk to humans. They're not going to want to
- 20:48talk to an AI agent. And you want to get that AI
- 20:50agent accuracy into the humans. And so we really think quality
- 20:54assurance systematically is so important in the AI era.
- 20:58So the AI is also enabling you to have a far more granular type
- 21:05of ongoing QA than was possible in the past.
- 21:10Allowing us to have more granular, more ongoing and
- 21:12what's the great thing is surface those insights where
- 21:15there are mistakes and propose actions, whether go build this
- 21:19policy, go change you have you have policies that are
- 21:23conflicting with each other. Go make sure that you create a
- 21:26policy here or a piece of knowledge to go solve it.
- 21:29So has more granular insights, it's more personalized.
- 21:32And with the Gentic, AI can propose actions and make actions
- 21:36to go hit the root causes of customer dissatisfaction.
- 21:39Let's jump to some questions. And the first question is from
- 21:44Greg Walters. He actually asks a couple, but
- 21:46I'm going to go to his second one and he says regarding
- 21:51disruption, AI eliminates the middle person.
- 21:55Do you see a time where Zendesk is more than a component of and
- 22:00a tool for customer service becoming quote customer support
- 22:04and service the department? I know some people, some
- 22:08companies are out there. There's something called BPOS
- 22:11that will say you can outsource a lot of your customer
- 22:14interactions to us and we'll be in charge of the technology and
- 22:17the people to solve that. I think in the future, our
- 22:20platform needs to be reactive for our companies, but also
- 22:24proactive. And again, it's going to be
- 22:26interesting. It's going to change the
- 22:28dynamic. Right now, we're all about the
- 22:30reactive customer issues. Greg, I think that's a great
- 22:33question. By the way, in the future, I
- 22:36think we're going to be solving problems before they even occur.
- 22:40You know, with our companies. I don't know if we're ever going
- 22:42to be the customer service department itself.
- 22:46I think we're going to be a platform to help the customer
- 22:49service resolve issues before they happen and when they do
- 22:52happen, solve them. But we have debates and I tell
- 22:56everyone this every week, customer service is massively
- 22:59changing. Every week there's new AI
- 23:01innovations and we have had discussions on how does our
- 23:05platform evolve and are we going to be taking more and more of
- 23:08the load from customer service departments in the future.
- 23:10I think it's a really insightful question.
- 23:12Another question from LinkedIn and we'll jump to those
- 23:16questions on Twitter shortly. This is from Brandon Byrd.
- 23:20And, Brandon says, are there specific verticals where Zendesk
- 23:25plans to double down with AI to gain a first mover advantage?
- 23:32We think over the next five years that vertical industries,
- 23:35the more deep that you can get into tasks and procedures and
- 23:39expertise, the better you're going to be able to solve your
- 23:42customers, employees or your our customers customers issues.
- 23:47And so we right now have a really, really good foothold in
- 23:50e-commerce, fintech manufacturing and I can go on
- 23:56with the list tech. And so we are going deeper and
- 23:59deeper and taking our learnings from those verticals and
- 24:03applying them to our customer service AI driven resolution
- 24:07platform. And so I think that's going to
- 24:09be one of the secret sauces where companies help with help
- 24:13their customers get more value by differentiating on a vertical
- 24:16or industry basis. And so it's already happening
- 24:19now. I think we're in the early
- 24:20innings or the early part of the match for your international
- 24:24audience. But I think you're going to see
- 24:26over the next five years a lot more industry specific solutions
- 24:30and industry, industry specific knowledge and industry specific
- 24:35AI. We have a question from Firestar
- 24:40who says it feels like we're finally past the AI is coming
- 24:45phase and into actually making it work for customers.
- 24:50The technology works, but I think where we are in the, the,
- 24:54the, the, the, the next evolution is getting it adopted.
- 24:58It's interesting we see customers maybe don't want to
- 25:01adopt it on the, on the one end. And so one of the things we tell
- 25:04our customers to do, I've seen some, and this is learning from
- 25:07other customers is you might want to say, I have an AI agent
- 25:10available right now for you or do you want to wait 5 minutes
- 25:13for a human agent? And so a lot of people will test
- 25:16the AI agent because they've had the bad experience, Michael, you
- 25:18had with the chat bots in the past.
- 25:21And so you kind of want to nudge them to go try the AI agent.
- 25:24We've also have the human agents where there's something called
- 25:26Copilot they're using and they've been using other tools.
- 25:29And so how do you go nudge them to go get adopted?
- 25:32And once they get, they see the power of Copilot, they'll start
- 25:36using it more and more. And then just implementing with
- 25:39the customer to make sure that you're getting the data right,
- 25:42you're structuring the policies right.
- 25:44You're really taking a gentic AI and starting to run through your
- 25:47business process and workflows and creating new processes in
- 25:51workflows with AI. There is still some adoption
- 25:55challenges in the marketplace and just the fact that you can
- 25:58do a lot with three clicks, but it might take you a couple weeks
- 26:01or a couple months to get the full value stack or beyond that
- 26:05journey is another problem with adoption right now.
- 26:09So I think we've gone from does the technology work to how do we
- 26:13get employees, customers and autonomous AI agents, all three
- 26:18really embracing the technology, the adoption curve, the
- 26:21implementation adoption I think is the next frontier that you
- 26:24know, Zendesk is dealing with and other companies are dealing
- 26:26with. It's a great point.
- 26:28And of course, across different industries and different
- 26:30companies, as you mentioned earlier, there are varying
- 26:34degrees of maturity which has a direct impact on that adoption
- 26:38capability. Definitely.
- 26:40And it's really interesting, we're seeing, you know, when you
- 26:44talk about a small customer, they're looking for
- 26:48instantaneous impact and they're looking for the three click
- 26:53solutions. When you talk to an enterprise
- 26:55customer, they're really looking for embedding AI deeply and
- 26:59tasks and procedures, changing their workflows, inventing new
- 27:04workflows, having actions that take say like I used the return
- 27:08example, it's used across the ERP, across the supply chain to
- 27:12really interconnect it. So different customers are
- 27:15asking for different things. And I kind of look at it,
- 27:17there's a maturity curve for customers.
- 27:19There's a segment whether you're SMB, mid market, enterprise.
- 27:23And there are also these geographic differences people in
- 27:25the Bay Area, for instance, are ready to embrace.
- 27:28I'm originally from Ohio, my parents are from Kentucky.
- 27:31When I talk to customers in Ohio and Kentucky, they might be a
- 27:35little more, you know, where the Bay Area was a year ago.
- 27:37So you have this kind of Rubik's Cube based on geographic region
- 27:41segment and maturity to the customer where they're ready to
- 27:44go. But I think largely we're
- 27:46getting over is this hyper real and people are accepting that's
- 27:49real. I have a lot of questions of my
- 27:51own, but I'm going to, I really like to prioritize the questions
- 27:55from the audience, The audience you guys are, I always say this,
- 27:58you guys are really smart. We have a question.
- 28:01I'm still on Twitter X and we're going to jump to LinkedIn in a
- 28:05moment. Question from Elizabeth Shaw on
- 28:08Twitter who says as AI enabled customer support functions
- 28:14increase, especially with AI agents and then agentic AI, how
- 28:21important will the human touch be and when will human touch be
- 28:26the right thing? So how important is the human
- 28:29touch and when do you need to loop in the human from the
- 28:33electronic agent? We still think 10 or 20% of
- 28:36interactions are going to be human LED and they're going to
- 28:40be the most complex interactions.
- 28:41They're going to be segments of your customer base.
- 28:44Maybe it's your top customers and, and there could be people
- 28:48that just want to talk to a human.
- 28:50And so we think there's going to be two or three areas where
- 28:54humans are going to be still the central focus point of a, of a
- 28:59resolution. And sometimes, even though I
- 29:00said sometimes the AI agents are giving more empathy, if a
- 29:03customer is really upset, a lot of times a human being can do
- 29:06that really, really well. So I guess that would be the
- 29:09first point that humans, we believe are still going to be in
- 29:11the loop. The second thing that we think
- 29:14is really exciting is I know there's been a concern that
- 29:18people are just going to start cutting human agents.
- 29:20And what we've seen is something different.
- 29:23We call it there's been a service deficit up until now.
- 29:26I think most people would say when they interact with brands,
- 29:29whether they're a business or consumer, they might not get the
- 29:33best experience. And so there's this service
- 29:35deficit. We think that AI is going to
- 29:38give a service dividend where the humans are going to be able
- 29:41to free up to spend more time on the root cause of some of these
- 29:45bigger issues. They're going to be able to
- 29:47manage the AI agents. So you might have a human agent
- 29:50managing a team of AI agents to what we talked about before,
- 29:53Michael, to make sure that there's accuracy using QA and
- 29:56other tools. And we think the service
- 29:59dividend is just going to go increase customer satisfaction
- 30:02again, whether that's B to B use case or B to C use case or an
- 30:05employee, because you're going to be dealing with more the root
- 30:07cause and preventing things before they happen and you're
- 30:09going to get more personalized. And so we think there's going to
- 30:13be a couple reactive use cases. We believe, you know, people
- 30:16need to remember that there's always a human at the other
- 30:19side. Even though, and I'm sure we'll
- 30:20talk about it, Michael, in 12 to 18 months, you're probably going
- 30:23to be using your own bot to go talk to a company's bot.
- 30:26At the end of the day, there's still that human on the other
- 30:29side of the interaction, and we need to remember that.
- 30:32And this is from Preeti Narayan. And she says as AI reshapes
- 30:38customer experience, how do you see industry verticals like
- 30:43retail, healthcare and financial services evolving their customer
- 30:49engagement models? Are there industries where AI
- 30:52driven CX has become a strategic differentiator rather than just
- 30:59an operational tool? And how is Zendesk positioning
- 31:03itself to lead in those transformative shifts?
- 31:07And I'm just going to make one request to people asking
- 31:09questions, keep your questions simple because I'm reading your
- 31:13questions and I need to make sure I get it.
- 31:15OK. So AI driven CX becoming a
- 31:19strategic differentiator rather than just an operational tool
- 31:22and how you're positioning to lead in these transformative
- 31:25shifts? What's really interesting is we
- 31:27are seeing some customers see it as a a transformational tool
- 31:33that can really differentiate their business.
- 31:34And I'll give you an example. We have more and more customers
- 31:38that don't want to be customer references because they don't
- 31:41want to let their competitors know that they are getting 70 or
- 31:4580% of their interactions solved by AI agents and they're having
- 31:49this service dividend that they can go reinvest.
- 31:52And so some of our best customers don't want their name
- 31:55out and don't want the results out.
- 31:57And to me, that's kind of the proof in the pudding that they
- 32:00are getting amazing automated, automated resolution rates.
- 32:03So the AI is solving the problem before it goes to human and they
- 32:07see that as a competitive advantage.
- 32:10The way we're trying to do that is we really focus on what we
- 32:14call the resolution platform or automated resolutions and the
- 32:17whole system of the resolution platform.
- 32:20So we are looking with our customers on a day-to-day week
- 32:23by week basis on what percentage of your interactions have you
- 32:27resolved and have you automate it.
- 32:29And again, like we talked about Preeti, this was a good question
- 32:32before. We're getting deeper from a
- 32:34vertical perspective on that. And when we get those high
- 32:37automation rates, we really think you can go into that
- 32:40service dividend land. And so that's how we're
- 32:43positioning ourselves with our customers.
- 32:44We're not perfect, but we're really trying to drive our
- 32:47customers to higher and higher automation rates.
- 32:50And that's why we've come out with our new pricing model
- 32:53called automated resolutions. We only get paid if we solve a
- 32:58customer's problem with you, if we resolve it, but it has to if
- 33:02it goes back to a human, we don't get paid for that
- 33:04interaction even though it costs US money.
- 33:06And that's a way for us to kind of lead the way in customer
- 33:10service and put our, you know, put our operating model totally
- 33:14in line with our customers because they care about driving
- 33:18up the automation rate, the automated resolutions.
- 33:21And we're in the exact same boat because about how we price.
- 33:24I like that, putting skin in the game.
- 33:27Historically software companies, you know, going back to on Prem,
- 33:32they sold the product, threw it over the wall, and if the
- 33:35customer used it or didn't use it, they still paid.
- 33:39And now you're going to the opposite extreme where Zendesk
- 33:45has a stake in the customer outcome as opposed to the
- 33:50process of getting to that outcome.
- 33:53We talked about transformation and culture a little before and
- 33:55employees. This is a big transformation for
- 33:57our customers, for our employees and for our shareholders because
- 34:02how we look at operating model. In the past it was seats.
- 34:06We still have some seats. Some companies made the switch
- 34:09to interactions and now the next evolution is to outcome based
- 34:14pricing or really aligning with your customers.
- 34:17I'll give you an example. I was on an airline, I won't say
- 34:19we're in the world recently and I noticed that I was going to
- 34:22miss my connection. I got on their app.
- 34:25I think they had a beta going on with AI agents and I got about
- 34:2990% done solving my problem of re booking my connection
- 34:34unfortunately took off, I turned off my phone, no Wi-Fi, I
- 34:37landed. The AI agent hadn't solved my
- 34:40problem. I'm confident that airline,
- 34:43maybe they didn't get charged but they would have gotten
- 34:44charged $2.00 for an interaction.
- 34:46Even though I was more dissatisfied because I called a
- 34:49human agent then and the human agent didn't have any of the
- 34:52context from the AI agent. And so I felt like I was
- 34:55repeating myself. And oh, by the way, I had to
- 34:57wait like 8 hours at the airport because I missed all the
- 35:00connections because they got booked in the meantime.
- 35:03And so that's an example where if you're on a seat based or
- 35:06interaction based, the software provider probably got more money
- 35:09from my interaction with that AI agent, even though I was more
- 35:13dissatisfied. And that's not what we want to
- 35:14do as a company. We really want to align our
- 35:17customers outcomes to our outcomes.
- 35:19Here's an interesting one from Ryan Smith.
- 35:21This is really nuts and bolts, he says.
- 35:23As AI begins to resolve tickets faster and automate more
- 35:29interactions, how should CX leaders rethink success metrics?
- 35:35Do legacy KPIs like customer satisfaction and FCR first close
- 35:41resolution still matter or do we need a new framework?
- 35:46I'm going to give you an example.
- 35:47There's a lot of companies right now that use average handling
- 35:51time. So how long does a human agent
- 35:53stay on, whether it's a phone and e-mail, a chat, you name it,
- 35:58a web form, how long does it take for that human agent to
- 36:02resolve the problem? And you want lower average
- 36:04handling time. One of the things that we tell
- 36:07our customers is once you implement automation, automation
- 36:12takes let's call all the easy and medium hard interactions
- 36:16away. And so the humans are left with
- 36:18the most difficult ones. And so we tell CX leaders, your
- 36:22average handling time is going to go way U OK, and you're going
- 36:25to be like, why did this happen? I just imlemented AI agents and
- 36:29automation and then I gave my human agents copilot, which
- 36:31should make them more efficient, but their average handling time
- 36:34is going up. And so we're starting to develop
- 36:37more frameworks with our customers.
- 36:38So they're not surprised when you explain that, you know, a
- 36:42customer CX leaders like of course, because they're getting
- 36:44the 10% or 20% of interactions that are really, really
- 36:48difficult that take a long time. OK.
- 36:50And so we're encouraging more AB testing and different
- 36:54frameworks. I do think still NPS or CSAT
- 36:57matters. I think there's going to be some
- 36:59new frameworks and some new metrics like how much of how can
- 37:05you reduce interactions not for reducing interactions sake
- 37:09because you're solving the root cause.
- 37:11And so I think over time how quickly you reduce the
- 37:15interactions that keep coming up is going to be a metric in the
- 37:18future. Average handle of time is going
- 37:20to be less of a metric. I still do think first contact
- 37:24resolution is important. If you look at the psychic, the,
- 37:27the, the psyche of customers, they rather wait longer to
- 37:32interact with an AI agent or human agent, read a rate on
- 37:34hold, quote on hold for 5 minutes and get it solved
- 37:37immediately rather than being moved around two or three times
- 37:40between different humans and AI agents.
- 37:42So I think first contact resolution, but Ryan, I think
- 37:44your prep supposition that we're going to have an evolution in
- 37:49metrics is absolutely going to happen with CX leaders.
- 37:52It's so interesting how AI is driving change in so many
- 38:00different directions. The the tentacles are far
- 38:03reaching and the implications are very profound.
- 38:07I mean, just just look at what you just described with your
- 38:11pricing policy. I mean, that's such a dramatic
- 38:14shift. I don't think we know, Michael,
- 38:17all the implications of AI and I think people that are certain on
- 38:22how it's going to go over the next, you know, 15 or 20 or 25
- 38:24years are probably wrong, but it is profound.
- 38:29You know, one of the things I do is every two weeks, I do a deep
- 38:32dive on what's changed in AI in the last two weeks and how could
- 38:36it impact our customers and how can we provide more value to our
- 38:39customers. And I would tell you, you know,
- 38:41if you're in the customer service industry from 2009 to
- 38:452019, like I was at another company, like the changes you'd
- 38:49have in years are about the same pace that we now have in weeks.
- 38:53And so there's just so many profound implications on the
- 38:56future of work. You know, I have a 22 year old
- 38:59daughter and a 20 year old son, and they asked me with AI, what
- 39:01should I do? And that's like kind of an
- 39:03existential question about what skills and capabilities they
- 39:06should learn. Because I think the nature of
- 39:09work is going to just massively shift over the next 10/15/20
- 39:12years for instance. As CEO of Zendesk, what does
- 39:18this rapid change? It's extraordinary that you have
- 39:21to look every two weeks to see what's changed in this AI world
- 39:26that could have some type of potentially profound impact on
- 39:30the market and on your company's Zendesk.
- 39:33It really changes that. You need to make sure you're
- 39:35getting a lot of signals from customers in particular.
- 39:39And like when we're doing well and when we're failing, I think
- 39:41is really, really important from the industry at large is really
- 39:45important. And I think it really what I
- 39:46encourage employees is you've got to be in a learning
- 39:49employee. You've got to learn AI, you've
- 39:51got to embed it into your daily workflows.
- 39:53You got to stay on top of it because that is a skill.
- 39:57And I think it's going to differentiate employees going,
- 39:59including CEOs in the future who really understands AI.
- 40:03Who's embedding it in their daily life?
- 40:05Who's using it strategically and tactically?
- 40:08And so I think always learning, I, I give advice to people, you
- 40:11have to always be learning and you have to be always learning
- 40:14about AI. It is going to impact your job
- 40:17in some shape, way or form over the next five years, no matter
- 40:21what job you're doing right now. This is from Arsalan Khan, who
- 40:26is a regular listener. He asks great questions.
- 40:28He says. When will we be able to replace
- 40:32humans with AI? What happens to people?
- 40:35Will manager AIS be able to write performance reports for
- 40:40other AIS and humans? I'm an optimist about this
- 40:44point. I think AI is going to enhance
- 40:46people's lives and I think we've seen with every major revolution
- 40:51in the world it create more jobs but create different jobs.
- 40:54So I'm still an optimist here on what AI is going to do.
- 40:59I think, you know, it's an insightful question about
- 41:01performance reviews. I think right now where we are,
- 41:04I think AI can give summaries. If you get a lot of 360 feedback
- 41:08summaries for performance reviews, insights for
- 41:11performance reviews, it's probably going to happen in
- 41:13151015, maybe even sooner that you are going to have
- 41:16performance reviews written by AI and edited by a human.
- 41:20I think it happens right now. I've talked to some startups
- 41:22that are doing that right now. And so I think, again, we just
- 41:25got to really understand at the end, the other end of that is a
- 41:29human getting a performance review.
- 41:32And you've got to be really, really confident in the
- 41:33technology that it's getting all the data sources, all the
- 41:36signals, you know, honestly in a good way and transparent on,
- 41:41again, that white box theory, Michael, on what it got and how
- 41:45it came to its reasoned conclusions about your
- 41:47performance. We're already there with quality
- 41:49assurance where we are evaluating 100% of interactions,
- 41:53whether it's a human agent and AI agent and scoring them.
- 41:56It's not a formal performance review, but it's like indicators
- 42:00of success. So we're already there.
- 42:02I think there's some really profound implications for human
- 42:06society over the next, again, 5101520 years about how jobs are
- 42:11going to evolve. Your long time listener, I think
- 42:15alluded to it where he said you're going to have different
- 42:18jobs. And we think there's going to be
- 42:19managers, human managers, managing teams of AI agents
- 42:23going forward. And that's going to be probably
- 42:26one of the new jobs that's created from this AI revolution.
- 42:29Arsalan Khan has been waiting a long time, so I'm going to pop
- 42:31another question from him and he says on Twitter, can you talk
- 42:37about the role of data in your product and AI?
- 42:43One of the things that we have an advantage as a company is we
- 42:46do 4.6 billion resolutions a year.
- 42:51And so there's a huge, huge data set that can help post train our
- 42:56models. So they're not just large
- 42:58language models off the shelf. You post train those models to
- 43:01get more accurate, OK, it's based on that 4.6 billion
- 43:05interaction data set. It's based upon to the earlier
- 43:09questions, deeper vertical information.
- 43:12And we're able now with knowledge bases and pulling
- 43:15things from Confluence or SharePoint or others, we're able
- 43:18to get even more knowledge and more data.
- 43:20But again, your models and your answers are only as good as your
- 43:26data. So I think, again, he's hitting
- 43:27on something really, really strong that you want to get as
- 43:30much data ingested, but you want, even if it's unstructured,
- 43:34clean scrub data that has some effectiveness in it.
- 43:39And so I think we're getting to a world where the more
- 43:43structured and unstructured data that you have that's scrubbed,
- 43:46the more precise with post training you can do to give
- 43:49answers to customers or employees or other businesses
- 43:54more accurately. And so I do think we're in a
- 43:57situation where data is really, really important.
- 43:59We've gone through a whole thing internally at Zendesk the last 2
- 44:021/2 years that wasn't that sexy. But we've spent a lot of time
- 44:07with our internal data, scrubbing it, putting it into
- 44:10the data lake so that we could go use it in a bunch of
- 44:12different AI applications. Until we did that, we kind of
- 44:15had a bottleneck on using our data.
- 44:18I'm not talking about our customer data.
- 44:19I'm talking about internal data to go run the business.
- 44:22I think smart companies have taken this on as a project
- 44:26because it's not fun, it's not sexy, but it's going to have a
- 44:33dramatic impact on how a business operates in the coming
- 44:39years. It's the foundation.
- 44:41It's the foundation and there's some really cool use cases right
- 44:43now. I've been talking a bunch of
- 44:44startups on reporting and analytics and we just acquired a
- 44:47company to do some really cool stuff in reporting and analytics
- 44:50that you're getting to the point where instead of getting a
- 44:54report, you can go natural query what is the year over year
- 45:01increase in automated resolutions in Malta for our
- 45:04customer base. And you can go get that data
- 45:08with natural language queries that can give you really, really
- 45:11insights. And you can go prompt reporting
- 45:14analytics to give you more forward thinking things.
- 45:17What should I be thinking about to run my business that I'm
- 45:19concerned about Malta? And I'm concerned about how our
- 45:22customers are getting value from our product.
- 45:23And you can get prompts and ideas back.
- 45:25And I think that's again, shifting from old static reports
- 45:29to real insights. And again, you can do that
- 45:31through natural language. Brandon Byrd on LinkedIn says
- 45:35when you increase self-service levels, how do you address agent
- 45:39burnout? And I'm going to ask you to
- 45:40start answering these, Tom relatively quickly because I
- 45:44want to get to everybody and we have, we have a lot of questions
- 45:47and I have my own questions. You can self-service, you're
- 45:50answering the easy questions and human agent burnout is real.
- 45:54It's one of the most stressful jobs possible.
- 45:57I think you can help them get better at their job through
- 46:00Copilot and so that you're serving them the answers, which
- 46:04I think can go really help with burnout because they're getting
- 46:07more accurate, more empathy, and they're getting it right with
- 46:10customers. Here is a question from Preeti
- 46:14Narayan, who says in her region there is a lack of Computer
- 46:21literacy. And what can a company like
- 46:26Zendesk do to make those computer interactions easier for
- 46:31folks who are not computer experts?
- 46:33We have a foundation and we have a program.
- 46:36We, we've trained over 10,000 people that might not have the
- 46:41same kind of access. There are, there is in other
- 46:44parts of the world to become customer service agents, human
- 46:46agents and giving them the computer skills and the customer
- 46:50skills. And so that's one of the things
- 46:52we're doing at Zendesk. You know, doing good, doing
- 46:55right is doing good. And I think, you know, more and
- 46:58more companies are going to have to figure out how to help people
- 47:01become more computer, a little bit more customer service
- 47:04literate, and become great customer service agents.
- 47:07Tom, how has this AI trajectory affected how you think how
- 47:14Zendesk thinks about talent, team composition and jobs?
- 47:21We haven't gone as radical as other people.
- 47:24I know they've said, hey, you need to justify any job
- 47:26requisition with why AI cannot do it.
- 47:29But when we talk to talent now, I know one of the first
- 47:31questions I ask is how are you using even for CXC suite
- 47:35employees? How are you using AI in your
- 47:38everyday life, you know, professionally and personally?
- 47:40Because in this AI era, if you're not using the tools, if
- 47:44you're not experimenting, you're not getting better.
- 47:46I don't think you're going to have the skills and capabilities
- 47:48to succeed. So that's one of the things that
- 47:51we've talked about. And then we're going to start a
- 47:53big push. We've kind of had a bottoms up
- 47:55approach up until now about how our employees use AI and we've
- 47:59opened up like an enterprise license for Chachi BT I use it
- 48:02everyday. I think you're going to see a
- 48:04little more top down pushes to encouraging employees to use,
- 48:09you know, these tools more because I think it's good for
- 48:11the company, it's good for their career.
- 48:13Can you talk about the future and where is AI customer
- 48:20experience headed? We believe at Endesk that 80 of
- 48:24interactions will be through automated resolutions in five
- 48:27years. So do not involve a human in any
- 48:30way, shape or form. We think 100% of interactions
- 48:34will involve AI in some way, shape or form within five years.
- 48:38That means that last 20% that humans are solving, they're
- 48:41going to do that with the assistance of AI.
- 48:43We do think there's another trend though that once you start
- 48:46getting quick satisfaction, it's the opposite of dealing with
- 48:50that chat bot. Michael, you and I talked about
- 48:52earlier. We think consumers, employees
- 48:55and other businesses are going to interact with those companies
- 48:58more and more and have more interactions.
- 48:59So we think interactions are actually going to explode,
- 49:03particularly when you start using your own AI agent and
- 49:06tasked them with doing something.
- 49:08And so those are some of the trends that we are going to see
- 49:10in AI over the next five years. What about the impact of AI on
- 49:16jobs in general across the economy?
- 49:20Is that something that you have thought about at all?
- 49:24Where, where is this going from A from a jobs?
- 49:26What jobs will be affected? What, what won't?
- 49:30And so forth. Definitely thought about it and
- 49:32I'll give more of a personal story and I alluded to it
- 49:34before. I have a 22 year old daughter
- 49:36who just recently graduated from college and I have a 20 year old
- 49:39son who's in college and they've asked me with the advent of AI,
- 49:43what kind of what should what job should I be doing?
- 49:45I think an answer would have been 15 years ago going to
- 49:48coding, right? Software engineering is a great
- 49:50place to go. I'm really telling them that
- 49:53again #1 be really, really knowledgeable about AI.
- 49:57Use it in your personal life, use it in your professional.
- 49:59That is going to distinguish you from other people #2 I tell them
- 50:03it's going to go back to the basics.
- 50:05If you can read well, you can write well, you can analyze
- 50:09well, you're going to be fine. Because no one can predict what
- 50:13the job markets going to be evolve over the next 40 to 50
- 50:15years. And that's kind of their career
- 50:17trajectory of, you know, young 20 something people is what's
- 50:20going to happen in the next 40 or 50 years.
- 50:22And we've seen what's happened in the last, you know, three
- 50:24years with AI. So really encourage people to
- 50:27get those basic skills. And 3rd, I encourage them to
- 50:30learn how to prompt. And it's kind of tied to 1 and
- 50:322, how to use AI. And I think I've experimented
- 50:36the last two or three years and my prompting is getting better
- 50:38and better. I think that skill is going to
- 50:41be great. And it's kind of interesting.
- 50:42I think some of the liberal arts education, when you write well,
- 50:45you analyze well, you read well, is going to be more and more
- 50:48important in prompting. On the prompting topic, folks
- 50:52listening, we did a show specifically on prompting for
- 50:57business leaders. So check out cxotalk.com and go
- 51:01back a few episodes and you'll find that show.
- 51:03It was really, really great very, very quickly.
- 51:07Data privacy and security, cybersecurity of your data.
- 51:12Can you talk about that? Because that's obviously it's an
- 51:15extremely important topic. We take data privacy and it's
- 51:18kind of that white box approach really, really seriously.
- 51:21And so we are out in the forefront right now on trust,
- 51:26the amount of certifications that we have to make sure that
- 51:29we have a lot of confidence that we're going to protect with our
- 51:32customers. Your data is really paramount to
- 51:36us. 2nd, we want to make sure that we can personalize things
- 51:42to you, but not in a creepy way with companies.
- 51:44And so we work with companies on that.
- 51:46But I think we're going to be in an era where that white box
- 51:49approach versus the black box approach and exposing that not
- 51:52just to companies, but employees, consumers and other
- 51:55businesses on how people are using your personally
- 51:59identifiable information, how they're securing that is going
- 52:03to be more and more of a topic over the next 5 to 10 years.
- 52:05And but it's absolutely the four front of our AI approach.
- 52:09And with that, I'm afraid we're out of time.
- 52:14Tom, thank you so much for being a guest on CXO Talk.
- 52:17I'm so grateful to you and for your team at Zendesk.
- 52:20Great questions, great flow. Really, really appreciate your
- 52:23time, Michael, and I really appreciate the opportunity to be
- 52:25on with you today. Folks, before you go, now would
- 52:28be an excellent time. To subscribe to the CXO Talk
- 52:31newsletter. Go to cxotalk.com.
- 52:33We'll keep you updated on upcoming shows.
- 52:36Have a great day everybody, and we'll see you soon.
- 52:39Take care.