← The Agile Brand with Greg Kihlström®: Expert Mode Marketing Technology, AI, & CX
The Agile Brand with Greg Kihlström®: Expert Mode Marketing Technology, AI, & CX
#761: Treasure Data CEO Kaz Ohta and CMO Karen Wood on the AI-driven reinvention of marketing
00:26:33
Loading summary
Transcript70 lines
- [00:01]
Pura Brand Announcer
Close your eyes. Summer smells like sunshine, fresh citrus and salty air. What if your living room could feel just like that? With Pura's new summer collection, it can restore your sense of well being with fragrances designed to move with your day. From bright, energizing mornings to soft, relaxing evenings. Make the invisible unforgettable. This season. Visit pura.com to find your new favorite summer scent.
- [00:28]
Strayer University Announcer
At Strayer University, we help students like you go from Will I to why not? For over 130 years, we've been innovating higher education to make it more affordable, accessible and attainable so you can reach your goals. Go from thinking Can I? To Yes, I can and keep striving. Visit strayer.edu to learn more. Strayer University is certified to operate in Virginia by Chev and its many campuses, including at 2121 15th Street north in Arlington, Virginia.
- [01:01]
Karen Wood
The agile brand.
- [01:10]
Greg Kilstrom
Welcome to Season seven of the Agile Brand where we discuss the trends and topics marketing leaders need to know. Stay curious, stay agile, and join the top enterprise brands and Martech platforms as we explore marketing technology, AI, E commerce, and whatever's next for the Omnichannel customer experience. Together we'll discover what it takes to create an agile brand built for today and tomorrow and built for customers, employees and continued business growth. I'm your host Greg Kilstrom, advising Fortune 1000 brands on martech, AI and marketing operations. The Agile Brand podcast is brought to you by Tech Systems, an industry leader in full stack technology services, talent services and real world application. For more information, go to teksystems.com to make sure you always get the latest episodes, please hit subscribe on the app you listen to podcasts on and leave us a rating so others can find us as well. And now onto the show.
- [02:07]
What if the very foundation of your marketing cloud, the one you've spent years building, is at worst actually giving your AI amnesia and at best missing key opportunities to connect with your customers? Agility requires more than just reacting quickly. It demands an intelligent foundation that can anticipate customer needs and empower your teams to act on them with confidence and precision. Today we're going to talk about the AI driven reinvention of marketing. For years, marketing clouds have promised a unified view of the customer and personalized engagement. But many of the leaders are still struggling with fragmented systems, disconnected data, and unpredictable costs. Now the demands of AI are exposing the foundational cracks in that legacy architecture, forcing a fundamental shift in how we think about the relationship between data and intelligence and execution. To help me discuss this topic, I'd like to welcome Kaz Oda, CEO, and Karen Wood, CMO at Treasure Data. Kaz and Karen, welcome to the show.
- [03:06]
Kaz Oda
Thank you very much for inviting us.
- [03:07]
Karen Wood
Hi. So happy to be here.
- [03:09]
Greg Kilstrom
Yes, yeah. Looking forward to again. Yeah, yeah, exactly. Welcome back to the show. Yeah, exactly. And you know, it's been great, you know, being at the show here. We'll talk a little bit more about CDP world in a minute here, but before we do that, why don't each of you give a little background on yourselves and your role at Treasure Data.
- [03:29]
Kaz Oda
Sure. So I'm Kaz Ota. I started this company with two of my co founders, and when I started this company, I was cto, but I became a CEO five years ago. So pleasure to meet you.
- [03:41]
Karen Wood
Yeah. And I'm an interim CMO at Treasure Data. Came up through the product marketing role and also have been in the data and analytics and CDP space for 15 over a decade.
- [03:53]
Greg Kilstrom
Yeah, yeah, sounds good. So Treasure Data has long been known by those of us in the industry as a pioneer in the CDP space. For listeners who might be familiar with that part of your history, you know, how would you describe Treasure Data today? You know, and what kinds of challenges are you solving for your enterprise customers? With the launch of the AI marketing cloud, of course.
- [04:14]
Kaz Oda
So our product vision is called connected customer experiences. Right. So we actually don't like the definition of current cdp, which is a part of Martech, but we believe customer data can transform the entire enterprise because there's no business who doesn't have customer. Right, right. So customer data can empower marketing. Obviously that gives you the shortest amount of time for roi, but it can transform sales, customer support after service, know, optimizing inventory that creates more profits, dynamic pricing compliance, et cetera, et cetera. Right. So we come in and saying, you know what, you have to know the customer better. And then that's why customer data. And obviously now AI can transform the entire enterprise. So that's our point of view.
- [05:08]
Karen Wood
Yeah. And there's also a new. A new opportunity to consolidate costs through the power of AI that I think we're capitalizing on with the launch of our AI marketing cloud. And then another thing I'll say is CDP's for a long time have increased operational efficiency by decreasing the amount of time it takes to launch a campaign or get marketing work done. And with AI, it's become exponential. And again, it's a natural progression of where we started.
- [05:32]
Greg Kilstrom
Yeah, yeah. One of the things that struck me, Kaz, when you were on stage this Morning talking about we've seen the chief Martech diagram grow and grow and grow. And yet that contrasts with, you know, the average number of SaaS platforms per company is actually, you know, it peaked. Yeah, it peaked a couple years ago and now it's decreasing, which I think says a lot of things. You know, there's, there's a lot of consolidation going on, there's cost cutting, there's, there's all of these things, but we still deal with a fragmented and very complex ecosystem in Martech and CX and things. You know, from your perspective, what's the single biggest architectural flaw in some of those, like legacy systems that the AI era has exposed?
- [06:17]
Kaz Oda
Yeah, I would say, you know, AI needs better data. If you're garbage in, garbage out, and the more tool you have, you just have more fragmented data. Right. So what's interesting is, you know, around 20, 20 people purchased so many solutions under the name of digital, you know, transformation. And there were some Covid boost people's, you know, behavior went digital. So a lot of company needed it, but now it got back to the reality. Right. So that means, okay, maybe a lot of world over invested a little bit. Right. And then the money poured into SAS and software actually created a lot of feature, I would say bloat. Right. You know, think about you. You probably use a lot of SaaS, but do you use 100% of your futures?
- [07:08]
Greg Kilstrom
Probably not. Probably never. Yeah.
- [07:10]
Kaz Oda
Right. Or probably never. Right. So that's happening everywhere. So and then, you know, the CIO and CFO is asking, okay, is this underutilized? You know, you have to get rid of it, et cetera, et cetera. So at best, the IT budget, market budget is probably flat or even probably declining. So that's one, you know, force for consolidation. The other one obviously is AI, which is if you are using one platform, you have clean data, it's more consolidated. Right. So that powers better agent. So within Treasure Data, it's kind of interesting, we have CDP that holds customer data. We're also consolidating ERP HR side of the system security too, so that we have, you know, consolidated data within a few platform, not hundreds. Right. So I think it's not just Martech, it's just happening broadly across all of the software space right now.
- [08:10]
Greg Kilstrom
Yeah. And I think another part of this whole platform ecosystem thing is this idea of the monoliths, the all in ones. A lot of companies investing in these, sometimes investing in multiple ones, even though they're all in one. So the paradox that that is. But you know, and these platforms turn into these walled gardens, so to speak. So, you know, and then on the, on the other extreme there's composable.
- [08:37]
Kaz Oda
Right.
- [08:38]
Greg Kilstrom
Which poses its own, you know, instead of technical debt, there's, I've heard it called integration tax and things like that. So, you know, your platform has been described as unified but interoperable by design. You know, how do you, how do you find the right balance there?
- [08:54]
Kaz Oda
Yeah, I think it's a really good question. I would say you have all these large gigantic software company who grew with a lot of M and A that creates Wall Garden. And in, you know, many reasons you have to adapt, it's safe, it's proven in brands. Right. But at the same time with the practical reasons, we need to have all these large platform talking each other. Right. Well, Treasure Data has 600 employees, but we're still 100 times smaller than let's say Salesforce or Adobe. So vendor like us, we have to have this multi cloud model interoperability as an edge against those walled garden vendors because it's natural strategy and that's what customers want. Right. And then composability is interesting because you now can compose multiple small software vendors and then create the one that gives you flexibility. But that means that's an overhead. So I think for buyers, it's actually really interesting time because you have a choice of those three, almost like a Lego block. Right. And then you have to carefully choose the right balance because you don't want to go all into Wall Garden because then you don't want to be relying on just one vendor. Right. But you don't want to be all the way to composability because then you almost need to manage everything by yourself.
- [10:26]
Greg Kilstrom
Yeah, I mean there needs to be kind of a hub in the hub and spoke model, so to speak. And yet, you know, to, to the point of the all in ones, I work with, you know, several enterprise orgs as a, as a consultant, none of them are like, they all own at least one or more of those all in ones, but yet they have so many, you know, they're averaging 80 some platforms at once. So even the all in ones aren't really all in one.
- [10:50]
Kaz Oda
Exactly.
- [10:51]
Greg Kilstrom
Yeah.
- [10:51]
Karen Wood
And they can't even, they can't prove the ROI in those all in ones too. So it's all kind of a mystery blurred together.
- [10:56]
Kaz Oda
And they're great solution. It's just, you know, companies needs flexibility over, you know, those large vendors sometimes. And then I think we cracked that. Yeah, right.
- [11:07]
Greg Kilstrom
Yeah. So another thing I want to talk about, you know, one of the big announcements here is on AI agents and an agent hub.
- [11:18]
Kaz Oda
AI marketing super agent.
- [11:20]
Greg Kilstrom
Yeah, yeah. So why don't we start there, Just, you know, tell us a little bit about what that is.
- [11:24]
Kaz Oda
Yeah, of course. So today at CDP World we release marketing super agent. So what it does is it Help you launch 10 next more campaigns with the health of this agent. So our core belief is, you know, at the keynote I talked about, okay, there's a lot of evolvement in the AI space and model is getting improved and everyone is saying, you know what, next year it's going to be in an AGI moment. And to be honest, in some sense or another, it's already happening because they are better programmer than myself already. Right. But at the same time, you need a well defined context like problem to solve and an output. Otherwise they can't solve any problem. So the point I made on the stage is right now there are models that can solve about, you know, an hour length of the time with 50% success ratio. And that's great because it's exponentially growing. The number of the lengths of the task you can solve with the 50% success rate is growing. But at the same time, I don't want the colleague who solved the problem 50% of the chance. I'd rather flip the coin and decide, right? So if you are actually adding more nines, which is okay, if you are trying to solve any problem with the 99.9% success rate, the length of the task AI, current AI can solve is only three seconds, right. And then that's short, right. Five years from now it's going to be 30 minutes. So it's getting into a little more reality. Right? So our theories right now is you build a bunch of agents right now we solve maybe three seconds worth of that problem. Next year it's going to be 5 and then 10. So it's going to be growing. Right. So the product we released today is we created bunch of small task agent and then we also built this one orchestrator agent that actually splits the task into multiple small agents task. For example, a lot of marketing starts from research. Okay, please research about my competitors. Put that into the slide. You know, the task consists of, okay, the web researcher deep research agent and an AI office agent. We can also have data analytics agent so that you can compare the competitor's revenue versus your own revenue within your cdp. We can also create segments and journeys and et cetera and et cetera. And what's amazing about this is, so we released one first AI agent called Audience Agent. It's like creating segments, explore some data about it. And a lot of tasks took almost like 10 minutes to finish. And after 12 months, we can finish within a minute now. So there was a 10x improvement. And right now, by using our marketing super agent, you can create the complete, nurturing journey with the email templates and Images within about 10 minutes. But think about it, next year it's going to be, you know, in one minute.
- [14:43]
Greg Kilstrom
Yeah, right, yeah.
- [14:44]
Kaz Oda
So there's a lot of tasks that gets faster and faster over time. And our core belief is, okay, we keep adding those agents that might take 10 minutes today, but one year after, it's going to be one minute. And it's become more practical. So that's why we're really excited. And every single marketer. I want to test this product.
- [15:05]
Greg Kilstrom
Yeah, yeah, it looked very cool on stage today, so. Yeah, yeah, yeah, yeah, absolutely.
- [15:11]
Karen Wood
I wanted to address Agent Hub, though, because you asked about that as well. So that was another thing that we announced this week. You know, the way that we're architected, on top of our CDP sits the AI Agent Foundry. And that's really the foundation for building like Audience Agent that Kaz just mentioned, that's in our product. Other productized agents, but also agents that our professional services team can build, our partners can build, can all be built on Foundry. And we've been building a lot of these agents. And so we have basically an ecosystem or a catalog or even a marketplace, you might say, of agents. And what we found is when our customers wanted work with us on agents, they're trying to solve for a specific use case with a specific outcome, which is why those agents exist. So the publication of the AI Agent Hub was really to make all of those available so that people could peruse this catalog. And we're developing so many of these agents that it's growing and growing.
- [16:04]
Kaz Oda
And a super agent can pick up which agents to use from the Catholic.
- [16:08]
Karen Wood
That's exactly right.
- [16:09]
Greg Kilstrom
Well, it also seems like, and correct me if I'm wrong, but it also seems like the agents that are kind of narrow in scope can be very. Not only built for purpose, but built for the customer itself. Because to your point, how one organization does their marketing in one, you know, in social media, you know, it's going to be different. I've never approached it the same way at two different orgs, so that the kind of, the narrowness is a, is a benefit. And yet it's all, I guess, can you talk a little bit about the, the benefit of Being them being also being tied to the CDP and the, you know, the, the diamond record as well.
- [16:46]
Kaz Oda
That's actually a really great question. So first of all, you know what we have learned last 12 months is a lot of people did POC and then according to MIT research, 95% failed. The biggest finding and learning for us is it's like a little bit of security and permission side. Right? So let's say CDP stores pii, which is the most important asset of the data in the company. So let's say Greg, you're CEO of the company, you talk with the agent, that agent should reply anything. But if intern comes in and ask any questions, you should restrict it.
- [17:30]
Greg Kilstrom
Right?
- [17:32]
Kaz Oda
You don't want to return any sensitive information and it gets leaked. Right. So Sati and Nadella probably a couple years ago said, oh SaaS is dead because you can build agent outside that operates SaaS. Now the argument is okay, if AI can operate all the SaaS, that means this has a super permission. If this agent gets hacked, you basically leak every single record of every single system. Not going to be sustainable. So what we have built is this permission system that tightly integrated with CDP and AI on top of it. So what it does is okay, agent itself has the permission and Greg, you're the CEO, so this agent recognize you're the CEO permission. We merge two permission together, agent and human and we operate in this real and it's same for in turn. Okay, you have this agent permission and you have an intern permission. You merge it and you operate in this limited permission environment. Right. So that you actually make sure there's no, you know, leakage of the pii. So what's interesting is the benefit of building AI on top of CDP is you can actually inherit all of the permission control you built for your own PII data. And that's a moat for us.
- [18:54]
Greg Kilstrom
Yeah, yeah, definitely.
- [18:55]
Karen Wood
And unique to treasure data too. I would say if you think about the clouds and the way they architected their CDP is to really just enable all of the tools that they acquired. They're not, they don't have the same solid foundation of security that then can transcend over to the AI.
- [19:09]
Greg Kilstrom
Yeah, because I mean otherwise then you're talking about, I mean there's, there's the consent from the consumer side, but there's also the internal permissions as well. So you know, consent management is its own challenge and all that kind of stuff. But the internal. Yeah, I, that, that does seem like a, like a unique, unique feature. Yeah, exactly. Yeah.
- [19:27]
Kaz Oda
And I also think, you know the other learnings we had was AI needs a lot of attention.
- [19:33]
Greg Kilstrom
Yeah.
- [19:34]
Kaz Oda
So you know, I know a lot of vendors say okay, you start our AI immediately, all of a sudden you have all these great things. I think it's a pipe dream to be honest. AI is a probabilistic software while SaaS is deterministic, which means probably, you know, when you introduce, it's not tuned right. There's no AI that works across everyone. Maybe your AI agent initially can accomplish the task 80% of the time, but if it goes to 99%, 99.99 then that becomes useful because that's exceeding human grade level of the task. So that needs a lot of tune up, that needs a lot of continuous involvement with AI. It's almost like you versus your manager, constantly directing together. Yeah, right. So I think the adoption of AI of especially on enterprise side is underestimating it. At the same time, probably software companies, you know, is probably setting too much of the dream. Right. So one of the difference I'm trying to make is we have this like post sales organization and the experts on AI comes in and then we actually hand hold them to go through this journey together by setting the realistic expectation. Right. So this is difficult. It's not like you have an AI and you're great, if only.
- [21:11]
Greg Kilstrom
Right?
- [21:12]
Kaz Oda
Yeah, exactly. So that's another approach we're taking.
- [21:17]
Greg Kilstrom
Yeah, yeah, well and to that end and to the study that found 95% of projects not achieving the desired ROI within your engagement suite you have an ROI reporting agent and to me that seems like a powerful component. Can you talk a little bit about that?
- [21:39]
Kaz Oda
So the biggest dilemma Treasure Data has was okay, we believe data and AI is really valuable piece. But if you go to enterprise, let's say if you're paying, I don't know, 1 million dollar per CDP, they actually pay about 3 million to 4 million for actual channels. Which means email, SMS and all these. Right. Because it's easy to understand there's no company who doesn't send emails either. So it's must have. Right. And then by having CDP and other system it makes it difficult to attribute the revenue.
- [22:18]
Karen Wood
Yes.
- [22:19]
Kaz Oda
And our dilemma is also a lot of attribution go to channel at cdp. So we thought, you know what, okay, it's actually great to use two solutions but only if you want, if you use our engage, we will do all of the sending within all these channels and we can activate, you know, attribute all those behaviors so that you can exactly calculate how much revenue you're making through our email marketing automation product in CDP that becomes a true end to end so that you don't have to prove that to your CFO or CIO on okay, how much money we're making. We have this exact number.
- [23:01]
Karen Wood
What happens so often with CDPs is they can't show the ROI because they have to rely on this totally different tool to go in and find the roi and then how could they trust it? Even within a cloud ecosystem, it's still separate and segregated into that external tool. So it's unique to have, you know, because we're all embedded in all in one. You can actually attribute the revenue of the CDP within the email part of our solution.
- [23:24]
Greg Kilstrom
Yeah, I mean that just seems like a win win for everybody involved, so. Yeah, definitely, definitely. Well, it's been great talking with you both. Couple last things as we, as we wrap up here. You know, we're here at CDP World in Las Vegas. Lots of great conversations. Let's say, let's go out maybe a year from now. What are we going to be talking about? What's, you know, what's what, what are your things? What are you thinking about?
- [23:48]
Kaz Oda
Agentic World?
- [23:49]
Karen Wood
Yeah, sure, I can bring. Yeah. So next year we announced today that next year CDP World is actually. We're changing the name to Agentic World because our market is evolving. The world is evolving. Of course CDP is still at the core of agentic and AI strategy but that's going to be next year in Miami October 6th and 7th at gentechworld.com you can pre register now. So that's exciting. Definitely looking forward to that. Yeah, yeah.
- [24:15]
Kaz Oda
And then for me, you know what we need to do is we actually bringing real world success of marketing SuperAsia. Right. We want companies presenting. Okay, here's what we have done. And then launch 10x more campaigns with these solutions. Right. So working really hard on building more agents and then guiding and navigating our customers to adapt it. A lot of hand holdings. Yeah. So you know, I think in the end we can say whatever but you know, ultimately customers have to succeed and we want to hear more story around that. Right. And then we want to be the company who navigate this, you know, AI wave. Right. And then we think we're best positioned to do that because we are CDP company.
- [25:06]
Greg Kilstrom
Yeah. Love it, love it. Well, yeah, looking forward to next year. So one last question for each of you. I ask this to everybody. I added this question since you were last on the show, Kaz. But what do you do? Yeah. What do you do to stay agile in your role and how do you find a way to do it consistently?
- [25:25]
Kaz Oda
I don't know. I think there's a little bit of a balance. Right. So I think there's, you know, I've been here for five years. There are period where, you know, things are going well. So at the time, CEO's job is really just watching and delegating the team. But there are some times, okay, you know, there are some threats coming in, then it becomes a little bit of founder mode. Right. So I'm trying to be really sensitive about when we become both like when I become idle modes. Right. And that's fun. But at the same time, you know, you know, we need different modes in different period of time. Yeah.
- [26:05]
Karen Wood
I think it's really important to stay agile by listening to customers and constantly being in conversation with them, hearing their needs, hearing how they're solving their challenges. It helps us stay abreast of the market changes and the market dynamics, but through the lens of the customer. So I just think always being close to the customer is the secret sauce for agility.
- [26:22]
Greg Kilstrom
Yeah. Love it. Well, again, I'd like to thank Kaz Oda CEO and Karen Wood, CMO at Treasure Data for joining the show. You can learn more about Kaz, Karon and Treasure Data by following the links
- [26:32]
in the show notes. Thanks again for listening to the Agile brand brought to you by Tech Systems. If you enjoyed the show, please take a minute to subscribe and leave us a rating so that others can find the show as well. You can access more episodes of the show@theagilebrand.com that's theagile brand.com and contact me. If you're interested in consulting or advisory services or are looking for a speaker for your next event, go to www.gregkilstrom.com that's G R E G K I H L S t r o m.com the Agile brand is produced by Missing Link, a Latina owned, strategy driven, creatively fueled production company. From ideation to creation, they craft human connections through intelligent, engaging and informative content. Until next time, stay curious and stay agile.
- [27:28]
Karen Wood
The agile brand.
- [27:36]
Pura Brand Announcer
Refresh your space for summer during PURA's Summer Savings event. Enjoy 20% off site wide with exclusive savings on our smart home fragrance diffusers. It's the perfect time to upgrade. Visit pura.com
- [27:51]
Liberty Mutual Announcer
and Doug there's nowhere I wouldn't go to help someone customize and save on car insurance with Liberty Mutual. Even if it means sitting front row at a comedy show.
- [28:00]
Greg Kilstrom
Hey everyone, check out this guy and his bird. What is this your first date?
- [28:04]
Liberty Mutual Announcer
Oh, no. We help people customize. And save on car insurance with Liberty Mutual together. We're married. Me to a human, him to a bird.
- [28:11]
Karen Wood
Yeah, the bird looks out of your league.
- [28:12]
Liberty Mutual Announcer
Anyways, get a quote@libertymutual.com or with your local agent.
- [28:16]
Greg Kilstrom
Liberty, Liberty, Liberty, Liberty.