2024 Top Trends and Predictions in Customer Engagement

Kailey Raymond: We all heard it. Data is the new oil, but according to one of today's guests, Jacqueline Woods, data is the new water. Especially in this age of Generative AI. Data is an essential resource that must be clean, usable, and trustworthy if you wanna gain valuable insights and drive excellent customer experiences. In today's episode, I'm joined by Jacqueline Woods and David Chan for a panel discussion on top CX trends, AI predictions for the year ahead and omnichannel real-time personalization.

Kailey Raymond: Hello everybody and welcome. I am Kailey Raymond and I lead campaigns marketing and ABM here at Twilio Segment. I'm really excited to be joined by Jacqueline Woods, the CMO of Teradata, and David Chan, managing director at Deloitte Digital for a little fireside chat today. Jacqueline is an executive with over 30 years of experience leading marketing efforts at Fortune 100 companies, including IBM, GE, Oracle, and Verizon. And David has spent his career partnering with clients to digitally transform their organizations by enabling key CX capabilities to creatively solve complex business problems. Welcome Jacqueline and David. Thank you so much for being here.

David Chan: Thank you for having us.

Jacqueline Woods: Thank you for having us.

Kailey Raymond: I would love for y'all just to start off by introducing yourselves, your role and how it impacts the customer experience and journey. Jacqueline, why don't we start with you?

Jacqueline Woods: Sure. I am the chief marketing officer at Teradata, which means that I lead really all aspects of marketing, which includes even our MarTech stack and our digital, everything that we do digitally as well as comms. For me personally, I started my career in finance and so when I think about the journey to becoming a CMO really is kind of rooted foundationally in what I would argue would be numbers and analytics. And so today I'm more focused on not just storytelling, but really focused on how do we bring what people are looking to do to life through stories. And as you can imagine, everyone is talking about AI and analytics. It is today's Zygox or what is the lexicon in messages of today. It is what our technology does. So I'm really having a great time of really sharing those stories about what we do in that area.

Kailey Raymond: I love that. A storyteller amongst us. David, tell us about yourself.

David Chan: I was gonna say Jacqueline's background probably gives a lot of people in finance like a great opportunity to take on a CMO role at some point. That's pretty awesome.

Kailey Raymond: That's so good.

David Chan: My current role is I am a practice lead within Deloitte digital focused on customer data platform, CDPs. And it's really about how do you create more hyper-personalized experiences, engagement, leveraging that intersection of the robust amount of data signals that now exist. How do you use insights and analytics to then drive and deliver experiences using that connected MarTech ecosystem? But what's interesting is, when I first started out my career, I actually started in industry and then moved into a consulting role. And through that it was really about just creating experiences. Before omnichannel was just cross-channel and cross-channel just meant you had a brand experience sitting somewhere in the channel, it wasn't even consistent. Now omnichannel is actually combining and making it consistent. And back then it was just a mobile optimized version of the site before even a native app. And during that time I don't think I really thought about the measurement side, it was just building the experience. But now to sort of Jacqueline's point around the data, the data in itself is good, but then how do you measure it and understand what the data's telling you in order to drive your businesses forward and make good decisions. I think that's really the next step.

Kailey Raymond: That's great. We have two folks that are data experts here, so very excited to learn from you. So to the topic at hand, we're talking about the realm of customer engagement. We know it's perpetually evolving and it's really hard to stay abreast of all these constant developments. So this is why we have these two experts here in the room. They're gonna help us decode today's top CX trends and share the predictions that they have for the year ahead. So Jacqueline and David have already mentioned the 900 pound gorilla in every single room right now, which is AI. To kick us off, I'm gonna take some trends from our recent growth report that we released, which was saying that half of businesses that we surveyed were expecting to spend more time and money. I would argue it's probably more than half at this point on AI driven campaigns in the next 12 months. That's no surprise. But David, I'm wondering from your perspective, and I know Deloitte's been deepening your investments in AI, and I'm wondering what that interest and demand has been like from your clients and in particular, are there any use cases that folks are looking to solve for?

David Chan: Yeah, I think AI is a very sort of broad umbrella and in general, if you saw what happened in the last year, there's huge interest. There's a huge demand and interest around GenAI as well as AI. And I think the biggest challenge with most companies is trying to understand where to even start. Do I buy something off the shelf? Do I build my own? You look at a lot of companies today, they're all, especially the agencies, they're all building and releasing their own AI initiatives with some sort of custom built core AI platform that they're running all their marketing plays off of. And then so you have these sort of now potentially black box solutions, but that they're supposed to be the secret sauce of what they're gonna provide to you to differentiate, the same time you have companies who also understand that they want to own that capability as well. And so there's this balance of, what do I buy that's differentiated, I cannot do myself, but also things that I need to get support from and build like custom, to give me this custom share of wallet with my customers.

David Chan: Now, based off of all the different options and possibilities a client can go towards, I think they struggle with knowing where to start. Like I mentioned and some, what I've seen is a lot of our clients either go with, "Hey, we're going to do some... Run some few... A few pilots. Yep. Low risk, low investment". Or they say... They take more of a gated waterfall approach of saying, "No, I'm gonna invest in a very, very comprehensive holistic AI strategy before I take the next step because this is so critical, and I accept the fact that I might not do anything for a year". I don't think there's a right or wrong way. I mean I'd love to get Jacqueline's thoughts on what she thinks, but I think to each their own, right, some just have higher appetite and desire to go faster than others and it's not so black and white that there's gonna be laggards and leaders. I think there's also these people who, this is where I would play. I'd play in the fast follower copycat, see what's working and adopt that, versus being on either end of the spectrum.

Kailey Raymond: I like that. That's really smart. Which is gonna lead into the question that I wanna ask you, Jacqueline, because we all know, I mean, what David's talking about is relying on this really solid foundation of data. And so it makes sense to me that you're saying that some people wanna wait a year because maybe they need to make sure that their data is in a standardized position to be able to actually get the right AI outcomes. I don't know. In our reporting, we found that 71% of people were saying that they felt like AI would be more useful if they had access to higher quality data. And Jacqueline, I know you had something similar come out with Forbes recently saying that only 24% of companies feel like they have the right data to get started with AI and make those informed decisions. So how and where should enterprises invest in AI for that maximum impact?

Jacqueline Woods: Well, to David's point, and I wanna pick up on what he said, is there the right way to do this? I think we're in unchartered territory and I think everyone knows that. But I do think that there certainly are things that we can learn from the past with other technologies that were launched. Everyone knows that the iPod certainly was not the first mp3 player. And so there was an appetite to have a thousand songs in your pocket so to speak. But the ability to do that and the platform to do it on actually didn't come out when the initial mp3 players came out. And we probably can't even remember the names of those companies that have those initial mp3 players, but everyone knows what an iPod is. And so to David's point of fast followership often not a bad idea for things that you don't have a solid idea and a plan on how to do it.

Jacqueline Woods: What I would say with data and the way that we think about data is data is everything, it is the holy grail. When you think about how much data there is in the world and how much data is being produced every minute, every second, and the number of zettabytes that everyone talks about like in the next year or two, and the replication of all that, most of it is replicated, so you just keep having things kind of replicate. It doesn't mean that everything is original data. And all of that means that you actually don't have clean and pristine data. And I often talk about data is not like oil, to me it's more like water, because when you think about the earth, the earth is really about 70% water, but only about less than five tenths of I think 1% of it is actually usable. So you have a lot of water that's not usable. You have a lot of things in data today that aren't usable.

Jacqueline Woods: Now, in order for AI to be really impactful in your organization, it has to start with data. And do you have clean data? Is that data pristine? Do you know the lineage of the data? Because AI is nothing if it doesn't have clean data to essentially build intelligence off of, particularly when you talk about generative AI, when you talk about machine learning, it is learning from patterns that are already in the data. So if you start with something not good and unclean, what's going to come out on the other end is also gonna not be clean. And so when you think about how can an organization really use that data, you have to start with, is my data clean? Do I know the lineage? Do I have all of the things that I need in order to make sure that that happens? Obviously those are things that we do at Teradata, but it's less about Teradata and kind of what we do. These are things that everyone who is thinking about artificial intelligence, that is where you must start. And if you don't start there, I think you're setting yourself up to not be successful.

Jacqueline Woods: And I do think that there is a fair amount of uncertainty as we enter this inflection point that clearly we are in, but I think there's also a tremendous amount of opportunity as well. So for me, it's something that excites me, but we all have to be responsible with how we're gonna be using this technology for the good. I think of everyone.

Kailey Raymond: Jacqueline, I love this idea of data being like water. I haven't heard that one. I think it's a really fresh take.

David Chan: Yeah.

Jacqueline Woods: Just give me credit because it is my original thought. It really is like water when you think about it, there is a lot of it, and honestly, it's not usable. So I have thought about this a lot, just conceptually. And also as you know, I always say I do love telling stories and I like to think of things that kind of take different concepts and just make them really very visual in people's minds. And I have found that that one has resonated really well.

Kailey Raymond: It stuck with me. I like it. We'll, TM you every time we say it, so don't worry.

Jacqueline Woods: Absolutely.

Kailey Raymond: Your stamp is gonna be on there. I wanna get into the kind of predictions for AI. I feel like you two are really kind of in the know and have your pulse on this. And one of the things that I was reading recently, and I love your take on it as well, is that Gartner came out with a new prediction that was saying that they were thinking that organic search traffic will decrease by 50% by 2028 as consumers start to adopt GenAI search, which I think is interesting. I'm not quite sure if I'm on the believer side of that or not, but this one's for both of you. Any predictions that you have for AI and how it's going to enhance customer engagement in the future?

David Chan: So put me in the camp of selling and not buying the fact that GenAI driven search is going to sort of take over the world. This was actually a very early one I think when GenAI first came out, someone proposed this and you see a lot of people writing articles talking about how well GenAI can spit out the information from web scraping, then there's no reason for them to actually click through right. Now as someone who's at a very young age been told they're really good at searching the internet and had people very young in my career say, "David, how did you find this out? You're so good at it". And I get, I think I screw up thinking that I'm a good internet search person. If you're searching for something very non-critical, non-mission critical, like what time does a store close? Yes, you might accept like the first search result that you even see today. There's some GenAI based search results that are coming up, but anyone who knows when they're trying to do real proper research, they probably click through at least like 10 different links. Some of them are just copy and paste and paraphrase version of it, probably just to drive some revenue to just land on it by leveraging some of the person's content.

David Chan: So that content isn't always accurate. And so you kind of have to do your due diligence. You have to click through link. I don't think anyone wants to click on the second page of a search results link, but my point is, when you do that you get a sense of, "Okay, this authoritative place said this about this thing". And now you're trying to triangulate on it. With a GenAI search result, it just summarizes it. It doesn't cite references, it doesn't do anything. And even if it did the amalgamation of it, you would still put into question. So unless searches fundamentally changes in the next four years and GenAI based search results reflect that transformation, that GenAI search on its own in its current state, it's not gonna drive 50% to GenAI search results. So that's sort of my stake.

Kailey Raymond: I think the content is the thing that you have to worry about more with GenAI is like how many aggregated GenAI articles are gonna be created by all these forms that are then gonna be served to the top of your search. That's a more realistic thing that I might see happening.

Jacqueline Woods: I tend to agree with David. First of all, even when you think about AI, machine learning, and anything in that realm, it's all gonna boil down to critical thinking. And I think the skills of critical thinking are gonna be even more important than they were two years ago, four years ago, or 10 years ago, because you will be required to really think through in a really methodical way where it not only where the information is coming from, but how is it relevant and contextual to what you're trying to do? I was talking to a friend the other day, because you can't see it now, but to this side of me, I have these encyclopedias and we were making a joke about it because obviously no one has encyclopedias anymore, but we've had them in this office for obviously more than 20 years. And I started thinking about when I was a kid, how I looked up things in an encyclopedia. So to David's point, having to do research, having to go in and look at all the things that you wanted on a particular topic. And what we were talking about was with Google and search, that imagine, I can't imagine looking up turtles, something that I was doing when I was in fourth or fifth grade when we were talking about reptiles and all of the things that are at my fingertips today versus what was just in that one encyclopedia when you looked up reptiles.

Jacqueline Woods: That said, the one thing that the encyclopedia has is the source and citation for all of that data and where it's coming from. It's just not copied from something to something else where you actually can't trace back to what happened. I was recognized in Fortune like in 2004 or something like that, and I have seen the citation of that article in reference to another magazine that actually was not the magazine that it was in. And somehow that is on the internet right now. And I know that it is incorrect information. And I don't know how it happened at some point someone was lazy and they kind of wrote something or an article or something that then was someplace else. But now that shows up. And I believe that generative AI can't filter through those mistakes 'cause it doesn't know that it's a mistake. I know that it's a mistake. And unless there's something that kind of rationalizes all of this, I think you would have more bias than the bias that already exists in some of this information today if we were only reliant on generative AI. And I like the way that David described this kind of, it's good for kind of general, like right now, if you use ChatGPT on OpenAI, it gives you some general things. It also can prompt you with other ideas to do other research. And I think those things are helpful. I do not think it's a replacement.

David Chan: I actually love the encyclopedia reference. I still have a copy from like, I don't know, like 30 years ago. But you know what it made me think, Jacqueline, you're like my GenAI. You're prompting me to think of something else. There are so many citations and sources of research that actually conflict with each other.

Jacqueline Woods: Sometimes, yeah.

David Chan: So to your point is who is reconciling that? What's that Oracle that's actually sitting in the center saying, this is right, this is wrong. I mean, we try to do that with Wikipedia, but you know how that turned out. So I do sometimes wonder, like how would GenAI rationalize and reconcile those conflicting arguments? What would it actually spit out as an answer to people's problems?

Jacqueline Woods: And who is in charge of that. Like there is no great AI GenAI rationalizer in the sky. That doesn't exist. And to David's point, there's not this great Oracle that says we will make sure that all this data has the appropriate citations. And when there are conflicts, this is how we're gonna resolve them. Who is the decider of the conflict and who is the decider of what the information is that needs to be rationalized? And then what gets prioritized because you can't do everything at one time. The thing that's always discrete and always absolute is mathematics. Now things that are formulaic and things like that, those things are properties and principles and they exist in the world and they're not arguable. But anytime you have something where there is what I would call other judgment and you're kind of feeding something to a machine to kind of look at patterns and come up with something that's a little bit judgmental, you have to have a way to break through that and to resolve it in a way that is in the best interest of the humans that are using it. And I think that's really critical and must be the first step when we think about AI. The impact to humans and the impact to us not just in 2024, but a 100 years from now, 2100. And so we can't just abdicate our responsibility to think that these things are gonna take care of themselves.

Kailey Raymond: Fully hear you. And we're gonna touch on this a little bit later. I think when we're talking about trust, which, who's the arbiter of truth here? Right now, I think it's probably newly appointed tech god, Sam Altman, but we'll see where that ends up at the end of the day. I wanted to see before we end our discussion on AI for now, any other predictions related to CX and how AI is gonna play a part of that in the future?

Jacqueline Woods: I think that personalization, it is the thing that marketers want to do best. It's the thing that we wake up every day and want to do better. It's the thing that we know drives stickiness and engagement. And to the extent that we can leverage these learnings to do that, I think it is we need to be all in. I think it makes our companies better companies. I think what we deliver to end users, whether that be a company or whether that be a person, is gonna be really important. And those platforms and technologies that enable us to do that, I am all in 100%.

Kailey Raymond: I love that. David?

David Chan: I think for going into 2024, obviously, Deloitte actually produces some research studies. Everyone's doing it around AI and GenAI. And one of the things that we're seeing is there is adoption of someone simpler. The use cases to pull off right now with like copywriting, image generation, content production, all rounded up into content production. But the other wrinkle is what channel are these content pieces being produced? And what's the risk, brand risk associated with getting it the fidelity level, perfect versus not. And so obviously, own channels like email, direct mail, SMS, what have you, higher level of quality of content required. But I think you're gonna see that going to this year, where you're thinking about paid media channels, where maybe there's already acceptance of like a lot of DCO solutions, the dynamic construction of some of the final outputs isn't even... Isn't that great to begin with. GenAI is gonna add more quality to what's there already, in my opinion. And that's where the adoption is gonna grow in for this coming year.

Kailey Raymond: I am on the same page with you. And to Jacqueline's point around personalization, I think that's a perfect segue kind of into our next trend, which I wanted to touch on, which is omnichannel real-time personalization, which is something that I think has been a driving force for a really long time. But with technology, we're getting even closer to being able to actually achieve this, being able to kind of create all of those personalized images and content pieces for every account or individual. GenAI is actually making that possible. We're also seeing, to your point, a change in kind of behavior as it relates to this proliferation of digital channels. The number of touch points has just only been increasing. Our reporting is saying that it's tripled in the last 15 years from two to six. Frankly, that seems low to me. I feel like it's probably even higher than six, especially for different industries. We also know everything kind of relates back to timing when we're talking about customer engagement. So the importance of real time, especially in certain industries, might be more important than others. So, David, do you have any insights or examples you might want to share for how your clients are approaching that need for omnichannel, real-time, personalized communications?

David Chan: Yeah, I would say a couple of years ago, coming up with an omnichannel strategy was pretty much a thing everyone asked for. Now that we've sort of accelerated our adoption of digital, I feel like I see less of, hey, can you help me out with an omnichannel strategy? Because I think most companies have already either spent the time to create one, and they're actively executing it, or they know enough or have the maturity to actually not need a partner like Deloitte to go and help them with their omnichannel strategy. Which, like I said, has changed to before just having a brand flag in the channels versus now creating that omnichannel experience that's consistent, that's connected across the journey. So I treat that as table stakes. Now, the real-time one is very interesting because in my world, everyone wants real-time personalization. And what that means is the data has to be real-time collected, data has to be a real-time process, data has to be real-time curated to be made of some sort of business sense to then activate it on in real-time. But what I've also seen is that most companies just don't need it. When we actually walk through their use cases, none of that or only parts of that have to be real-time. And if only parts are real-time, it's not real-time. It's sort of a batch-based delayed response, which is acceptable.

David Chan: And to me, what matters more is less about whether it's real-time 'cause just faster is not always better. It's about how contextually relevant the message is being returned to the customer from the brand. That is more meaningful. And so for me, my advice to a lot of companies out there that are so focused on real-time, you have to understand there's a penalty to pay for not only implementing the systems and technologies to do it, but also the support and the licenses and the operational costs when in reality, you probably don't need it if you really ask yourself those hard questions and talk through it with your teams. And so really think about whether you need real-time or not and invest in those areas where it's required and don't in those others.

Kailey Raymond: That's great. And it kind of touches a little bit on personalization when you're talking about the context is key. It's kind of like riding the line between personalized and creepy if you're just like immediately sending an email to somebody based off an interaction that they had, depending on, I think the industry is really dependent on this as well. Like B2B probably doesn't need to be as real-time as maybe retail. If somebody buys your shoes, maybe you want to suppress ads for that shoe from that person so you can save that dollar immediately. So I think there's some interesting use cases there. David, I'm wondering, do you have any thoughts on what might be next in this realm of real-time personalization?

David Chan: Yeah. So I mean I think the logical extension Kailey of how do you create that, personalized, contextually relevant experience is that most people think, all right, I've got something that caught the signal. We've now had the information to build that audience. They're part of it. Okay, I'm going to now just deliver the experience. But the missing piece is the content. And not to harp back on the content, I always joke to people who go, yeah, we want to do one-to-one people-based marketing. All right, you heard that. And I said, okay, that's cool.

David Chan: But do you have a piece of content, one-to-one, do you have that content supply chain set up to scale and deliver a piece of content to every person? And I'm talking about companies who have millions of people. Now multiply that by your 2 to 6X expansion of channels, which has a slightly different rendition and format. So how are you going to achieve that? Which is why I think from a content perspective, I know Jacqueline earlier said, hey, is data the new fuel? No, I think about it as water. Well, if you roll that back and still think about data as a fuel, content is almost like an alternative fuel or another fuel that doesn't allow you to deliver great CX unless you have that as well, another element, another dimension. And that's where I think that some of the studies that Deloitte has had is that there's actually 54% increase in the amount of content that's required by most marketers today compared to the previous year.

David Chan: And what's that driven? I think it's all the stuff that we talked about. It's the explosion of channels. It's the desire to be more personable, more relevant. And so because of that, I think that GenAI will probably play a pretty big role in that space because it can pull off that use case. It's much closer to pulling off that use case versus all the other interesting and exciting and emerging use cases that are out there, which they will at some point, but these are very ready and baked in my opinion.

Kailey Raymond: I love that.

Jacqueline Woods: Yeah, I do think if you just kind of expanding on your example, I do think if you go to Kailey's, oh, there's six channels minimum, maybe there's eight, but minimally six, and you do your, "one piece of content that's email. And then you say, how does this get served up in LinkedIn, TikTok, Instagram, how does it get served up in text or blah, blah, blah, like all the places that it could show up. I think that GenAI would be a great place to start to be able to do that because it knows what the character requirements or limitations thereof for Twitter. It knows how the content differs between Instagram and TikTok. It knows when LinkedIn has to look like this in order for people to engage and to capture attention. And I think that is an opportune place, honestly, for AI to help because those are things that kind of, whether it's format or trends or other things, you could take one thing and then say, this needs to be in all these places. Give me an example of what that could look like. And I think that that would be awesome. And What that does, it doesn't take anything away from the person that's developing or creating original content. It just helps them quickly propagate that content across all the channels.

Kailey Raymond: That's really smart. And it's a really nice transition to the last topic and trend that I want to speak to you all about, which is trust and privacy. So we've been talking about some of these one-to-one people based, you need all the content, 54% increase in content By the way, David, it's just astounding and feels like it'll resonate with my team as well. So I'll bring that back and I'm sure it'll make them feel very heard. So Thank you for that. But consumers are less comfortable with using their personal data for personalization purposes. We're seeing that in our state of personalization report. They're not quite sure how businesses are gonna leverage that. So Jacqueline, I'm wondering if you can share your thoughts on how to best address some of these privacy concerns coming from consumers while still reaping those benefits of personalization that we've been talking about.

Jacqueline Woods: Two things. I think people opt in to the personalization, I don't want to call it platforms, but things that they actually care about that they feel that they get value from. So in other words, what we see is that from a personalization perspective, if people believe that the company, product, or service that they are opting in to provide personal information is creating value for them, they are willing to give pretty much all the information that is required in order for them to optimize that experience. And so what I think companies really need to do is understand what do they need the data for and how they're gonna use it and how they're gonna use it primarily to create value, not for themselves, not monetizing the thing themselves, but how does that information help them deliver a better experience and create value for that end user. When that happens, people actually do want to opt in.

Jacqueline Woods: So imagine if it was Weight Watchers as an example. I mean, you're kind of giving everything, your age, your height, your weight, even your ethnicity, because those can be factors for other things. Do you have hypertension or not? Are you diabetic or not? You're kind of giving a lot of information to kind of an entity that doesn't necessarily know you. However, the belief that you are gonna be on a journey to better health is worth it to you. And therefore, you know that if you do not give all the information, you're not gonna get the optimal output from whatever the platform is. And those are the times that I think personalization can be really helpful. So even if you translate what I just said to a company and someone said, I need to have these kinds of insights for this reason to help me create more value for my own company, then you are going to be willing to kind of figure out what data that you need across your enterprise to make that happen. And if everyone understands the value that they're gonna get out of that on the other side, then they're gonna be more willing to share that data.

Jacqueline Woods: If there's no value creation, then it's not worth sharing the data. You don't need to know everything about me. You don't need to know, I was born in December just in case Kailey, you wanted to send me a present or a Christmas card. You don't need to know that. But if there's something valuable to it where we're exchanging value, then I am more willing to share that information that's across the board on B2B or B2C.

Kailey Raymond: The Weight Watchers example is so on the money. I love that example of value exchange. It makes perfect sense to me. And I wanna talk about the other side of this a little bit, which is, well, it kind of touches both, but like acquisition, I think related to trust and privacy, right now, we know the world is going cookieless. It's on the top of every marketer's minds. It's related to how consumers are calling for more control over their customer data and they're placing top priority on identity data. So David, I'm wondering how are you advising companies as they're managing this cookie transition this year? And even more broadly how they're managing consented data. Do you have a biggest piece of advice for folks?

David Chan: I would say that a lot of people within the market, or what I see is they immediately gravitate toward first party data solutions. As alternatives. Well I need a more durable identifier, I need to get right with my first party data. And so they go and try to build data lakes, data warehouses, customer data platforms and the such. And so that's sort of like where they immediately gravitate towards. What I tell people though, you have to understand that when third party cookies first was initiated, like within sort of main and existence, it was a little bit of the wild wild west. I mean, people were using third party cookies almost in a way in which it wasn't intended. And what they did was they did things that consumers didn't realize that hey, you could track a user cross-site using this third party cookie that followed you around.

David Chan: So, just so it's clear, third party cookies are being deprecated. First party cookies, like the one you go on your banking website and it pre-fills your sort of username and whatnot, or a token that's still sticking around. And I think that when you wanna first start understanding how you want to go about third party cookies, there's sort of like three types of people in this world. There are people within organizations who think, you know what? I am ready. I've been planning for this because they are a planner, they're getting ready and we'll see what happens. And I think because Google has started to... Made the changes to Chrome and in sort of like a slow rollout fashion, we'll see, they'll realize whether what they had... All this planning had done was right or wrong. Then you have the group who says, I'm not ready, but I don't care because I'm not really convinced that this is going to have impact that I think it will on my business.

David Chan: They're not getting asked by their CFO or their CEO on this topic. And they're like we'll see what happens. And then the last bucket are just people who have no clue. [laughter] they've been living under a shell or rock. They have no idea that third party cookies or what their impact is. And so my recommendation is there are actual, a lot of companies who have come up with some sort of way to help you measure the impact of cookie loss and signal loss on your business as a whole. And they vary in degree. So like Deloitte, we offer this cookie calculator, which is almost like one of those simple surveys which you can just kind of fill out and say, okay, if I spend this much media, if I run these number of campaigns, this is the potential impact to my business if I do not address.

David Chan: Other companies will offer services where they'll actually do a deep assessment. Or a combination of both. Either way, I think it's first measuring how much does your business actually rely on third party cookies? And then figure out from there what's the right level of investment into some set or suite of first party data services that can combat it. And it won't be just that DMPs are going away, but you're gonna have to leverage first party data platforms. You're gonna have to leverage data clean rooms. You'll also have to leverage third party services to support the gap that will exist once third party cookies are fully deprecated.

David Chan: It's a challenging world to navigate in 2025, honestly, Google's been telling us about this for how long though David, I feel like companies have had plenty of time to prepare and now people are maybe finally waking up and they're in that category of, oh, I need to take action. [laughter]

[music]

Speaker 4: This podcast is brought to you by Twilio Segment. According to Twilio Segment's state of personalization report, 69% of businesses are increasing their investment in personalization. And with the world going cookieless, first party data is the most reliable, trustworthy, and compliant way to create tailored experiences. Imagine suppressing consumers that just purchased that pair of shoes from your ads and decreasing acquisition costs by 5X or sending the perfect text promoting your other beauty lines and converting 50% of customers. That's the power of Twilio Segment. Learn what's possible @Segment.com.

Kailey Raymond: I have two more questions for us and we'll round it out today. Are there any predictions or trends we haven't discussed today? So, so far, the three that we've talked about, AI, omnichannel in real-time and trust and privacy as some of the biggest things we're talking about in CX any other ones that you both wanted to throw into the room that you were thinking about?

Jacqueline Woods: I do wanna pile on with, we all know that AI it has the potential and likelihood to change everything about how we think, about how we do work and get work done and how we interact. What I wanna throw in is that the importance of trust in the people. So when people say responsible AI in some ways I cringe because I do not think technology can be responsible. Responsibility or something being responsible is generally an attribute that you put on people and humans and humans can be responsible because we have a consciousness that makes us believe that we should do something and that we have ownership to be accountable and have accountability. And so I think that the trend will be really doubling down on making people responsible as we move forward in this world of AI. And that what you will be able to build trust in is either the data that you've cleaned and that you can serve up.

Jacqueline Woods: And certainly those are things that Teradata does. Not to put it in a plug for Teradata, but I noticed that David was putting in plugs for Deloitte. So I feel guilty.

[overlapping conversation]

Jacqueline Woods: I feel guilty that I haven't plugged as much as maybe I should, but I feel very strongly that what we'll need to really do is be able to trust data and trust the output and that we need to have transparency and that the responsibility squarely falls on us as humans to each other. And I think that will be a trend that comes out and the approach around the ethics around AI will become extraordinarily important over the next year and in the years to come.

Kailey Raymond: Beautifully said. David?

David Chan: Jacqueline being my GenAI and promptering.

[laughter]

David Chan: Spinning my wheels in creativity. I think what you said and going all the way back up to our initial conversation where we started talking about good data and from search like the Oracle of like governance around data. I think what's gonna happen in this coming year is this idea that people are going to realize that the GenAI solutions that are based on public data sets. Those LLM based on public data sets will probably realize that there's not a good solve. I mean there's this concept of AI cannibalism, one of my coworkers told me that it's called AI eating itself, where the LLM is starting to consume data that was generated from an AI thinking that it's truth and it's creating this like circular loop where it's almost collapsing on itself. The point is because of that, there might be a pivot to trying to understand, okay, the public LLM based solutions might be too hard to pull off.

David Chan: Maybe we gotta focus on the private ones where there's governance around the data and maybe there will be third party data providers who are going to say, well I have the clean data water, [laughter] for you and I will supply you my clean data sets that I've refined not oil, but water, [laughter] purified for your LLMs to drink in a private setting. So that's one that Jacqueline, you made me think about. The second one is, I really think that the last couple of years everyone's been focused on data platforms to try to figure out how do I turn data swamp... Data lakes from data swamps to data lakes and create some order to the chaos of data. Great. But now the next step is, okay, I know how to move data from A to B, but that in itself does not make it usable.

David Chan: The data is still just like taking your analogy, Jacqueline, of the 70%, you've just moved it from one area to the other, but it's not still not usable. And I think to make it usable besides ETL and transformations and all that, I think it's actually identity resolution. It's actually for customers there'll be identity resolutions for non-customer domain, it's gonna be entity resolution. But it's this idea that the data now has to make sense and be tied to make it usable. And so if you don't do that, then your LLMs, your personalization campaigns, your experimentation, your measurement frameworks that you're now trying to unify those reports will mean nothing. And I think this coming year will be the year when a lot of people have invested in these data consolidation projects, but now they're gonna realize that unless they figured out this identity resolution, they might not be able to take it for forward.

Kailey Raymond: Beautiful. I love that one. And to your first one, I feel like the media industry, the publishers just got a whole new business model. Like imagine New York Times sells OpenAI because they can actually trust the information that's coming from... A lot of possibilities for the next year.

David Chan: Yeah, I'm pretty sure that wasn't an original idea for me, but thank you. Or you're welcome rather. [laughter]

Jacqueline Woods: And I...

Kailey Raymond: You're welcome.

Jacqueline Woods: And I actually think that is coming up on an interview that I think I had read about or seen with Satya Nadella from Microsoft. Because I do think to your point, some of the LLMs were kind of bringing in data from the New York Times, but I think they had not paid for it.

Kailey Raymond: Not paying them.

Jacqueline Woods: And I do think that's wrong because I do think that that is IP that someone owns and the owner is the New York Times and the people that wrote those articles. So it's not like this is just in the public square and it's free. And so that, I don't agree with that. That should be free. I think that may be the position of some folks in technology. I don't think so. I think the reason when you have research and researchers and that's expensive.

Jacqueline Woods: It's expensive to do what journalists do, it's not free. They work really hard and I don't think that people should just be able to take that information and learn from it because it is using something that otherwise would've had to be paid for if the "internet" did not exist. [laughter] You would have to go buy a paper, [laughter], you would have to read it and a bunch of people would have to do that and then they would all have to synthesize it. Now, whatever the cost of that should be, I don't know. I just know that in order... That if you're taking information that someone has essentially research, developed and curated for a purpose to put something together that we all learn from, I don't think that we can just take it. I think we as people who respect the work of others should expect to have to pay for that.

Kailey Raymond: I think it's just a sort of fun thought where we originally might have thought that GenAI would kind of eat journalism as an industry, but in fact maybe it's actually paying their bills. It's like...

Jacqueline Woods: Correct.

Kailey Raymond: It's a really interesting flip in the story...

[overlapping conversation]

Jacqueline Woods: It's a flip in the story, it's a flip in the story. Yeah. You gotta get all that information from somewhere and you can't and...

Kailey Raymond: Exactly.

Jacqueline Woods: And You can't make it up. And David already talked about the hallucinations that occur when AI's getting information from AI and then pretending that it's AI. [laughter]

Kailey Raymond: Totally.

Jacqueline Woods: And generating more generative AI. [laughter]

Kailey Raymond: Okay, you all have been extremely generous with your time. I have one last question for you, which is what steps or recommendations would you both have for somebody to stay ahead of the curve as it relates to customer engagement?

Jacqueline Woods: I think you have to start small and then scale. Like I think you need to start with something that you can develop a use case for, pilot that, if it's successful, scale it out. If it's not successful, make sure that you obviously tweak and course correct until you get it to where you want it to be. And then scale. I think that's gonna be important to David's point at the top of this call, you just can't just go all in without a plan in a prescriptive way to actually go down this path. This is a marathon, this will not be a sprint. And I think those people who train well and do all the things to be well trained, I think they will be the winners of this race.

Kailey Raymond: Train well everybody. David.

David Chan: Yeah, I like what Jacqueline said. I think it's a combination of when you do your pilots and your POCs, it's to prove a point. But once the point is proven, [laughter] then you have to come up with a plan. That's actually holistic and maybe waterfall. I like that very much so Thank you Jacqueline again for being my, [laughter] GenAIcon.

[laughter]

David Chan: For this conversation.

Kailey Raymond: This is a match made in heaven. There's a lot of ideation that's happening today. This is great. Well, Jacqueline and David, thank you so much for being here. We so appreciate you sharing all of your insights with us. Thanks again.

Jacqueline Woods: Thank you.

David Chan: Thank you.