PegaWorld iNspire 2023: Panel - The Future of MadTech is Cookie-free

The clock is ticking, but there is still time for advertisers to future-proof themselves against the gaps and budget shortfalls created by data deprecation, most notably the demise of third-party cookies. But, instead of seeking solutions from multiple external providers to augment and manage data, the most effective and efficient way is to use AI and machine learning to activate first-party data across functional areas and channels.

To do this, the customer care center and the marketing department need to connect and access data signals from every functional area to obtain a full view of the customer in the moment. Cookies enable short-term sales. First-party data optimization, powered by AI, enables relationship building and long-term sustainable growth.

In this session, find out from our panel of experts how AI-based tools enable marketers to constantly analyze, aggregate, and act on data signals in real time, creating an interactive dialogue with the customer based on their needs in the moment. The result is more dynamic, empathetic consumer engagements, better brand-customer relationships, and lasting brand value.

Transcript:

- Hello everyone. Welcome, we are here. Let's see if my little clicker thing works. It doesn't, but that's okay 'cause we don't need those slides. I'm your host, Tara DeZao. I'm a product marketing director here at PEGA for CDH, Customer Decision Hub. And I have been tracking very closely the trend that we're here to talk about today, which is the deprecation of third party cookies. I'm a long time ad tech nerd and friend and I'm joined by four outstanding industry experts here today. So I'll introduce them, I have Liz Sleyffers here. She is a senior solutions consultant at PEGA. She specializes in one-to-one customer engagement and she's an expert in digitization and personalization. Liz, welcome.

- Thank you so much.

- Next we've got Paul McVicar from British Telecom. He leads a team of data scientists, analysts and decisioning experts who ensure that BT can provide a more meaningful and enjoyable customer experience. Paul.

- Great to be here.

- Thanks for coming. Neil, also from BT, Neil Hodgetts is leading a transformation to make BT a more data-driven, personalized and agile organization. And he's responsible for the customer base of 30 million customers across all of their channels. Welcome Neil.

- Thank you Tara.

- Then I have Jacqueline Leng representing DFN Strategy. She is the SVP of Global Client Solutions at Kinesso with IPG. And she helps her existing clients focus on their data strategies for ad tech and MarTech. And she has also held many roles across the ad tech ecosystem. We've worked together at several companies. She's got an agency background. She's amazing, welcome.

- Thank you.

- So before we get started, just a little bit of housekeeping since this might be the last time you can ever engage with cookies. There are some cookies in the back for when you're exiting the room so that you can have one last hurrah with the cookies. Now let's jump into why we have this brilliant group here today. So in 2021, Google announced that they're gonna phase out third party cookies in their Chrome browser, which captures 60% of internet traffic. It's the way that we've all sort of connected consumer touch points, especially when we don't have enough first party data. This was widely expected, but it sent shockwaves through the ad tech community, although it's the last thing in sort of a long line of deprecations and regulatory environments that are making it impossible to move forward with cookies. So let's get right into it. Liz, what are some of the challenges that you see with organizations who are trying to connect the full customer experience, the full customer journey with just third party cookies?

- Well, I think the main thing, it's basically two things that I can sum up. 'Cause I'm sure there are so many other challenges that the organizations are obviously going through because we've been used to working with third party cookies of course 'cause it was available to us. But with just third party cookies alone, it's just impossible to create that omnichannel experience because you cannot really track all of the information of your customers and their whereabouts across multiple channels because you're tracking in channel. And it's obviously all done in silos, you cannot really connect it. So that is a big problem. Of course, moving forward, if you want to achieve more personalization, that is a huge roadblock, obviously. And second to that, when you're actually tracking in that manner, like in a channel and just on third party cookies alone, you have such a big chance of having inaccurate data. Because also, of course, like a simple example is always if someone is using same computer in a household.

- Yeah.

- That can be multiple people of course, but you're just tracking it as one, so highly likelihood for inaccuracy. And also of course when it's tracked with third party data, it's collected and reviewed at a later stage in time, which means that there's already a lag of what that data is going to tell you. And that usually we would use that data to then take action upon. But that would take days, if not weeks of course, 'cause we would look back at a week's worth of data or a month's worth of data and make our tweaks. So that is already quite a long time. Of course, that might change the context even in just a few hours even.

- Absolutely.

- So those things combined I would say is the two crucial pieces of course, of not being able to have a 360 view and create that consistent customer experience.

- Fair, fair point. Okay, I'm gonna kick this now over to Paul 'cause Liz just discussed some of these issues that we're seeing with third party cookies and brands trying to connect the consumer experience. But we're all marketers, we've known this is coming for a long time. What do you think has like prevented brands from taking the steps to sort of get in front of this moving train?

- Yeah, I think first and foremost, the fact is you have a solution in place and in an organization as large as BT, I mean, the case for change is always the hardest case to make. So it's just been a lot simpler to live with what we've had that's sort of been working for us to a large extent. And no one was really looking at how can we evolve beyond that. It was just, I think the teams that were managing the platform and the teams that were driving the paid media outcomes were pretty comfortable with how things were going. It's really been like a bit of a kick up the backside from the big tech companies to say, you've gotta move on here, you've gotta evolve. And then secondly, as Liz mentioned, I mean we run two brands, that are consumer facing within the BT group. Even within those brands we have siloed X stacks for different product lines, for mobile, for broadband, et cetera, et cetera. So it's only really been over the last two years where we've made a concerted investment and effort to bring all of our first party customer data together into a single locale and a single platform. So I think even if we had looked at doing this a couple of years ago, we probably would've found that the value proposition was somewhat deprecated because we just didn't have all the customer data there to support an alternative solution. But now we've made a lot of inroads there. So I think the foundation there has been well laid. Neil, you've been working on it for probably even longer than I have. You might have a perspective as well from the marketing side.

- Yeah, so I think as, as Paul said, the barriers are pretty big. I mean, you have to make a lot of investment in CapEx on things that aren't that sexy. Data platforms clouds. It's not things that boards traditionally wanna invest a huge amount in. And added to that you've got the cookie journey today that sort of works for, we spend tens of millions of pounds every year on paid media.

- Yeah.

- And there's obviously cases to be made with finance around keeping that budget. It's very easy to measure ROI with the way it's set up today and changing it is a big move for an organization like as Paul says. So that has been some of the barriers. But you know, as Liz says, it does lead to a fractured customer experience. I mean, we send hundreds of customer interaction, outbound interactions every single month. We take in millions of interactions into our inbound channels and what we're pitching to customers through NBA there effectively is very different to what's happening in our paid media channels. So our sort of own channel experience and our paid media experience is very different and less richer because of it. But I think one of the challenges we've had at BT is trying to join up, create that omnichannel experience before we even think about plugging it into paid media. So that's been a big challenge for us and Paul's team have been working quite hard on doing that side of things.

- Yeah, yeah, I think, I mean the way we look at this at a holistic layer is how it ladders into our next best action solution. That's basically the future of our MarTech, the future of our customer engagement is next best action. And I imagine at a conference like PEGAWorld, there's gonna be a lot of discussion around MBA. And different approaches and the way different organizations are tackling it from where I sit, I conceptualize MBA as an architecture ultimately because unless you have all of your channels fully integrated and on board with a single decisioning engine that's able to drive a consistent customer engagement and offer program across all of your channels, then no matter what you do centrally, it's never gonna see the light of day where it really matters. Obviously a key part of a next best action architecture is a decisioning engine. And while PEGA brings that world class decisioning capability to bear, but from a strategic rollout, we looked at it sort of channel by channel. How do we make sure that we get each channel to line up against what we're trying to build from a next best action solution. So for a number of different reasons, we started with our digital platforms, our own digital platform. So our app and our web were the first ones to be fully integrated. We've been working on outbound sort of in parallel because that's been a tougher one to crack organizationally more so than the technology challenges. And then we've also moved into our agents assist for our contact centers and our retail outlets. Paid media to a large extent is the last one to go. And partially that was because it's just harder from an operational perspective. We've talked about some of the silos earlier where the paid media team inside BT, although they're part of marketing, they have sort of run to their own agenda to a large extent and not really integrated in a ways of working with our base management teams that Neil runs. But it was segregation of technology, segregation of data and segregation of business operations. Now we've using the the next best action solution to sort of bring all of that together under one roof. So we've just probably broken ground on paid media in the last quarter.

- Okay.

- That's where we're seeing a lot of the activity over the next year.

- Great, you're not alone. Many brands are facing that siloed paid media situation. Speaking to that, Liz, do you see this across other brands that you work with and what industries, what's it like out there?

- Right, no, I think that of course BT is not alone in this when it comes to the complexity and the ecosystems and the willingness to change might be there of course, but it's not an overnight thing to achieve obviously. And I think it actually makes sense that, well all of the organizations I've obviously been working with have a similar situation, a similar challenge. 'Cause if you just take a step back, it makes sense, right? In our marketing world over the past decade as new communication channels continue to arise, new management tools for those channels came along with it. And they were all just built to work on their own and to have their own brains basically. So it is just been building and building and building. It was bound to happen of course to see these spider webs of ecosystems.

- Yeah.

- And trying to streamline that is a challenge. You literally have to vision the short term and a long term vision and then how do we get a centralized indeed piece in there, a hub if you will, to indeed make sense out of all of that piece, all of the data. That is definitely something we always see of course. And all of those, what you also just mentioned, the pieces of data that are tracked in silo and also stored in silos, obviously that needs to also come together in this hub. So it's a bit of a dual thing that you have to look at the architecture first of which piece is doing what. And secondly of course then where's all that data and how do we make sure that we make sense out of that. But yeah, those are very recognizable challenges and especially the complexity to make that change, it really requires more than just the marketing team to understand, of course, what that will look like and what are the steps to actually get there. So it is very recognizable. I think also over across industries, I think the one, the couple of unique companies I've worked with who had maybe a less of a complexity were the pure players.

- Yeah.

- Like abull.com, which is our Dutch Amazon.

- Okay.

- Because they were obviously born online basically. They never had offline. So they have a different set of challenges. But not to the same extent of having such a complex ecosystem where you have to merge offline with online, yeah.

- Yeah, I mean, MarTech stacks, MadTech stacks, they're so complex. We know this, they're often full of technologies that move around third party data. What's gonna be a better long-term sustainable solution once we can no longer activate these cookies?

- Yeah, well I think Jacqueline will have obviously way more to add to that as well from her background. But from what I can see and already see what's happening, of course, the first party data focus, it's not probably new to anyone, but that is certainly something that will continue to grow because third party data is just never as high quality obviously as first party data will ever be.

- Agree.

- So having that focus on the data that most organizations actually already have. And utilizing that in a better substantial manner is going to be crucial. And leveraging that piece of data without that reliance of third party data, that would also hopefully mean that, and I've seen that happen already, that third party tooling's can be deprecated. So that should hopefully get rid of some of the legacies that is holding down some of these complexities. Of course, so hopefully that is one way next to that though. Just the first party data would also not be enough of course. So the focus on behavioral data is very crucial 'cause making sure that you actually understand what someone is doing on the website in real time is obviously just as crucial to match that up with the information, like first party data that they already have of that person. So marrying those two up is really going to provide that 360 view of course. And making sure that the interactions are actually also stored in the profile. For example, I think we already alluded to it, a centralized hub to make sure that everything is streamlined is going to be crucial because having multiple sets of data everywhere without having one logic centralized hub to make sense out of it all is just going to be impossible to create a consistent experience. Right, because you would, if there's still silos, then you would also provide a siloed approach or customer experience basically. But of course we can't do any of this without AI.

- Yes.

- So the audience will hear many times towards AI, of course, it's only PEGA World. So without AI, obviously you will not be able to scale or do this in real time. And that's another of course crucial piece is that when people are doing something on the website in real time, the ability to understand that, recognize that, and immediately analyze what the best response is. You couldn't do that, as a either of us. So how good a marketeer you are, you need AI to be able to support that and that's gonna make a difference in relevant interactions with your customers.

- Absolutely, okay. Jacqueline, you're out there in the world with all these brands. There's been a lot of solutions floated, a lot to replace cookies, but there are no like for like replacements. Could you talk us through some of those other technologies and sort of maybe the challenges and opportunities of those?

- Sure, certainly, I think last look, there's some sort of list where there's over 100 identity solutions out there. You know, LiveRamp, ID5, UID and the list goes on and on. And coming into this event, we were doing some research and data points and eMarketer has done some research last year around Q2 looking at marketers and surveying them. You know, the stat shows that 70% of the marketers, they're just gonna rely on their activation partners, the DSPs to come up with the cookie solution. And 57% of those marketers are just overwhelmed. There's so many solutions out there what are the differences, what are the nuances? What suits my needs and what are my use cases, right? And 29% of those are just stop. They don't know where to start. And I think that's sort of what we are seeing across the board. There's a little bit to what Paul mentioned earlier, it hasn't happened yet. The cookie deprecation has been postponed a few times. So we're just gonna wait and see and in the meantime we're not ready to do a lot of these things. So I would say opening up to the room who here raise your hand? Has the ultimate cookie free solutions?

- Oh no one, no one shock.

- We're getting there slowly, slowly. But I wanted to share an example of some of these examples of tests and learn that's starting to happen. A lot of this in the pay media space that Paul mentioned, one execution is looking at partnering with your DSP space, like a trade desk, getting that log file information in, put that into environment. And this is where we're starting to see clients with their analytics claim room, working with their agencies to give them access to their instance of this analytics claim room. Bring that DSP log file from Trade Desk on performance clicks, connect that to an anonymized ID so you can attribute to an individual based level. At the same time, the analytics claim room has client's first party data from CRM, for example, anonymized also connected it to a pseudo anonymized ID. This is where you can get into some rich insights on the transparency of different audience you're reaching, the reach and frequency mapping in a very precise way. And this is just sort of on the paid media side, I think analytics, lots of possibilities there. But I think ultimately it's about partners like PEGA, Trade Desk partners like Snowflakes, all working together with the agencies and educate the clients and implementing.

- Fair, okay. Well, British Telecom has the largest first party install base data in the UK, which is amazing and impressive. How do you use that data to unlock value for your customers and what role does AI play in that?

- Yeah, I mean we do have a very large customer base. I think we've got a commercial relationship with about half the population of the UK. So definitely gives us a very rich first party data platform to build all of our customer engagement processes, marketing processes on top of, I think first and foremost, once you can get that data together, which obviously is not too easy and the use cases we have started with have been a lot more at the traditional marketing use cases around how do we drive better cross sell, how do we do better churn retention? And that's using our traditional channels, over well, at least as long as I've been there and probably for a lot longer, BT has been trying to utilize AI to improve those customer engagements. They varying degrees of success. But we have recently I think started to see a lot of fruit coming outta that. And I think Neil can maybe point to some of those examples a little bit later. As I mentioned earlier, our next best action solution will be sort of at the heart of all of our ongoing customer engagement. At the heart of that, you create a measurement foundation and then on top of that, the ability for AI to come in and really stretch its legs and start to show how the ability of AI to I guess, start to show some incremental value around those customer touch points. Whether it's through talking to our agents in the contact center using our app or our web. We'll start to measure that and grow that over time. We've got a fairly lockdown plan for that when we look at how we're gonna use AI on top of our first party data in paid media. To be frank, it's a little bit more of an unknown territory.

- Okay.

- I think we have some ideas and some concepts, but the first part of our strategy there is, look, let's build a solution that allows us to live without the DMP, without the cookies. Let's get that in place. Let's also look at how we can start to build integrations right next to our next best action solution that allow us to influence and change audiences in the ad platforms. Let's start to build feedback loops going into our agency to help improve and optimize that. And then I guess conceptually we're seeing over time the optimizations that we're bringing into our own channels where we are using AI to drive a better customer experience, sort of going on faith. There will be a way for that to then drive better audience selection, better targeting, greater, more consistent customer experience across the channels, including our main ad partners as well. But in terms of using AI, I guess Neil, you might talk to some of the learnings we've had over the last couple of years on how to best utilize that within the company.

- Yeah, I mean, in terms of using our decisioning and our MBA, I think it's been a real journey for us. I mean, we've effectively had to change the way of working as an organization. We've been quite a traditional organization used to sort of manual interventions as a marketing team. So over the last couple of years on the old legacy PEGA platforms, we've been working quite hard to improve the skills of the teams. And it really has been about the ways of working and the commitment to utilize the technology from the teams. I mean, part of that has been around the way we've set up, we've got agile squads working now on this with areas from all around the business, from the trading teams, analysts, data science, et cetera, all working together as a collective squad with business outcomes at the heart of what they're trying to do. That's been a big step forward. And I think in terms of what we originally had in our NBA engine, we originally had 150 business rules driving what customers were recommended.

- Wow.

- And bit by bit, part of what these guys have been working on has been taking those rules out and effectively trying to give more scope for the MBA to actually make some decisions. And that has been a big part of the journey for us in terms of AI and the generative AI, this is a hot topic in marketing, talked about it this morning. I think we're sort of making tentative steps on that. We're certainly looking at the use cases of around copywriting. Again, you really have to work quite hard with the teams who are responsible for copywriting today, for instance. You don't wanna sort of make them think they're gonna be out of a job. It's actually more about the skills that they're gonna have are actually gonna be slightly different. And we actually call the squads human and machine. So that we're saying it's not about a machine taking everyone's job or decisions away from, it's actually about the skills of the people are gonna improve to make the most of technology. And that's what we are really aiming for. So we are doing tests around subject headers of emails around PPC, copy, et cetera, where we'll potentially have thousands of generated copy that can be optimized and targeted customers. It's early steps and we're gonna sort of take it relatively slow at a company like BT. But I think there is definitely an excitement from everybody that actually, we can really do something that's changing the way we we do marketing.

- Sure, so Jacqueline, adopting AI is often scary for many brands and it requires organizational transformation. So what prevents brands from taking the leap and jumping with two feet like BT has into the AI revolution here?

- Well, I think my analogy is AI is akin to when the internet first started blowing up many years dating myself here. But the corporate strategy essentially then is just adding a .com at the end of your company name. So that's kind of this excitement that we're seeing with AI that's blowing up. But there's also a sense of that dauntingness for companies. So like where to start, you know, you mentioned some of that. It's like the companies that have silo team and legacy systems in place, they might be risk adverse. So a big complex organizations, it's gonna take time to shift. So those are all potential barriers that can slow things down. And a lot of companies here, and I think for the audience included, digital transformation has been part of that sort of similar journey, right? It's a multi-prong approach when you think about transformation and an evolution of a enterprise company. You don't just hire a consulting company to come in and give you a solution. You don't just have a senior executive that lead the charge. It's a multi-prong approach. And same thing for AI. A lot of it, I think what we try to recommend is have these specific use cases and have the squad team in place so you can get that, gather that learning, and think about it as this crawl, the walk, the run kind of phase approach. So it doesn't have to be all or nothing essentially.

- Can you talk to us a little bit about the organizations that you're seeing use AI to replace third party cookies and how that's sort of going for them?

- So I think there's a slow evolution of that and there's instances where we're I think touching on some of the things that BT is doing with personalization, now personalization we all know is not new, right? The custom experience, the email that they get, the content that they receive, the landing page experience, all that is part of the personalization. You know, how do you connect with loyalists, new customers, prospects? I think with generative AI to what Neil mentioned, it's that versioning that's available. How can we hone in on your segmentation? I'll use the financial services. A lot of the clients here, attendees here are from that sector, we have these high earners, but not yet rich. You know, it's a common segmentation in marketing and how do we further look at that? Do we layer on the delivering life stages? Do we layer on the first party data and many so on iterations that we can get to? So that's the sort of the power of AI that can come in. And I'll touch on on the analytics side, predictive AI has been in use for many years.

- Yes, yes.

- You know, how do you analyze the data to run cluster analysis? The data science team, a lot of companies have that in place. And connecting that to your marketing activity, your digital activity for pay media, that becomes the threat that can bring you through from a microsegmentation. But ultimately I think with AI, it's that hyper-personalization at scale that's possible.

- Amazing, all right. To close out the session, Paul, Neil, if I had a magic eight ball and I shook it up and I said, what's the outlook gonna be for the future of MadTech without cookies? What's the outlook? Good, not so good, tell us later.

- I think for us it's gonna be fantastic.

- Yes.

- I think as we mentioned earlier, the position we're in with sort of definitely the largest relationship in the consumer marketplace, at least commercially in the UK, I think strategically from probably from the CEO down, they're seeing this as a moment of great opportunity for us to not only look at how we can improve the way we use the paid media channel to create a more seamless integration with our on on premise activity, but also how can we start to think about our customer base and working with our customer base to go into new market segments or go into new product categories as well, adjacent to the core telco. So definitely opportunities to optimize and improve how we sell mobile and broadband and TV propositions. But then potentially opportunities to leverage that sort of competitive advantage we have with our first party database and start to look at, well, how else can we exploit that? What other products and services can we start to offer? So I think that's seen that we're at that tipping point now that the change that's sort of being forced upon us forces us to make a good change.

- Sure.

- We're making the investments, I think that we're probably about halfway down the journey. There's still a long way to go.

- Yeah.

- I guess there's always a long way to go, but yeah, I think the outlook's fantastic. Neil?

- Yeah, I mean, in my 20 plus years or so aging myself, I'm working in sort of below the line marketing. I think I think as Alan touched on this morning, it feels like in marketing in particular, we're at a real tipping point now in terms of using data to drive customer outcomes. I think we talked about it for a long, long time, but I think actually we're taking a step over the horizon almost now. And certainly at BT, we've made such great investments on the data platforms, the automation, the channel integration, the world class MBA, thinking about steps into language AI, et cetera. You know, all those sorts of things. We sort of feel like we're taking that next step. It's also really exciting, I think for the people in the organization. You know, I think for the traditional marketeers sort of learning new skills, embracing a different way of working, even the agents embracing a different way of working a different skill set, I think it's exciting for everybody working in this area now and really ultimately changing what marketing is effectively changing the way we think about it. So I think it is really exciting and I think these sorts of changes only expedite it.

- Yep.

- So they just force us to do things even quicker.

- Absolutely. So I think from our point of view, it's a massive opportunity and in terms of business outcomes, customer outcomes, and people's careers and enjoyment of their roles, I think it's only a fantastic thing.

- Absolutely, awesome, well thank you guys so much, experts. Do we have time for questions? Yes. Any questions from the audience? Sure, step right up.

- So apologies if you answered this in the beginning. I was a few minutes late, so disregard if so, but I'm wondering how you're going to test your models, your internal models in paid media against Meta's and Google's models. I know we are using real time data through CDH to create exclusion files in paid media, but we have yet to find success in our models versus the really honed in models that Meta and the other companies have already created.

- Yeah, I can talk there. I think as I mentioned in one of the answers, to a large extent we haven't really worked out how we're gonna apply AI into the paid space. But the current thinking, it's not so much around our models versus Meta's. We're probably looking at it more around a how do we augment what we're already doing with met the lookalike audiences and some of the advantage that Meta and Google bring to the table. There's definitely things that we know about our customers. So particularly like if you think about a mobile contract, which might run two years, we know when the customer's coming outta contract, we know a lot more about what's going on within the customer's geographic region in terms of the new broadband technologies that be coming available. So I think the approach we're taking there is more the insights that we're generating there. The activities that we're generating through our own channels, how are we using that to perhaps maybe create some separate audiences that are driven by our first party data, but it's probably currently being looked at more as a how we augmenting rather than competing with the Meta and the Google models.

- Great, any other questions? Close out on this last yes or no question, starting with Paul, do you think cookies are gonna be deprecated?

- At some point, yeah.

- Neil.

- Oh, I think there's a good chance it'll take a while, but yes, probably at some point.

- Fair Jacqueline.

- I'm just gonna say no, Google is just gonna keep pushing it out.

- Okay, Liz.

- I'm very torn, but I wanna say yes. I hope it's yes.

- Yes.

- But I do agree. I'm afraid that Google is going to keep at it, okay. I'll stick with, yes.

- Okay, I'm gonna say no. Thank you everyone.