PegaWorld 2025: The Future of Customer Engagement: From Blueprint to Bottom Line with Customer Decision Hub

PegaWorld 2025: The Future of Customer Engagement – From Blueprint to Bottom Line with Customer Decision Hub

Welcome everyone. I see some people trickling in so you can come on in. Um, this is the breakout on the future of customer engagement from Blueprint to bottom line with Customer Decision Hub. So today we're going to get into the vision and roadmap for CDH Customer Decision Hub. But first I'd like to introduce everybody.

Um, I am Kyle Munderville, a product marketing manager for CDH. And joining me I have Shoel Perelman VP of product for decisioning and I Vince Jeffs Senior director of Product strategy for marketing and decisioning, and Annette Kerlin Senior director of Specialist Solutions Consulting. So thank you all for coming today.

day, so you should have all been super pumped for Blueprint during Kareem's keynote this morning. Um, it was very cool to see a demo of customer engagement Blueprint right on the main stage, so we were personally psyched about that. And I hope you all are very well versed in customer engagement. Blueprint.

If not, here's a QR code to scan it, bookmark it, keep revisiting it because we are constantly updating the product. So on that note, I want to get right into Blueprint um Chel to kick us off. What are your favorite new features that are coming from Blueprint right into Customer Decision Hub? Well, I guess I'll just start by saying, um, I was thrilling for me to see Blueprint for CDH on the main stage.

I know there's been a lot of fanfare and tremendous progress with the case management version of it. Um, but I'll just even go back to 2023 when LMS hit the scene. We got all excited, and we could feel deep down that this was going to change marketing. But the first few iterations, I feel like we missed the mark.

We had this feature to be able to generate treatments from within CDH, right? But the reality is, I think what we were missing there is kind of what we're all getting to now, which is that it's not enough to just have a prompt. You actually have to have a workflow, which is actually what's behind Blueprint that tries to take the content and extract what you need out of it to put it into the structure that CDH really.

I think we've come to a good point with NBA designer because it is that structure. So I think like my favorite feature is just the ability to actually look at your real marketing content and put it in in context of a way that marketers can understand it. I think that's been a like a gap that we've had for some time.

And this really gets us to that mantra that we've been trying to get to, which is meeting marketers where they are. Cool. It is certainly an exciting time for us. So we're going to get into a lot of the details further on in this panel, but to kind of level set, can you just take a minute and kind of get everybody on the same page in the audience and just recap some of our recent updates and what they should know about CDH beyond just Blueprint? Yeah.

So in the past year 2024, as you know, we've moved to these yearly releases. We had two last year, which was kind of special. That's not a normal thing we plan to do anymore. We had 24 124 two. But I'll just recap some of the biggest things that happened last year. So one of the biggest ones was the focus on outbound. We introduced message stream, which means that you no longer have to contract and do integrations with an external email service provider. We have that built into CDH. It's actually I'll just make it known here. If you're not doing that yet, it's actually going to be free for the first certain amount of emails just so you can get started and try it out.

We really want everybody to be using outbound with CDH if you if you might not be. The other big thing was customer journeys. So you know, we've sort of come from this place that next best action will just consider each action individually and choose the best one. But it's always been a little bit hard again, to meet marketers where their minds are, where they want to be able to see them placed in context of each other in a sequence, even if it's not a hard sequencing.

And so I think that was one of the things I was most excited about in 2042 was customer journeys. And I think finally the features Key Audiences and Reach Explorer. Those were also making it easy to be able to get counts right. When I run this Next Best Action thing, I don't want to have to do a simulation to know how many people are going to get each of my actions.

I want you to just hit the database and and tell me. So those are some I think the key highlights. All right. Awesome. Thank you. That's a good level set for everybody here. Now I want to introduce Vince to kind of bring this into a larger context for us. So Vince I know you are very tapped into what's going on in the market and the industry as a whole.

So how are you seeing what we're doing, fitting into some of the trends right now? Well, it's interesting times. I think nobody would probably argue with that. Um, and the one thing I'll say about Blueprint is I think that one thing that we've struggled with for years is to to give people a way to understand what we mean by real time interaction management and AI and decisioning.

And now with Blueprint, I don't know if you saw the little feature where you can simulate the, you know, what's actually happening as you're arbitrating and you're getting to that next best action. So to me that's really exciting. But in the market, I think what we're seeing is we're seeing, um, that no surprise at all.

Data is extremely important to any kind of marketing solution and certainly to, you know, a solution like CDH. And so what we're seeing is there was a, you know, CDP's kind of came up about 7 or 8 years ago, and we saw a lot of startups pop up. We saw then the big marketing automation platforms adopt their own CDP.

And now what we're seeing is I think that's shaking out a little bit. So we've seen a lot of these CDPs, they've been bought by larger companies or they're struggling. And some of the big marketing automation platforms that have CDPs are struggling a little bit with data management. And we actually saw, I don't know if you saw the news the other day, but there was a very large one that bought a very large data management company.

So that's no surprise. Data is important, insights are important. And that's going to continue to be a really important theme. The other thing that I think is really exciting to all of us that are responsible for the product at Pega, is this problem that's existed out there for years, which I will call the content bottleneck.

There's always been the ability to create lots of, you know, small segments, target customers. We have a very unique way of targeting customers. We think it's a good way to target them. But, um, to get the variations of content that you need to do personalization, that's been a struggle for organizations because most of that has been manual. Most of these, you know, these treatments and these images and these different variations of copy that you have to come up with have all been created by humans and tested by humans and done in really very labor intensive ways. So I think it's really exciting to see the market wake up to and GenAI is helping this, this whole problem of content variations and how to test those.

The other thing just real quickly is we saw it this morning, but this whole idea of how we design for these systems, how we design actions, how we design strategies, journeys, um, that no longer needs to be done on whiteboards that can be done using things like Blueprint, which is pretty exciting. So you don't need to do these little yellow sticky notes on whiteboards.

You can use technology to really do that. And the last thing I'll mention, obviously, I'd be remiss if I didn't say something like Agentic because that seems to be in the news every day. Um, but the reality is, I think that's exciting. But the way I look at it is it's like having a lot of, uh, excited teenagers in, uh, in a classroom that want to learn.

They're very eager, but they make a lot of mistakes. So we're at the early stages of really sort of educating all these agents on how to do work for us in a quality and controlled way. But it's an exciting time, and I think it's a really exciting area that we're going to see develop. Cool. Yeah. And Alan's example this morning was just perfect for that of the eye that doesn't quite get it with the chest.

Exactly. Um. So cool. Thank you. Vince. Now, Annette, I want to bring you in here. Um, so everything that they both said. Um, how does this relate to what you are seeing out in the field? Um, I'm sure a lot of our customers are here with us right now. So what are some of the aspirations that you've heard and maybe some of the things that might be blocking or preferably enabling them to use these great new features that we have to navigate the landscape that Vince just explained.

Yeah. Thanks, Carl. So I think what we're seeing is this massive desire to harness the power of GenAI, but many of our clients don't know the best way to deliver the right business outcomes. There's so many capabilities and tools out there. It's hard to see the wood for the tree sometimes. And if you think about a typical enterprise level organization's challenges, you know, you often have organizational silos, data silos, lots of different ways of talking to customers.

There's a whole challenge of speed to market. So how quickly can you get your propositions out of the door to satisfy your customers needs? So actually, what we can do with our technology is to democratize the technology so that actually the people that really understand the needs of customers have the tools then to meet the needs of the customers and aren't hamstrung by all of those challenges.

So I'd love to share an example of this in practice with one of our newer clients, Nationwide Building Society. So they they signed up and chose CDH as their personalization tool less than a year ago. They've got a wonderful vision, so I need to make sure that I get this right. So it's to be to deliver the best outcomes through personalization of any bank anywhere.

So I love that as a vision, because what it's telling us is actually when they think about personalization and decisioning, what they think about is the entire customer experience. So there are member owned organizations. So all the value that they generate for their members gets plowed back into that value relationship. So over the space of six months, they built out their program around four key areas. So technology data consent. Because consent is obviously really important because if your customers don't want to talk to you, then it's very difficult to personalize your offering to them. And then the final piece was around um skills and people.

And obviously this is really crucial not just from a skills perspective, but actually if you have a big vision and you really want to drive value, you need the hearts and the minds of the people in the organization to drive it. So I think that's really important. So over the space of six months, they've built out two channels.

Um, so they've built into the app and onto the web. They've built out multiple use cases across various products, and they've delivered a 20% improvement in their outcomes in just six months. So I think this is incredible. Um, they've had 100% improvement in their engagement. Clickthroughs. Um, they've doubled the number of models they've got, and they've got big aspirations to really build out their model library.

Now, any client that is using our adaptive model capability, you know, I think can tell you what the benefits of this is. So one of our clients, NAB, has got 2000 models in their repository. Now actually, if you think about how long it would take to build 2000 models, that's quite a lot of time. And there's a lot of people and actually, you know, being able to orchestrate that is the next challenge.

So I think for Nationwide, having that capability, being able to use out of the box tooling and to really learn from all the best practices we have has really helped them to get to speedy value. So I'm really excited about the journey. I think you're going to see some some great things from them. Cool.

Yeah. And Nationwide, they're actually speaking today I think at 315. So if you want to hear directly from them, I encourage you all to attend that session. Um, so I want to get into some of the big themes of our roadmap moving forward. Now, we kind of touched upon the importance of a variety of different channels to get the most optimal customer engagement.

So what is happening in our product to kind of rise to those challenges. And this slide has a couple of the key themes. But let's jump into the first one which is outbound. Outbound outbound. Outbound. And I think you might see some of our colleagues have a Pin. Ask me about message stream or outbound.

So yeah that's huge for us. Yes. So I'm very proud in this release. So we now have a world class email editor. So I know that many clients in this room, even if you've been using CDH to decide what to send outbound, you may have been using some other tool somewhere else for the marketers to actually design the content.

You don't need to do that anymore. And if you haven't started doing outbound, you don't have to worry about that. We now have the ability to create email templates, and those email templates can be filled with pieces of content that come from the next best action decisions, and all merge together and then sent out using our own message stream email.

Or it actually will work with your own, ESP if you want to continue using that. It's free to sign up for. We do have a piece of paper that needs to be signed, but it actually just is to give you access to that, to that capability. So I'm very happy about that. One other thing that's that's coming in the next release that I just want to raise visibility to is without a lot of fanfare. 1 to 1 operations manager has now come of age. It's become the primary way that our large clients are getting new content into the system. Inbound as well as outbound treatments. And so we've started to recognize there's problems that happens when you have 50 people all doing that. I never would have imagined that five years ago.

And we've now started to come across problems where there's conflict. Multiple people are trying to push things the same week. Someone's trying to pull something out and put something else in. So we're now piloting a new capability in 2001 called Concurrent Change Management, which is very much like what it sounds like.

Cool. Thank you. Um, so, Vince, a little while ago, you talked about the market problem of testing variation at scale. So can you explain how we are addressing that? Yeah. So, um, we're going to be working on that. This is Cheryl's been talking about some things that are coming in our next release, which is out shortly.

Um, but I'm going to talk about things that we're planning for our 26 release. So this is a little further out, but recognizing that, um, if you've used CDH before, you know that there's some things you can do to we always call it going under the hood, and you can go under the hood and you can you can wrangle strategies to do things.

But the problem with that is it's not something that an average business user can generally do. It costs, you know, uh, you know, to have a consultant that's skilled in that, it also makes it more customized. So it's more difficult to deal with and upgrade. The debugging is not pleasant. Yeah. So we want to stay away from that.

So we're going to build a user friendly design of experiment capability. And that'll be something that you'll be able to do, you know, in a in a proper user friendly interface. Um, and that will enable you to set up many of these experiments at once. So you're not just doing an A, B test, you're doing what I like to call an a, B test.

The end is, you know, many tests at once. Um, and the, the other thing that we realized as we're doing this so as you get, as you get, um, you know, Blueprint creating more inspiration for you in the variations of actions and treatments and the, and the, you know, the, the versions that you can have.

What do you have to do? You have to be able to handle a lot of those. So currently, if you ever have dived into some of our documentation, you'll see something called service limits. And we talk about limiting, you know, the system to 5000 treatments in 2500 actions. Well, what we recognize is that's not enough to deal with this.

You know, this content bottleneck breaking. And suddenly you may need to be testing a lot more than that at scale. So we're going to make sure that the system can handle many more times the number of actions and treatments at scale. Now, bear in mind you really don't have to to test all those variations at once.

Generally, I like to think of it as kind of like a, you know, like a conveyor belt through time. So you're, you're testing variations over time. But still, we know that you need to be able to handle a lot more than 5000. And the other thing is, we know that you're going to want to have the system automatically monitor those experiments. You don't want to have to have humans looking at all these experiments. So we're going to have some automated facility for monitoring that. And underpinning all that is our realization that we really do need a proper business intelligence platform underneath it. So we're we're going to be looking at pulling in, you know, a more modern business intelligence and reporting platform and using that so that we can provide this reporting and analytic monitoring capability.

Cool. Thank you. So our next topic on here is Blueprint and transformative AI. So I know we've kind of touched on that a little bit. But how will that be baked into CDH moving forward. Well, so we'll start from where we were in 2042. So at the end of 24, I feel like I'm like recapping a TV TV series.

Previously seen on CDH. In the last episode at the end of 2042, you could get to the end of your Blueprint. Download a Blueprint file to your laptop, go to 1:1 Ops Manager and you could seed a new change request from that Blueprint. So you'd upload the file and it would unpack it. It would show you everything that you created in your Blueprint.

You could sort of pick and choose which ones you wanted. You had to kind of map it to your system, like the issues groups don't show up in Blueprint because you're trying to keep it business level. People who don't necessarily understand CDH, but ultimately you have to sort of draw lines between them.

Now, where we're going is what we're calling round tripping, which is the one of the first questions that techies asked me, right when they, when they see Blueprint is. That's nice. How does it relate to my real CDH configuration? So the way we want to solve that is because we recognize there's a line here, right? There's your system that you own that's in your private VPN.

And then there's, you know, there's Blueprint, which is a real SaaS system. It's shared. So we want to be able to do is make it such that you can go into CDH, click a button, which will take a snapshot of your configuration, not your customer data, but like the schemas, the taxonomy, the engagement policies make those available such that when you go in Blueprint, it can create blueprints that are cognizant of how your system is configured.

Then you'll click a button which will then publish it and make it available to your CDH system. So instead of having to upload a file from your laptop, you'll go into ops manager. It'll say, hey, you have three blueprints ready to be imported into your real system. Which of these would you like to import? So that's really where I see this going next.

Cool. And then just kind of talking about ops manager. Um, so and you kind of mentioned how this opens up the usage. Um, I thought, Annette, I think you have a good example here of why do you think it's important to kind of get beyond that kind of older usage and bring it to marketing operations? Yes.

I think this is a really interesting topic, because one of the amazing things about CDH is it's it's an enterprise level capability. So if you think about the snippets of conversation that you have with your customers, actually what CDH enables you to do is to join up all of the dots across all of the journeys and products so you can have a really seamless conversation.

What that means then, is that you have to build out your action libraries. You have to connect up the journeys because otherwise you miss crucial parts of the conversation. But from a business owner's perspective. What that then means is as they're rolling out CDH, they have to translate for each area of the business what value this is going to bring to them. So it is, you know, it's a transformation journey. So actually, when we move out of the business ops area and we use things like Blueprint to help people to understand how CDH works, to brainstorm new use cases and to federate out that usage. It then means that journey to really deliver an exceptional outcome with CDH becomes a lot easier, because people are part of that, that journey and that culture.

Cool. Thank you. And then for our fourth bubble here at Pega, we've had a pretty strong reputation for being market leading in data and analytics. Upper right hand corner. Yes. Yeah. Proof of that. Social. What are we doing to maintain that upward trajectory? So I guess the first thing that that I'll talk about is the maturation of adaptive gradient boosting.

So those who have been using the product for a while know that our adaptive models for a long time have used the tried and true naive Bayes, and I always get heckled by data scientists. They say, why are you using an old model like Naive Bayes? And I say, well, because it works really freaking well.

It learns from sparse data. It's really fast. But we kind of knew in our heart that there are better models out there, like gradient boosting. And so for the past 3 or 4 years, we've we've actually had this latent capability in the code to use gradient boosting, but we hadn't switched it on by default because in our own test we actually used CDH internally.

So like when you get emails trying to persuade you to come to PegaWorld, those are actually arbitrated by CDH, right? Somehow it worked. Yeah, you're all here. But so we've been battling the Naive Bayes versus this gradient boosting for the past three years, and we kept wanting it to be better than gradient boosting, but it kept not being better.

And so this release 25.1, this is the first time that we've actually gotten it not only to be as good, but to be better than gradient than Naive Bayes. So that's now actually the default in. So if you're starting, if you've just licensed CDH this year and you're about to go live for the first time, you will start in with your models using gradient boosting.

So that's one. Vince talked about the having being able to move from low thousands to tens of thousands, if not even hundreds of thousands of treatments. That puts a lot of pressure on the models. So we have work that we're doing at the model level to make those a lot faster, because we still only have a 200 millisecond budget for making a decision.

So if we're arbitrating a lot more, you know, a lot more pieces of content, we have to be able to cut a lot more of those out much earlier in the process of of the strategy running. And then I'll talk a little bit about conversion modeling and external data access. So conversion modeling is really the the idea that right now if you've been using CDH The models are learning based upon what people click on.

We call that engagement. So engagement models right. Conversion models are trying to show people what they're likely to buy, not just what they're likely to click on. It's somewhat of a non- trivial problem because we get the clicks immediately right. Within a few minutes on when it's on the web or mobile.

But if somebody actually gets a credit card, you might not know for three months, right? So we might get a feed, and then we have to come up with a way of being able to take the engagement models and the conversion models and blend those together. So that's something that we're we're working on for the next version. But just to highlight something that we've been on a journey for for the past two releases is much tighter integration with Google Cloud. So in 2014, we had the ability to ingest data from BigQuery into CDH. We added the ability to listen to data changes from Google Pub / Sub. Now in 2001. We have the ability to be able to do segmentation on source.

So you can actually run a segment query against BigQuery, use that for an outbound schedule. And going forward, we want to even allow you to use the results of those segmentation as engagement policy criteria. So apologies if you're new to CDH. I just said a lot of stuff that was probably gobbledygook, but if you're using GCP, you will appreciate this stuff.

Yeah. So thank you. That was good. You all helped paint a good picture about current state of CDH and the kind of themes moving forward. Um, now, Vince, I wanted to get a little bit strategic here. Um, what is your perspective on what other vendors are doing right now and what they're offering compared to the stuff we all just talked about? Yeah.

Well, a good segue from what Joel was just talking about with data, because remember, I talked about CDP and that whole sort of shakeout that's going on and some of the troubles that some of the big players are having with their own CDPs, we don't believe that, you know, you need to necessarily have a CDP hooked up to, you know, CDH directly.

We believe that you should be able to use a big hyperscaler in these large data management platforms that seem to be sort of winning, that, you know, that battle over time, not only Google, but also Databricks and Snowflake. So there's, you know, a couple of those big players that we think that's the way that the world will shake out eventually.

Um, so that's that's one difference. Um, the other difference is I think there's we have a very different approach to, you know, how we're how we're using Agentic. I think, as you can see from what we're doing with Blueprint and the kind of the governance that we're putting around, um, what those agents are able to do and how they are all orchestrated and ultimately how they're held accountable for what they do and how we sort of maintain the quality of that.

Alan talked about how a lot of this is done in design time, and then at runtime, we're doing things that are more predictable. So we still think that's a big difference between how other people are talking about Agentic and how we're, you know, sort of the reality of how we're going to employ Agentic in our, in our systems.

Um, and I think the the other big difference is the, um, the way that we do, you know, the, uh, the, uh, channel, uh, proliferation, the handling of channel Decisioning Lead, um, many of the, uh, players out there really still look at that. As you know, we talk about Center-out. They they still put a lot of logic in their individual applications that are very channel specific.

So you have some new players emerging that are that were mobile first companies. That's great. They're they're putting a lot of pressure on the market to think about mobile and to think about digital. But they're really handling mainly outbound channels. They're not handling the whole proliferation of channels that you have to.

To give a great experience to customers and to be able to arbitrate all these things that you can do and then send them out. So we've always thought that that's a big difference in what we do. We handle a, you know, some of our customers will talk about 20, 25 different channels, and they're both inbound and outbound channels. Cool. So, you know, we like to think of ourselves as a class of our own. So thanks for highlighting that, Vince. Um, and we kind of went through some of these heavy topics related to AI like governance and transparency. Um, but Annette, how are you seeing this resonate in the field? Um, how are customers actually reacting to the use of AI, and what are their plans for using it to streamline their business goals? Um, obviously there's certain things that AI does really well, automation and whatnot, but there are other things that humans are just simply better at.

So I think this is a really interesting topic. I imagine everybody in the room must at some point have wondered what what impact is I in GenAI are going to have on my role? What does it mean for me? What does it mean for my customers and for my business? So I think trying to sort of evolve that conversation and help our customers to get that balance right is really important.

And if you think about the things that humans are really good at in terms of strategic thinking, thinking outside the box, um, sort of having judgment and empathy and being able to consider lots of different variables to make sure that something is is meeting your organizational strategy. Whereas maybe AI is better at the task oriented activity, mass scaling and execution.

So actually the sweet spot is between making sure that you plug your AI in to do some of the things we've seen our clients do. So I'll give you an example in a second of National Australia Bank. Um, so where we become the strategic orchestrators of AI, the instructors of AI. So we can see how it's performing, and then we can then make informed decisions about how we think it should be driven.

So National Australia Bank is a fantastic client. Um, they have big ambitions with CDH. They've already delivered some significant value for their business. They think about CDH as a capability that can optimize every interaction with their customers. So last year they they delivered over 350 billion decisions.

Hundreds of millions worth of decisions, interactions. And actually, you know, what's helping them to scale is I mentioned before that ability to activate the modeling and the decision within the brain. So they've got 200 actions in their action library and over 2000 models to really optimize the conversation with the customer.

So they're seeing some fantastic results. Um, what's also interesting about the way they've they've built out the solution is the skills of the people that they've looked for. So actually, because we've spent quite a significant amount of time and investment trying to simplify the interfaces with CDH, CDH and Business Ops Manager.

What it means now is you don't need to get people that have a heavy technical skill set. We've federated the usage of that so that actually you can get more people in the business using it and scaling out the solution. So I think getting that blend right is really important. And you know, we're seeing more and more clients being able to do that with with the capabilities we've built into the product.

Cool. And yeah, there are another good example. They have a session I believe tomorrow. So that's another one that you should all check out. Um, and Annette, what are some other innovative ways that you're seeing people use CDH? Like companies that are really kind of embracing this wave of innovation.

So Sheryl talked about outbound. I mean, outbound is a massive topic for us because I think historically, sometimes marketeers and I'm a marketer by trade, by the way, marketeers sometimes struggle to get their heads around, you know, with my campaigns, how am I going to implement my campaigns in CDH? What if I don't want it to be always on? So we've really invested heavily in customer journeys as a capability to help marketeers to map out the things that are more deterministic than maybe something needs to be scheduled, but it also is a mechanism to engage with the rest of the business.

So one of our customers, OP bank, they've been experimenting with customer journeys. And what they found is that the journeys are a really excellent way to work across the business teams and the tribes to look at where there are gaps, maybe in customer engagement and obviously some of the tools that we have also help you to spot gaps and to map out where there are gaps in their actions and their interactions with customers.

And what they found is that's really helped collaboration across the business as well. So obviously now that we have Blueprint, that is going to be a brilliant tool that we're already seeing our customers use to help to bring more of the marketing people on board that maybe, you know, before didn't quite get all of the capabilities in CDH.

So I think that's, you know, a really exciting tool for us to be able to use. Cool. Thank you. So you touched on some really impressive client metrics in it. And we all know that clients are seeing tremendous value from CDH. Um, earlier this year, Pega commissioned a Tei, which is total economic impact study from Forrester, and that found that companies can see upwards of $200 million in yearly incremental revenue when using CDH to its fullest capability.

So we all know that the possibility exists, but we also know anecdotally that doesn't quite happen for everyone. Um, so hopefully you can help us out here. What is going on in the product to make sure that everyone in the audience is able to realize the true value of CDH. Yeah. So we we recognize that not everybody has the luxury of having a data science team, like digging around and looking at how the models are doing it every, every day.

Um, but you don't. If you're running on Pega Cloud you're familiar with, like, you don't have to have it. Teams looking at the CPU and the memory and the heap. Right. We take care of that. So we're taking that same philosophy and bringing that to decision health. So even if you're not on the latest version.

If you're on 24 two, you could. If you go into Pega diagnostic Cloud, there's a new tab. It seems somewhat hidden because you have to choose. Look at things over the past month, and you'll see a tab that will show you metrics about your decision health. It'll show you how many actions are in the system, how many, how many treatments.

Now, what we're going to be measuring is how many actions make it to arbitration. And the reason why this is important we actually have a word for this. We call it optionality. And it comes down to if you only if you have such narrow filtering criteria that by the time everything is filtered out and you get to the adaptive models, if you only have 1 or 2 things remaining, what can the models really even do? I mean, their arms are sort of tied together.

So it's historically we've had the ability to measure this using value finder where you would run a simulation. But we now realize we don't need to ask you to run a simulation. We could just take this sampled from real decisions being made every day. Right. And so we're making that a metric that you can look at. And we're also going to be monitoring it ourselves to give you a heads up, a notification if that number is drifting too far down. Another one like this would be simply number of decisions where nothing's returned at. All right. You had an opportunity, a, you know, a precious moment to interact with a customer, and you had nothing to say that was relevant to them.

We want to surface that up. So that's I think, what we really we're really striving to get to in, in terms of helping you maintain the health of, of your decisions. Cool. Thank you. Yeah. Because we would love everyone to succeed. Um, so I do want to leave time for Q&A at the end. Um, thank you all for sharing your thoughts right now.

But before we close out, let's all just take a quick look in the crystal ball and see what do you predict for the next 2 to 3 years? Um, so, Vince, let's start with you. I'll try to make it short so we can leave a little time for questions. That's that's a big question. Um, but I do think that that the what we're seeing in agentic, it's going to improve.

And I think these specialized agencies, some people call it mixture of agents. They're going to get better and better. There's going to be more and more of them and there's going to be better orchestration of them. And I think there will be. It won't just be internal. So it won't just be about Pega agents, it'll be about other agents that are actually talking with Pega and back.

So there's more standardization that's going to come with that. There's you may have heard of agent to agent protocols and model context protocols. So there's these things that are starting to form in the industry right now. The other thing that I think is likely to happen more and more is that the way consumers interact with brands is going to change, and it's going to become more automated and consumers are going to take more control over their agents.

And so when you think about that, also, what's going to happen? Well, the brands are going to want to influence the agents. So what you're starting to already see is years ago there was this big area everybody knows called SEO, which is how do you get your stuff at the top of the Google search query, right? Well, now there's a new term.

It's called Geo, which is generative engine optimization. So it's how do I get the chat bots of the world that consumers are using to surface my products and their answers? And I definitely think we're going to see that shake out more and be more influential in the next 2 to 3 years. Cool. Thank you.

Annette, what are your predictions? So just a quick prediction from me. Um, so what I think we're going to see. Just going back to my comment earlier on about us becoming strategic orchestrators, is that many of the screens and the ways that we build out some of the interactions we have with customers, a lot of those interfaces are going to go away.

And instead, when we're talking agent to agent, actually, maybe if your head of customer retention and you want to understand Stand, you know, in the next month, who are the customers that are most likely to churn? What's the most prevalent predictor, and what are the propositions that have worked most effectively for these kind of customers? Before your agent can go off and get that information for you.

You don't have to do lots of digging around in reporting and analysis to find out who they are and why they're there, and what to do with them. But then the onus then will become on the orchestrator to then work with the business, let's say, and and then think about, you know, what should a new proposition be for this customer then. So I think it will remove lots of layers of sort of business operations and allow us to be much more strategic in the way that we engage with customers. And I think it would be wrong not to finish with the comment about Blueprint, which is that's just going to grow more and more in sophistication as an orchestration tool for actually, you know, already we're seeing massive value from people using it to to try and, you know, um, help the business to understand the vision of CDH and to really understand how it works and then to grow out the value of their own implementation.

So I'm really excited about what's ahead with Blueprint. Cool. Thank you. And, Joel, you can close this out. What are your final thoughts? So I know we're all doing agent predictions. I'll I'm going to take it the other direction, which is I think agents are going to help a lot more with the mundane problem solving and testing, painstaking work that I know many people sitting here go through every day, because I know we're working on trying to get them to to solve that problem.

So help is on the way. Cool. Awesome. Help is on the way. That's a good way to end. Um, I do want to. We have about five minutes left. Um, so if you have any questions, you can stand up. There are some microphones here, or if you project, you can just speak your question and I'll. I'll. Yeah. Come on up.

Hey. Oh. Is it on. Is it. On. I don't know. You can speak and I'll repeat your question. Okay. I can be pretty loud, too. Okay. Thanks for talking about the cool capabilities. I also really liked how you guys talked about how people interact with it and engaging with marketing people, engaging with the tool and whatnot.

And I'm curious, like for a company like roughly like 5 million people that wanted to build a really strong operation around CDH and operationalizing that and really feeding that and training that. How big of an organization would you expect to see? I mean, collectively, like from marketing and technology to really be kind of feeding that engine to get the most out of it.

How big of an organization? Yeah. Like, how many people would you expect to be kind of driving and feeding that tool and operationalizing? Yeah, I think. It's all to leverage all the capabilities that you guys are planning. Yeah. Great question. I think it does depend a little bit on the size. That's why I asked the size of the enterprise you were talking about.

But 5. Million, 5 million. 5 million customers. Okay. That's a relatively kind of that's not a mega organization, but it's it's a good size. I mean, I think that in the past we've seen that it, you know, it easily takes 20 or 30 people to run that whole operation. I think you're going to see that sliced in half, if not more, in the next five years at the most.

I think we're already seeing that. We're already seeing that, you know, consolidation of the of the needed labor, um, because of the, you know, the automated capabilities. Yeah. Okay. So ten to a dozen. Cool. I'm not sure. If the mic is on, but I'm standing over here and I can. I can be loud. Sorry, I couldn't see.

Um, I'm really excited about what Blueprint can do, especially for CDH. But given that I work for lifetime, um, and we are not an insurance company or a bank or, um, really sort of anything in that space. How are we looking at expanding Blueprint to, you know, be a part of other industries and having other models and other personas and things in other industries to make more sense for other companies and other industries. Yeah, I guess I'll say that's the beauty of the fact that it's sort of it's backed by generative AI is I think if you were to try it right now and you just put Other for industry, I'll apologize that it doesn't have your specific industry there. I think you'll find that it would do a pretty reasonable job, because we've tried it for a couple of industries that are not sort of like, as you say, financial services comes up with some good stuff.

I think where it's probably going to take a little while to catch up is going to be on the data modeling side. Yeah. Yep. Awesome. Thanks. Thank you. You're welcome. Thank you. You mentioned that there are a lot of changes that have been made to the customer journey features of CDH. Could you throw more light on what's been the new changes that have been made to the customer journey? Sure.

So a lot of it is focused on the entry stage criteria. So in the past, entry stage criteria pretty much always had to write a strategy to do anything useful. Now we've added the ability to have time based entry criteria. We've also made it much easier to not have to drop down to expressions. So that's probably the biggest.

The biggest thing is time based movement across stages. And then with Blueprint the journeys that it creates, it's going to already give you a good sense of which action should be and which journey stages and steps. So, you know, even if you don't import, you will be able to import that. But even if you don't fully import that and use all that, it'll give you a good sort of mindset.

Again, we like to call it inspiration about which actions are appropriate at which places. And then, you know, the system can arbitrate based on the logic that it applies, knowing the context of the customer in that journey stage. The other big one was it used to only exist in CDH portal, which you know was a place to start.

But now that we have widespread adoption of 1 to 1 ops manager, we had to make that available in ops manager as well, which we did in 2042. Got it. Got a follow up question. Uh, we just upgraded to Infinity 24.2 and we were on 8.8. In 8.8 eight customer journeys, worked at a customer level only and not at an account level.

In our context, we have a customer who could have multiple accounts and our servicing strategy is more at an account level, but we could not really use journeys because journeys operate at a customer level. Do we see that also changing in the near future, maybe in 25. That may already have? I would definitely say, um, there's a booth which is called Introduction to CDH, but it so happens the person who's at that booth is the expert in this particular topic.

Okay. So yeah. Good question. Thanks. Thank you. You're welcome. Oh, I thought nobody's over there. Um, we have about 30s. If anyone has any last questions. Oh, yeah. One more. So in this meet, we talked a lot about legacy transformations. So let's say we started with DSM on our CDH journey. So is there something going to come up like where we can just lift and shift the whole CD DSM, what we have done into the CDH part.

Probably not a lift and shift. And the reason why I say that is there's usually a like a paradigm shift that a lot of DSM systems I see, like the way they make decisions is not even aligned with like the way NBA designer strategy framework works. But I heard just over drinks last night that there's at least one partner in this room that's working on something to extract.

You know, that stuff out of DSM system to be able to use it as a starting point for, for CDH, for Blueprint. So I think that's probably worth chatting about. Thank you. Cool. All right. I think we got through everything. If anyone thinks of any other questions, um, reach out to any of us and or come back down to the Innovation hub and all the people in the CDH section will be happy to help you.

Um, thank you, Joel, Vincent, Annette and I hope you all enjoy the rest of PegaWorld..