Skip to main content

We'd prefer it if you saw us at our best. is not optimized for Internet Explorer. For the optimal experience, please use:

Close Deprecation Notice


PegaWorld iNspire 2023: Panel - How Customer Data Platforms Supercharge Real-time Decisioning

Customer Data Platforms (CDPs) were designed to help businesses develop and manage a unified customer database and provide a full-spectrum view of customers and their potential needs. With the recent explosion of behavioral data and digital experiences, these platforms are quickly becoming an integral part of the marketing stack, with tremendous adoption numbers (and huge growth expectations). Join Pega’s Norma Suarez as she interviews decisioning practitioners about their customer data implementations and explores the powerful relationship between behavioral data, customer context, and real-time decisioning.

This session will examine the origins of CDPs, demonstrate how they solve for specific use cases in today’s MarTech environment, and enable attendees to evaluate the value of a CDP investment within the context of their broader goals.


- Welcome everyone, to the Pega Customer Decision Hub session, "How Customer Data Platforms Supercharge Real-time Decisioning". And thank you very much for joining us today. So please help me welcome our esteemed guests, Tom Barker, Senior Vice of Personalization at Wells Fargo.

- Few Wells Fargo here, thank you.

- And Fiona Kirk, Head of Customer Decisioning at NatWest. Not working . And Vince Jeffs, Senior Director Product Strategy for marketing and AI decisioning at Pega Systems. So over the last decade, customer data platforms also, you know, more colloquially known as CDP, have become an increasingly important technology for customer engagement practitioners. Prior to CDPs, organizations have many times struggled with the limitations of traditional customer data sources. For example, operational challenges, like disconnected customer data have effectively constrained many organizations from a full view of customer behavior. So enter CDPs into the market, course were designed to provide businesses with a unified customer view. Now, with the ongoing explosion of digital data and devices, it really is not surprising that we see CDP adoption, as they become more amenable to...

- Right, thank two or three? 2 beds.

- The mic.

- Yeah?

- Take on, you know, some of these... Overall, important consideration is that it's really need to have an in strong alignment with where-

- Title wouldn't fit here so.

- They need to to get to that solid customer. Now, manage their... At the end of the day...

- He's gonna get you

- Time, and I used my professional services career. Kind of felt it was...

- I think we're on a third party line.

- Different frequency?

- Yeah, my apologies. I really don't, I wasn't aware if I was back or not. Anyway, so my final thought on that was that a critical driver for CDP success is to ensure that there is a strong alignment with where the organization's at and what else it needs to do to close those gaps, to get to that solid... Now many organizations already have ongoing projects addressing various aspects of their data management. So in order for a CDP project to successful, it can't just be a departmental rea, it needs to be a fundamental component of the enterprise's customer data blueprint. And with that, hi Vince and Fiona and Tom, welcome. Can you please give our audience a brief understanding of your role in the organ?

- You wanna start with me? Yeah, so, hi everybody. My name's Vince Jeffs, I'm responsible for product strategy for Pega Systems. And I've been with Pega Systems for almost nine years now. But I've been involved with, we're having all kinds of technical fun. I've been involved with customer data. I'm wondering if, like, I'll have to climb a telephone pole to present soon, you know, but all my career I've been involved with how customer data needs to be organized and used effectively for, you know, for decisioning and for marketing purposes. So I'm really thrilled to be here and to, and especially thankful that we have Tom and Fiona here to talk with us about this new phenomena, CDPs, which I actually think is not so new, but interesting. And I think we'll get into a little bit of why that is.

- Yep.

- Hi, I'm Fiona Kirk. I work at NatWest, and my current role is head of customer decisioning and contact. I'm responsible for delivering data led omnichannel customer communication strategies across the retail bank, powered by data, advanced analytics and AI. And ultimately, the reason for our existence is to bring value to the customers through communications.

- Tom Barker, Wells Fargo. Don't let the accent confuse you, I am based in the states, so I do work for Wells Fargo over here. I have the pleasure of being a product owner, product manager for multiple parts of our marketing technology stack. So I sit in a team and organization called CDEP, which are customer data and experience platforms. So I have been on Wells Fargo's Pega, their CDH deployment since the beginning. So we've been running Pega for three plus years now, and obviously we're about to talk about what we're doing with our CDPs as well.

- Thank you, Tom. So let's get started with you then. What would you say were the top reasons that Wells Fargo decided to add a C and what was your CDP vendor selection?

- Yeah, sure. So, interestingly enough, sat here in PegaWorld, our initial journey wasn't actually looking at bringing the CDP into work with Pega. So I'm sure many of you in the room can attest to the fact there were other reasons Wells Fargo were looking for a CDP, whether it was decommissioning our DMP, whether it was identity management, many of the usual business cases, reasons for bringing in a CDP, they were the initial reasons we actually started looking. We actually changed our RFP along the way though, as we started to identify and unlock, actually this can do a lot for our Pega CDH implementation as well. 'Cause we went throughout that process, we actually changed our reasoning for needing a CDP, how can we use the data out of it in Pega, in our deployment? How can we use it in audiences? You know, we're gonna talk through some of those use cases in a minute. So we went on an interesting journey, I'm sure many of you can can relate to that. And it's grown since. So the relationships between the two has grown ever since. And we did choose Adobe at the end of the day, I think that we're not gonna hide that fact here, so we went with Adobe AEP.

- Great, thank you. And now Fiona, same question.

- Thank you. The first reason we looked at a CDP was to try and understand that identity management piece. So we were really keen to understand what our customers were doing both online and offline. And that was something we didn't have at that point in time. And the second reason was the real time element of what a customer was doing and how quickly we could find that out and then do something with it. What we were going to do with it, we had lots of ideas, but not really sort of formulated a complete plan around how we were going to implement at that time, and then we ran a POC to understand what we could get from it. And I think the amount of excitement we got around the retail bank was just phenomenal. And mainly around the two things, the retail bank and our exec team finding out that they could understand much more about customer journeys than they ever had before, just seeing that online and offline journey put together. So there was, it then started to grow much wider than what we can do from a communication strategy perspective. We started to think about our operational areas and our telephony centers and how they could benefit from understanding customer journeys across offline and online.

- Thank you for your answer. And Vince, you know, we're gonna pivot a little bit. I wanted to ask Vince about, in your opinion, what do you think are the advantages of a productized versus, you know, sometimes when an organization might not even need one or why they don't wanna go down the route of...

- Yeah, it's interesting because I always have the opinion that, you know, pre-packaged software, I'm biased toward pre-packaged software. Through my whole career, I've thought if you can buy pre-packaged software, you're getting something that, you know, has been the vendor that built it, if they're good, they've put a lot of time and effort into it, and they've got experience across, sort of the wisdom of the requirements of a lot of industries and a lot of different use cases. So there can be real advantages that you're gonna be more, you know, myopic in your attempt to build that yourself. On the flip side, CDP is not so much an application, it's really more of a framework for a data repository. So when you think about it that way, you gotta look carefully at what you're really purchasing from that vendor and how much, you know, help they're gonna get you in getting time to value, or if your data's not very well organized, there could be, you know, help there from that vendor and that CDP. But also remember that these days you're buying a lot more than just data collection, data management. Those CDP vendors are trying to sell you on the whole package, right? All the way through activation channels. And, you know, sometimes if you look closely, you'll realize that, you know, it's really sort of maybe good enough in some of those areas if you're at a certain level of maturity. But be careful about how, you know, fully functional those other capabilities for activation and decisioning are, because those are important, you really want good capabilities there. Gartner did a study in 2020 about MarTech stacks and how much organizations like yours use those and at the time it was like 58%. And then they did it again two years later, last year, and it went down to like 42%. So the amount that people are buying software and it overlaps, and they're not using some of those capabilities is pretty large and it's increasing. But more importantly is, what are those capabilities are not in the package that you're buying. So just make sure that you get, you know, a package that you're really gonna get what you pay for and what you need. You want real time decisioning, you want next best action capability, you need those activation capabilities that may not be in those packaged CDPs.

- Thank you Vince, for those insights. So let's pivot a little bit, Fiona, where are you, where's NatWest in the process?

- Where we are is, we have implemented the digital side of things, so we've got our CDP completely filled with all the digital data, which is stitched together. We are using it within the data and analytics team, but also in the marketing team and also in the digital team as well. And each area is doing their own thing with it. Where we're using it within decisioning is we have it hooked into Pega and we're using it to decide, lemme think, we're using it to bring in dropout journey orchestration across a number of different channels. And we're also using it on the brochure ware side of things to stop showing customers things that we know that they've just done in real time. So that's where we've had the biggest success so far.

- Great, thank you. Tom, where is Wells Fargo in here?

- Yeah, so it's remarkably similar story, believable or not. Yeah, well surprise hey. I think the main difference for us though really when we did start looking at bringing it into Pega, was where we actually started using it first and foremost. We actually went ADM first. So we thought about how do we bring in our data from the CDP and actually get it to influence the models within Pega. So a lot of our initial focus and the business casing actually from a Pega connection perspective was how do we improve the models? So how can we bring this data in? Can it be used as predictors? Can it help improve our overall performance and ultimately increase the performance of our next best action capability? So I think that's where we were slightly different from a, it had all of the usual business casing as I said earlier, the RFP itself wasn't necessarily there for a Pega CDP connection, it has other reasons to exist within Wells Fargo. So we've also been working on those use cases. But I found it interesting that we went ADM first. We've since expanded obviously, how it can impact audiences, how you can build out your eligibility, your engagement policies within Pega as well. A real point and purpose for our CDP, which obviously Fiona's already alluded to, was latency. So straight off the bat, can we improve latency? Can we get better data sources into Pega? They're very much a lot of the analytics we're doing at the moment to prove out its performance and evidence its contribution.

- Excellent, thanks. And you kind of started ask

- I did, sorry.

- Answering my next question, which is, how is the CDP adding value to, at your, I think, you know, to paraphrase for Wells Fargo, it's been accelerating and enhancing those data feeds that are going in H adaptive models, right?

- First and foremost, that was the quick evidence we were using is kind of the POC, the base of the POC. We've obviously, as I said, since moved on to building audiences, bringing audiences, aiding data latency

- Improving

- You know, we are a big bank, 250,000 employees, lots of technology, so where we can, increasing speed to market is key. We want to put the customer at the heart of everything we do, so we believe this is a key unlock.

- Excellent. So, and Fiona, so how about it now?

- Yeah, NatWest exists to be a purpose driven organization. It's there to help customers, people, and businesses thrive. And I guess having that additional quality data just helps us look smarter, be smarter with our customers. Obviously customers are dealing with us in a very different way today than they were a few years ago, they're much more digital in their nature. And to be able to be there in the channel of choice of customers, and make it much more worthwhile, less time wasting for customers, that's the beauty of it, I guess.

- Thank you for your answer. So Vince, talking about implementations and integrations, is there anything you would like to tell our audience on how Pega CDH best with these customer data platforms?

- Yeah, sure. We, you know, I'd say three or four years ago when the CDP kind of space started to heat up a little bit, we had already had partnerships with some of the data collection companies like and Tealium, and you know, there's other places where we had integrated, we didn't have productized connectors. So we always kind of realized that getting, you know, data fast into CDH and contextual data coming from, you know, touch points such as web and mobile where things are happening like today, or you know, whatever with customers, you know, things happen overnight, right? And they get processed and you get this, you know, learning about customers that accumulates over time and CDPs are pretty good about collecting and curating that view of a customer. So it was in our vision to be able to handle both, right? Always. And when you look at CDPs, look closely at where they came from. CDPs kind of came from at least three or four different places. They came from the ad tech space, they were DMPs that suddenly started calling themselves CDPs. There were, you know, analytics data collection companies that, you know, really were web analytics companies that suddenly were CDPs. There were, you know, traditional marketing automation companies that happened to have a data repository of some sort that called themselves CDPs. And then, you know, of course, there were more organically built CDPs. There were some that really were, you know, startups, that started with the whole, you know, we're gonna curate a marketing view of the customer. So with that in mind, we thought there's probably four connection points that we need to support with CDH and our mission in CDH is to be an open decisioning engine, right? To be able to get data in and to be able to get next best actions out. So the data in part is this real-time streaming part. We knew we have to have connectors, whether they call it a CDP or whatever, we want to have connectors to those people that can collect data and feed that raw event data in. One of the disadvantages is you maybe can't ID that always so easily. So that's can be where A CDP comes in. CDPs as was mentioned, are very good with identity management and identity graphs, and curating a view of the customer. Again, it's gonna be a little latent, but we wanted to be able to get that profile, those segment memberships into CDH. So we built connectors, once the big, you know, marketing automation vendors like Adobe and Salesforce jumped into the CDP fray, we were, you know, quickly building connectors to those platforms. And then the third area is really the, you know, the next best actions that we make inside of CDH, we want to be able to share those with other platforms. That's a natural. And then the fourth area is the history that we generate, the interaction history, the decision history, we need to be able to feed that into analytical platforms. So those are the sort of the four areas of connection. Yeah.

- So I guess our audience is probably a little interested in hearing from our clients, from Fiona and Tom, about what kind of results they're date and of course they're very new implementation, so it's an ongoing process, but anything you would like to comment on what you're seeing or what you are anticipate by this?

- One of the things that we've seen is the brochure ware side of things. So we've been able to half our displays but maintain our click through rates with the addition of real time signals telling us that customer's actually just done something so don't bother showing them that message. And that's my point about, you know, we're wasting customers time less and less whilst the commercials aren't any better, the click through rates, obviously, you know, it's just a product of those two numbers, it just gives us another opportunity to talk to a customer about something else that we feel is important, or nothing at all. But we've been able to reduce the displays and double click through. That's one of the examples. The other place where we're getting a huge amount of traction is in our short term borrowing space, so loans and credit cards specifically, they have a real problem with customers, you know, showing a level of interest in the product journey and then dropping out at different points. So we've kind of captured those signals across the journey and depending on how close they are to actually drawing down and taking the product, we send a different type of real time message, whether it's in a push notification or an SMS, or an email with different dialogue each time, as I say, depending on the level of interest the customer did show and that's getting really good click-through rates as well and draw down rates. What we're finding is obviously the closer the customer was to drawing down, the easier we can nudge them back into the journey and help them on their way again. But those, the collection of MBAs that we've produced as a result of that new data are consistently performing at the high top 10 out of our 3,500 MBAs, they're always up there, showing really good performance.

- Tom?

- Yah, similar. I'm not gonna share stats, but a boo hiss, but a similar deal

- Boo hiss.

- Of obviously inventory management is critical. So how we make the best out of our placements, obviously when we see a customer having that next best conversation, next best action. I always like to say next best does not equal good. I think a lot of people overemphasize next best, the winner is the best, it doesn't necessarily mean it's good. So really focusing on how we use our data, how we can stitch together these experiences, really provide that next best. A genuine good experience for our customers is really critical. Going back though, just to actually talk some OKRs and some targets, we have obviously set those out, I've mentioned ADM a few times now, setting percentage thresholds with how many models and how many times the predictors need to pop, we are doing head-to-head tests in the ADM models as well, just to show does one model perform better than the other? Obviously we can then calculate the business performance increase off the back of it. So we do have our internal OKRs, which will be held to account, but obviously also just really focusing on that customer experience as well.

- Thank you so much for the... So to close it out, I think a great topic would get your take on the powerful between behavioral data, customer context and real time decision moving forward with projects. Let's hear.

- Sure, sure.

- From a market perspective.

- Yeah. Next better kind of better action doesn't sound as good as next best-

- Yeah, it's not actually.

- Though, does it? No. So yeah, the way I think about it is, you know, in thinking about the event, again, the, you know, behavioral event data and the curated view, I think that, you know, that those two need to be combined into a profile that, and we call it now in CDH our customer insights cache, that can be decisioned on in real time. The analogies I'll use is kind of a one that's, you know, on everybody's mind today, GPT. Think about GPT, think about how you can ask it questions and it can give you really very deep understanding of, you know, the answer. But it's learned only from 2021 data back. So it's got this gap. It can't tell you about what has happened in the last two years. Imagine that with your customer. You need to be able to have that curated, knowledgeable view of your customer that's over years of their interactions with you. But you also need what happened, you know, a week ago, a day ago, more importantly, an hour ago or a minute ago. So the ability to combine that long-term memory of the customer with the short-term memory, and do that in a way that you can then immediately identify that customer effectively and immediately decision on them and provide them with a next best action is the way I think about it. Yeah.

- Thank.

- We just heard from the master, I don't know what I can add to that . We should have put Vince last. For me the bringing together of what customers are expecting from us now and how they interact with us now is so different and it's so immediate, their expectations and you know, in terms of they think that we should know everything that they've just done and there's probably no reason why shouldn't. So bringing together the sort of collective memory from our CDP together with the kind of more static data and this more static memories that we've got within CDH really gives us the power to be there in the moments that matter for customers and help them when they need it.

- Again, I'm just kind of building on Fiona's unfortunately.

- She said what he said.

- Yeah, we'll just keep going.

- Yeah.

- To try and build on it though, obviously I've touched on latency, I think there is a lot in here where you do need to spend a lot of time with your data and understand how fast is fast enough, to actually understanding how that experience translates off the back of the data. So having that customer experience front and center and really focusing on what their needs are, how can we enable that as a bank? And then obviously we've got the omnichannel play, which somehow we've gone through a full 30 minutes here without mentioning the word omnichannel as well, which is amazing. So we're in some significant channels with our CDH deployment, mobile app, online banking in the branches and some of our telephony centers as well. There are still channels not on that. So we have to also acknowledge how do we move data, think you alluded to it, Vince, can we move our decisions also out of Pega and make it available for other systems to use as well. So again, focusing on that customer experience, how do we make sure it's all tied together no matter where they turn up? You know, the pandemic changed the way our customers interacted with us. That's slowly changing again now with everything having reopened. We need to be in the right place at the right time for our customers.

- Well thank you so much for this very insightful exchange, despite some of the initial technical difficulties for that. And with that I'm sure our audience...

- Any questions? There must be some. Somebody's got a question. There we go.

- You can shout it out.

- We can hear you.

- I'll shout, okay.

- Go for it.

- We can hear you.

- I can go, the one I mentioned around the recent successes in our short term borrowing space was a bit of a surprise because it was a complete mindset shift for our managing director who was a fan of Pega and he got a lot of commercial value from the work that Pega delivered. And then the CDP came along and I think he thought, I'm not too sure I can get an awful lot from this, but once we shared a lot of the POC outcomes with him, he was kinda all for it. So much so that he'd asked us to create a whole new way of working and dedicate people to just picking up new signals and thinking about what they can do with them. And we've got a whole kind of train of work now, just burrowing through all the backlog of AEP ideas. So that was a nice surprise. One of the other surprises that probably isn't as nice, is I think you kinda alluded to Vince around what people think a CDP can actually do. And of course in that way certainly you have very much, there's some people that are wedded to CDH and Pega and what it can do, and there are other people who are wedded to the digital side of things and definitely wear their Adobe cap. And I think we have chosen Adobe as ours, and the sort of surprising piece was people thought we should be able to choose one or the other and just go with that. But to your point, neither of them can do everything well, and for me it's always been about how do we work them together to make them sing as a community of MarTech rather than, you know, just one piece versus the other. So I was kind of tired of those conversations around the one versus the other, but I think we've got the organization to a place that it understands certainly internally, why we need them both and what they can do for each other.

- Any other questions?

- CDP specific? Pega CDPs?

- I think for NatWest, I can go for NatWest. We've obviously been having, we've had a CDH for since 2014 I think. So the organization was quite in the way of using AI to deliver commercial value already. And the element of trust was definitely there with our team. Had it been 2014 when we were looking at CDPs and all that's kind of unrealistic, but if it had been a while ago before we had a CDH and we had gone through that journey and built the trust up and the history up there, I think it would've been a lot more difficult. But for us, it was such a big spend bringing in a CDP, we had to have something quite tangible to share. And that proof of concept that we built, we did a three month proof of concept to demonstrate the value of that, was really key for us getting the exec to put their hand in their pocket and actually buy it in the end.

- I'll add one thing and what I love about, you know, the story from Fiona and Tom is that they both were able to get decisioning up and running without a CDP initially. So they're seeing the value now though. So they had some, you know, initial value to already build on and you know, if anybody has used CDH, they know CDH just loves good data, decisioning loves good data about customers and it'll prove very quickly whether that data is good or not, right? In the adaptive learning and in the, you know, ranking of the decision. So, you know, that I think is in stark contrast to some other clients that I've seen that have done it in the reverse. And they've tried to get all their data house in order and I literally, I won't mention the client, but literally there was a client that has spent three years trying to get their data house in order and get a CDP up and running, meanwhile getting no value out of decisioning at all. So I think one of the things that will help with, you know, selling the whole kit and caboodle, right? It's not one or the other, is getting success early so that you can build upon that.

- And I think, sir, you had a question as well. Yes.

- Sir. Okay, holding me, you're keeping me honest. Well, keeping Rob honest, well and me. So yeah, you're probably referring to back then what we called CDP choice. So there was some integration and actually we have a booth in the innovation hub that specifically is about the integrations that we're doing and the CDP choice story, which is the story about being able to connect virtually to any CDP using the methods of data ingestion that we have through what we call the customer profile designer. So yeah, I mean I think that maybe some names have changed in terms of what we called it in the product, but those features are continuing to evolve.

- Always on insights.

- And always on insights is the overall umbrella on everything, yes. Thank you.

- Any other questions? Oh.

- Sure.

- Well, when we think about empathy, we really think about making the right next best action under the circumstances of what the customer would really want, so the propensity score being, if the propensity score is good and it's not messed with, there's ways to mess with it in CDH, you can use levers, you can use your own value calculation, which may, you know, not represent the value very well, we've seen that, so we think of empathy as really listening to the data about the customer and using that to produce the propensity, the adaptive models, and then letting that actually drive the next best action. So when we talk about empathy scores and we have a tool called Value Finder that will help you see whether you're using MBAs empathetically or not, they're looking at how much you're levering things and how much you're changing what we call the P times V times L calculation. Does that help? Well, the human could get in and you know, lever things. So that would be the human in the loop. Yeah.

- The only interesting part on that as well is we talk about propensity a lot, the P in that calculation is what is that even learning on? If you speak to every other CDH customer in here, someone will be learning on click, some will be learning on conversion, there is different ways for your models to be trained and dispositions you'll be feeding back into your engine so actually really understanding that before you touch any of the levers, any of the values which we try not to, I think that's critical.

- Yep. Absolutely. Absolutely, so the whole idea of an engagement policy, having, not just eligibility but applicability also is all about empathy. The applicability is, you know, for instance, you know, even though you can offer to this the customer, you know, can the customer actually afford it, right? So that's empathetic if you're taking that into account in your engagement policies.

- Vince, I think you said that you recommend starting with the decision engine and then adding the CDP but, and not waiting for the CDP in order to start decisioning, if that's the case, how do you explain, you know, you talked about getting quick wins on decisioning, how do you explain the results if you don't have the CDP to, you know, provide that level of granular data?

- Good question. I think that, you know, when I say that I think you still need organized data, but you have to have, if you don't have organized data, then you've got a problem, right? So you still have to have that and as we said, you have to feed the results of what's happening in next best action back to analytic platforms. That doesn't have to be a CDP, CDPs are also saying they've got analytic capabilities and reporting and such, but many big organizations already have those capabilities somewhere else so as long as you can get those results back to those analytic platforms, then you might not necessarily need a CDP.

- Could one of you give a overview of the platform? Like how did you set up a platform that is robust enough to handle the massive amount of data? So we all understood about the enrichment, arbitration, propensity, everything, and you're connecting with the different analytics data shows and pulling the massive amount of data, but how do you ensure that the next best action is time sensitivity, contextual, personalized, but in real time, right? One of the key challenges in my past experience is data matters and if the data is not coming from real time, then you're gonna miss the next best action for that personalized customer. So could one you share your experience, what problems you had getting the real time data archived and the platform, how do you scale up the platform to support that?

- Well, I think both of these two have great experience with scalability.

- I kind of alluded it to earlier, like understanding your data's key, like knowing what latency is acceptable can actually help you on that journey, right? A mortgage, depending on which country you live in, takes nine to 12 months, right? There is an acceptable latency when you're talking around mortgages usually in a product. A credit card, you usually engage with a person two or three times in 24 hours. And I know I'm being very financial services specific here, but actually being able to categorize your data and understand the acceptable latencies, then key with how you even engage with your partners. So sitting down with them and saying, for me, like this needs to be as real time as possible, these elements, then I can actually work on the platforms, work with them and say, this is the data we're going to move across, through whatever connector it is ultimately, whereas a significant amount of our data still move batch nightly. And that is acceptable 'cause we understand what the customer experience is behind that, what it's driving and what it's doing. But having that overview and that understanding of your data model and what latency is acceptable, and then dealing with your internal partners as well who are probably responsible for delivering that, that becomes critical. And I think you do ultimately need some real senior buy-in to be able to drive that and push it. Because even though we were saying that you can do CDH without the CDP, which we did, trick question by a Wells Fargo employee that by the way, we always looked at our latency, we always had significant data coming from our internal systems, whether it's the data lake, just even IH and the data Pega produces. So I did it that way, I understood the need and the speed for each data element and what segments had to come through at what speed. That way you don't impact the platform straight off the bat by saying everything needs to be moved in at speed.

- Thank You. Appreciate it.

- Yeah, and I'll just add that I think that's true of channels too. A lot of times people, you know, we promote that we can make decisions in two or 300 milliseconds, but the reality is there are some channels where, you know, you might not need 200, 300 milliseconds. So looking at it from those lenses is really important and when you do that, then you're gonna be able to, you know, figure out your scalability and tune it appropriately.

- Any last questions?

- Great questions too, thank you.

- I know. Excellent. Well, thank you. Oh, one more, fine, please.

- For a company considering bringing this in-house, what advice would you have for them in preparation for what is almost the expectation of instant customer service and this instant way that we're now doing, like what advice would you give to prepare for that shift?

- And what are you bringing in-house? A CDP?

- Yes. A product.

- Well, I think if you're bringing something like that in-house and meaning that you're not gonna use it on the cloud, you're gonna have your IT group build a CDP? I'm assuming that was what you were alluding to? I think that's a big endeavor to, you know, again, think about the agility of your IT partners to be able to provide you with, you know, the change management you're gonna need in evolving that repository over time, because it's, you don't want a sort of a big bang approach to trying to build it. 'Cause you know, we've seen those projects take 18, 24 months or longer. So I'd be very, you know, inquisitive with your people that are gonna help you build that to make sure that they're set up to be agile for you.

- I think a real challenge is they are at the end of the day two separate standalone products as well. So skill sets, people, process, backlog management, all of a sudden you have two products, you're managing the backlog for each, but there also needs to be overlap depending on what you need outcomes for as well. So just I would really think people and process skill sets are hard to come by still. Focus on those areas and make sure you have the right things in place for quick success. Obviously if you have a longer term, you can survive.

- Well, thank you so very much. This turned out to be a quite the lively conversation and very insightful. I hope you all have enjoyed the session and some good learnings too out of it. So I hope you enjoy the rest of PegaWorld too. Thank you.

- Okay, thank you.

- Thank you.


Product Area: Customer Decision Hub Topic: AI and Decisioning Topic: Customer Engagement Topic: PegaWorld Topic: Personalized Customer Experiences

Related Resources

AI in Innovation

Unleash the power of AI innovation

Learn about enterprise AI and how it'll drive the future of business outcomes.

Why Pega

Why Pega?

Uniquely powerful software isn’t the only thing that sets us apart.

Share this page Share via x Share via LinkedIn Copying...