It’s no secret that data clarity underpins all AI optimisation. However, this data can often appear complicated and in silos. At Pegaworld 2025, FutureWeek sat down for a conversation with Tara DeZao, Product Marketing Director at PEGA, who explains why great customer engagement depends on these silos being broken down, and data being orchestrated across all departments, as well as delivering timely, relevant, and respectful interactions at every touchpoint.
What Marketing Challenges Do You See AI Solving?
Customers have issues around unifying all of their channels, and being able to push data to the right channel at the right time, in real time. Martech stacks are so disconnected and bloated that if you don’t have AI at the centre, you’re going to be over-serving messages, or serving irrelevant messages. I once had a client serve the same consumer different offers in different channels, and that was bad for the brand’s reputation. The second piece is hyper-personalisation. I hate the term hyper-personalisation, and I think that most marketers are kind of sick of hearing about it. But the ability to serve the right interaction at the right time, in real time, is the ultimate level of personalisation.
Can You Give an Example of What This Personalisation Looks Like?
Let’s say you’re booking travel, and then you just skip over and start doing something else. It’s no longer OK to talk to your customer the way you would have 30 minutes ago. You need to be right there. If you’re in a coffee shop, and you leave the coffee shop, and then the brand texts you and offers you a free croissant. That’s not a great customer experience. Instead, you could make a fan for life if you had real-time data in the moment. Because of adaptive AI, our technology actually re-decisions based on your signals. We can help brands pivot right away if they need to move into another channel, if they need to not send an offer in that moment. Every interaction shouldn’t be a sales interaction. It should be a mixture of nurture, retention, and services.
Do Brands Need to ‘Value Add’ to Avoid Being Seen As Invasive?
It’s a critical piece. Gartner put out some numbers last year about no personalisation, no help, personalisation without help, help with no personalisation, and personalisation with help. Help without personalisation was more desired than just the personalisation. I think it’s becoming something where we live in such a fast-moving, fragmented society that the customer journey has to be populated full of value-added moments for customers.
What is Unique About the Pega Product?
I would say the power of the adaptive AI, but also the fact that it’s an end-to-end solution. I have a slide in one of my decks from a report from 2023 that said the average martech stack has 77 applications, but their marketers are only using 30 percent of the applications. For example, we have a built-in outbound email tool. You don’t need to have an ESP, or you can bring your ESP to the party and connect it to the Customer Decision Hub and then send real time interactions through that ESP.
What’s unique is its interoperability in-between applications and connectivity, because you can just have the Customer Decision Hub as your full marketing solution, or you can bring your existing investments to the party. It’s a really good interplay.
What is the Best Approach for Utilising AI to Build Efficiencies in Your Marketing Process?
Firstly, you need to audit your Martech stack and understand how your tools are performing – who’s using them and how they’re using them. We have a lot of clients that have technology in there, and maybe one person is using it, but they have 35 licenses. Something that’s really important is to understand what your company policy is around AI, and to get marketers really comfortable using it.
I think there’s a lot of fear and uncertainty on the part of marketers around AI, because a lot of folks are worried that they’re going to lose their job, but it’s actually a partnership. If you make AI and human labour partners, that’s the best way to approach it. It’s going to help you make better decisions for your customers. It’s going to help you remove manual complexity from some of your processes. It’s going to give visibility for compliance and data privacy folks. It’s really about articulating it as a partnership versus a replacement technology.
What Are Some Ways You are Instilling an AI-Positive Attitude Internally?
On my team, I de-stigmatise using Gen AI for things like copy. De-stigmatising that for people and educating on what’s available. Our marketing team does a great job of putting on workshops and making those available to anyone in the company that wants to come and understand how to use different things. There is a lot of good self-service type education.
One thing with marketing that has historically been true is that you get the access to the things you get access to, and then you’re not really supposed to create anything like the design and brand teams do. I think enabling your marketers to make some of their own stuff with Gen AI is amazing.
Does Marketers Having Ownership Over Their Work Create a Positive AI Culture In Teams?
It’s a big part of it. I’m a writer, and I’m really particular about what my name goes on. I make sure that I write all of the stuff that I put out. One of the things I have trouble with a lot of the time is brainstorming angles – Gen AI helps me with that a lot. Making people comfortable with using Gen AI is one part of the process.
What Challenges Does Agentic AI Address?
Currently, agentic AI is along the CX line. In a customer service or preemptive service framework. One of the things that we talk about is taking data and ‘extracting the signal from the noise’. Agentic AI is going to be good at helping marketers with that – slicing and dicing and getting what we want out of the data they have. ‘Performance’ has historically been terrible for marketers. It’s very disconnected and siloed, and there’s very little closed-loop marketing, unless you’re in a direct-to-consumer organisation where you send someone an ad, they click on it, and then buy something right there.
That’s not most marketers’ experience. I think agentic AI is going to help marketers better understand what their customers want and need. We’re going to have the customers who only want to do things online, never want to call, never want to talk to a human. But we’re always going to have those people that do want to talk to a human. So I think it’s important to not fear the future.
What Role Do People Play in Agentic AI?
There will need to be human oversight. There are certain tasks you can leave to agents, and there are other tasks you will need to have human oversight over what the agents do. At PEGA, we talk about responsible AI a lot. One of the things that you don’t want is bias in your AI algorithm. There are certain fields that marketers collect data on that have the potential to become biased. You need to make sure that even if you are using AI or an agent for a process, that your bedrock or your data foundation doesn’t have bias in it, because that’s where the bias really originates from. And then anything with brand reputation or brand safety is a risk.
If a salesperson is doing research on a target, and then asking their AI to craft a communication. And instead of crafting the communication, the AI crafts and sends the communication immediately. That’s not something that you want the AI to do. It’s important to always be checking and making sure that your models are adapting, and your data doesn’t have biases in it, and that AI is exposed to various situations and trained on the right things.
What are the Main Data Factors Marketers Need to Pay Attention to?
Data freshness is always a consideration – making sure that it’s as fresh as possible. So many organisations say that they’ve mastered personalisation when they haven’t. Many organisations say that they don’t have data silos when they actually do. So, making sure that you don’t have data silos and visibility to as much data as possible across the organisation.
Contextual data is the key to creating a great customer experience. I use this example a lot: if I’m having some sort of a medical panic and I’m on WebMD, I’m not going to pay attention to you if you try to interact with me, because I am in a moment of fear and panic. I think really being data driven and paying attention to the signals, a lot of organisations put their thumb on the scale of their AI.
They don’t let it actually do the work. They give it rules and restrictions more over-the-top than just risk mitigation, and you really need to make sure that your data is available in the right place at the right time, and that you’re letting your AI really listen to that data and learn from it and adapt.
Why Do Companies Keep Their Data in Silos?
Data silos aren’t easy to break down – it takes organisational transformation and agreement from the functional areas across the business on your overall data strategy. I think your marketing is going to be how your organisation is. If you have a top-down approach that’s transparent and seamless, that’s going to be great for your marketing team. If you have an organisation that is protective, where each functional area is protective over their own little kingdom, it might show up in your marketing.
For example, if you’re an organisation that has a million product lines, and each one of those product lines is not in competition with, but in competition for, resources, you’re not going to have the holistic nature of an audience experience. Because if two teams want two different things, which team gets what they want in that situation? Whereas, if your data is connected, and you can see what decision will drive the most traffic, you can pick the interaction that drives the most value. The data says it all.
