AI Reshapes Marketing Through Adoption and Governance at EY

Dylan Cosper:

Hi, I'm Dylan Cosper with the Infosys Knowledge Institute. Today I'm joined by Cecilie Burleson, MarTech leader at EY. Cecilie, thank you for joining us.

Cecilie Burleson:

Great to be here.

Dylan Cosper:

I want to start kind of generally how you're seeing AI change client marketing organizations.

Cecilie Burleson:

AI has been just incredibly impactful across organizations and you see it continue every month. I mean, there's something new coming out, something significant. So with organizations that we work with at EY, we're really seeing it accelerate the workforce. So specifically around marketing, that content acceleration to development, the different types of variations, workflow automation, how do you kind of bring that to life and get through approval processes quicker with accelerated speed. And then that speeds the sort of insights to make actionable strategic decisions and optimization.

Dylan Cosper:

Since you're client focused, I'm sure you encounter a lot of interesting situations and probably a diversity of situations around AI, be it maturity or things like that. So what kind of surprises have you seen with clients around AI marketing?

Cecilie Burleson:

I think it will vary, of course. A lot of things that we see with AI is really more of the adoption and change management when it comes to some of those hesitations. Well, I'll take a step back here. So adoption and change management is one, but when they're actually starting to think about AI, there's a lot around privacy transparency regulations, that there's a lot of legal just stepping stones and kind of processes to go through to ensure like, hey, are we safeguarding our data? What are we actually doing? Are we getting good outcomes from it? Is it trustworthy? And then once that starts to be a barrier that's overcome and they start to implement AI, it becomes, hey, how do we actually roll this out and communicate to the organization that it's available and folks can use it? And so really then getting the adoption towards it and having people actually use it in their day-to-day becomes where some of the struggle and where some of the AI implementations can fail.

Dylan Cosper:

Okay. So really focused around just actually using it.

Cecilie Burleson:

Yes.

Dylan Cosper:

I mean, that brings up a good point around experimentation. Have you seen some approaches that are maybe, have better outcomes when it comes to experimentation in your work with clients so far?

Cecilie Burleson:

Absolutely. I mean, there's a lot of AI initiatives that really start with more of that POC or that kind of experimentation phase and they just test out really, some tend to be really aggressive and then some tend to be more foundational and small. I've seen more success with kind of taking that foundation and building the building blocks around it and really communicating with those groups that are going to be involved versus, hey, we're going to look at something that's a POC, that's a bigger milestone and actually take a lot longer to implement, take a lot more of that kind of change management that I had mentioned to really get people to change where there's a little bit more struggle in the adoption there.

Dylan Cosper:

Okay. So kind of shortening the distance between having the idea and then getting it into people's hands as opposed to these big large scale kind of POC projects.

Cecilie Burleson:

Exactly. And you're really looking at how do people actually work now? What are they actually using for AI personally to then say, "Okay, if we're going to adopt something for the enterprise, how do we make it an easy stepping stone for them without completely upending their world just overnight?"

Dylan Cosper:

Yeah. With clients, what are kind of some of the most important decision points or decisions that you're seeing that they should be making right now so that they can set themselves up with successful AI adoption in the future?

Cecilie Burleson:

So part of what I just mentioned, really being able to understand how the workforce is using AI right now. So a lot of individuals within organizations are already using for personal reasons or just their sort of regular day-to-day. And so really kind of taking a step back as an organization and identifying, hey, what are some big use cases or some use cases that we can actually foundationally start to put into play, knowing what some of those around our organization are already doing, and then what's also going to align with our business objectives as well so that we can get leadership buy-in? So how do you kind of mirror that kind of macro high level with what people are doing day-to-day and bring them together so that you have buy-in from both sides and start building to more of a common ground?

Dylan Cosper:

We hear a lot about the efficiency that's coming from AI, decreasing the time from ideation of a campaign, a marketing campaign to the actual implementation of it. What kind of value are you seeing clients get out of AI? And ideally, I'd love to hear kind of beyond the efficiency and productivity, which can kind of sometimes be a softer, squishier target. Are you seeing some very high value being driven by some of your client organizations?

Cecilie Burleson:

Absolutely. So when I mention things around privacy, so one of the significant things, there's content at scale, but when you start to think about the squishy, oh, we're producing more content. But when you think about it, especially with certain organizations that have high regulations or maybe IP, there's a lot that has to go into getting all of these approvals. So things that could take weeks, days, weeks, even in some rare instances, months to just go through because there's all of these different players that need to approve it, you can do it a lot quicker with AI and you can almost generate this content and asset in mere seconds that's closer to the compliance approval. It's not always 100%, but a lot closer to reduce those cycles in time.

Dylan Cosper:

What are some of the critical challenges that you see clients likely to face or currently facing in the next six months with AI adoption?

Cecilie Burleson:

The challenge becomes that there's so many things out there and there's so many possibilities. And so the market is very saturated with a lot of AI organizations that want to come. The existing organizations that they work with, especially enterprise tools are also in there saying they have AI solutions and then you have sort of like, how do we consider building in-house? So there's just a lot of noise that people have to kind of navigate through and figure out what's the best solution for them and how does it actually align again to their business objectives and what they actually need versus buying the shiny thing, right? This is going to be cool. We're going to just go from zero to 1000 overnight. It really, again, is the stepping stone that's really going to make a difference to the organization and the people that work there.

Dylan Cosper:

We did some research this last year, and I think one of the things we uncovered is there's thousands of new marketing tools every single year. Well, now all of those marketing tools that are coming out every single year claim at least to have some sort of AI capability, further muddying the water of what do we need, what should we pursue? So how would you recommend marketing leaders identify what is the right tool? Is it something that kind of goes back to defining the problems they're trying to solve? Or in your view, what do you think they should be doing?

Cecilie Burleson:

It is. It's really what kind of problems are they solving? What are the use case? What do they actually use now? Because implementation and integration is going to be very important with their existing systems, with how data flows. Also, how comfortable are they with understanding, hey, if we're sending our data over or sharing our data, or is it something in-house, like how do we start to really trust what's coming out of it as well, and how do we train it to continuously adapt to our needs? So there's a lot of things to consider when looking at the different solutions to come in, but there's always foundationally a good place to start, and then you evolve it based on your use cases and your needs.

Dylan Cosper:

We've talked about kind of the value, we've talked about the challenges. Where have the disappointments been, at least in your experience and work so far? Have there been any big AI disappointments that you've seen?

Cecilie Burleson:

Maybe in some of the tools I've tried, but it might've been just lack of maturity at the time.

Dylan Cosper:

Content creation through AI I think is one of those that's really proven to be a boon for a lot of organizations, right? Are there any use cases that you've seen that have been particularly disappointing, at least in AI marketing?

Cecilie Burleson:

Originally, I'd say when we kind of really started to see the acceleration, there was disappointment. Things were like, that hand doesn't look right or that foot doesn't look like, or you can definitely tell it was AI generated. It's accelerating so fast and it's building off of its learning that you're only going to see it get better. So I think it's a lot of learning from the things that may have been disappointment early on. The part I will continue to always just be cautious about is just the hallucination, right? Ensuring that the outputs that you're getting are accurate and just having that human intervention and sort of that hybrid approach to ensure that you're actually getting quality output that you can be confident in.

Dylan Cosper:

Okay. And so I mean, going back to the disappointments, it's almost not a disappointment in the use case itself. It's a disappointment that we're not able to do it just yet kind of. It's not fully there. You're mentioning the issues with the hands or little errors here and there, but yeah, we'll get there.

Cecilie Burleson:

I think we'll get there. And it's also expectations. I think personally for me, I didn't have necessarily grandiose expectation that it was going to do everything at once. And I think that's just our culture is... With everything at our hands, our fingertips, we want things instantaneous. And so while I might fight with some of the tools, I also have to recognize some of it's probably user error in some of my prompting. But again, I've seen it evolve to, hey, either I prompt it a lot better or the technology has started to evolve to be able to look at, understand what I'm asking, take previous conversations to heart and actually build upon it as well.

Dylan Cosper:

Are there any ways that you see how AI is really going to change the way the effectiveness of marketing is measured?

Cecilie Burleson:

Oh, absolutely. A lot of organizations right now look at measurement and sometimes I think AI will actually help to accelerate on KPIs that actually matter, things that are going to move the needle drive some change, really being kind of more data-driven, data purpose versus some of those vanity KPIs that we've typically seen where I know impressions and clicks has some value depending on where they are in the funnel, but oftentimes you tend to see that where maybe you want something that captures a little bit more detail, gives them a little bit more insight. And so with leveraging AI for just measurement effectiveness or campaign effectiveness, marketing effectiveness, you're really going to start to see value in being able to lock that down. I mean, even at this conference, we talk about being able to look at performance marketing, which it's a little bit easier to measure, but then there was a discussion from the CEO that was like, "No, it's actually like, now how do we take this and look at brand, the impact of brand, the influence of brand?" And I think having this will enable to have more data-driven insights.

Dylan Cosper:

What's your advice for folks as they continue on their AI marketing journey?

Cecilie Burleson:

One of my key advice I would say is, again, anchoring on, hey, what are people doing within the organization already so that adoption will be easier? And then anchoring again to business objectives. So how do you get the leadership buy-in and then the folks that are actually going to use it to kind of cohesively work together? The other thing not to be undermined in any way is just, again, the privacy, regulations, anything dealing with that and the transparency continues to be so pivotal. I mean, while a lot of things are advancing and people are like, "Let's start pouring our information in there and seeing what we get out," you have to trust what you're actually doing. And the folks that are getting the output back from it or benefiting from it, even our customers need to be able to trust that you're being responsible with information. So ensuring compliance, that adhering to regulation is absolutely key, and I think we'll continue to see that theme as things continue to progress.

Dylan Cosper:

Well, excellent. Well, Cecilie, that's all I have. And thank you so much for sharing your insights and some of your stories here today. Thank you for coming on the interview.

Cecilie Burleson:

Thank you for having me. It's been a pleasure.