Media Monitor’s Conversation at Cannes| Why AI Still Needs a System of Record| Drew Kane

AI may be reshaping media planning and buying, but not every business problem requires an AI solution.

In this episode of Media Monitor: Conversations at Cannes, Guideline Chief Product Officer Steve Silvers speaks with Drew Kane, Chief Product Officer for Prisma at Mediaocean, about media workflow automation, agentic buying, and why trusted systems of record become even more valuable as machines make more decisions.

The conversation begins with Prisma’s role as a financial, contractual, and workflow system supporting the media industry from planning through payment. Drew explains how Mediaocean’s wider portfolio—including Prisma, Innovid, and Protected Media—helps agencies and advertisers manage creative workflows, buying, verification, and financial accountability.

Steve and Drew then discuss the new integration between Guideline’s MediaTools and Prisma.

While information has moved between the two platforms for years, the new API integration reduces manual work and helps keep planning and execution data synchronised. Objectives and budgets can flow more directly from MediaTools into Prisma, with actualised results moving back through the workflow.

The result is less friction, fewer manual errors, and more time for teams to focus on business outcomes.

Their discussion then turns to AI.

At Cannes, many companies are trying to apply AI to nearly every workflow. Drew argues for a more disciplined approach: begin with the business outcome, then choose the appropriate technology. Sometimes that means an AI agent. Sometimes it means an MCP-enabled workflow. And sometimes a deterministic API is still the better answer.

Steve and Drew also examine the emerging world of buyer and seller agents, where systems could identify inventory, exchange campaign requirements, and accelerate media transactions. But as the number and speed of automated decisions increase, so does the need for governance.

Someone still has to track what was approved, where the money went, what changed, and whether the advertiser received the expected value.

That is where systems of record become essential.

In this conversation:

• What Prisma does across the media planning-to-payment workflow
• The partnership between Guideline and Prisma
• How the MediaTools and Prisma API integration reduces manual work
• Why business outcomes should determine the technology used
• Why not every form of automation needs AI
• The difference between APIs, MCP connections, and AI agents
• How buying and selling agents could change media transactions
• Why agent orchestration may matter more than individual agents
• The continued role of human approval in automated workflows
• Why auditability becomes more important as AI scales decision-making
• How systems of record track media spend, approvals, and outcomes
• The convergence of linear television, digital video, and broader video investment
• Why media efficiency must ultimately translate into effectiveness

Media Monitor: Conversations at Cannes is a special series featuring leaders across advertising, media, measurement, and technology discussing the ideas shaping the future of the industry.

If you’d like access to the benchmark report or want to suggest a topic for the next part of the programmatic series, reach out to press@guideline.ai.

If you enjoyed this episode, be sure to follow or subscribe so you don’t miss future conversations on advertising, media strategy, and cultural marketing moments.

And if you’re listening on Apple Podcasts or Spotify, a quick rating or review helps more people discover the show.

Transcript

SPEAKER_00 0:04

You're listening to Media Monitors Conversations That Can, a special series that we're dropping just this summer. Thanks for tuning in to Media Monitors Conversations That Can. You're about to hear our Chief Product Officer Steve Silvers interview Drew Kane from Prisma.

SPEAKER_01 0:24

Hey guys, welcome to Media Monitor from Cannes. I'm Steve Silvers, the Chief Product Officer at Guideline. And with me today is Drew Kane from Prisma. Drew, tell me a little bit about what you guys are doing over at Prisma. We're doing a lot of things at Prisma. We're doing a lot of things at MediaOcean.

SPEAKER_02 0:38

I'll make a quick, if focused on who we are. Many do, but from a Prisma standpoint, Prisma, think of us as for media system of record financial contractual workflow, plan to pay. Plan gets interesting, but $200 billion in media spent annually. Second piece of us is innovative. Think of that as creative workflow and optimization and protected, which is all about trust, you know, viewability and verification. So together, we are really servicing the advertising and media industry in all capacities. From the Prisma side, hugely exciting things going on. Of course, AI is going on. No, no, that we'd we'd have to set up by default. We are hugely focused on the value we bring to our customers. I'm our chief product officer for Prisma. I like to optionally insert chief value customer because everything we try to do, we're delivering value. The value is around automation, workflow, convergence. Clearly, there is an AI and ejective components to what we're doing, and we're doing it on all levels and running on all simulators.

SPEAKER_01 1:36

Well, obviously, as a product guy, I love thinking about that chief value officer. I like that. CVO. Yeah, yeah. We should, we should, we should, you know, I'm gonna I'm gonna talk to my boss about that. So that, you know, obviously, or maybe not so obviously, you know, Guideline and Prisma have a very strong partnership. We just announced a new integration uh that Prisma has with Media Tools, our media plan management software. Why don't you talk a little bit about what that integration is and and why you you think it's important?

SPEAKER_02 2:08

Sure. I'm I'm I'm more than happy. That integration, again, I I mentioned that we are we're plan to pay. Um but planning workflow and aligned adaptivity is happening inside of media tools or inside of guidelines. This is taking what over the years has been a semi-automated integration where there's manual steps to kind. Steve, jump in where you feel like I'm doing you wrong or doing you not right. But but it's really about saying how can we make the use of APIs and a more direct communication between our platforms to enable at an objective level, at an actualization level, and really in between get to a place where we can better keep our systems in sync. Uh what you do inside of media tools, for example, objectives for local and national, or even budgeting, that directly is, if I think of you as the planning, we're execution. So planning flowing into execution, that flowing in and tying a bow on it at the back end for actualization is huge. And to do it with immediacy, to do it with a level of automation that hasn't been there is is where I really see we'll use that word value, maybe about 15 more times, but that that's real value. And it's coming at for what what folks have been asking for, which which I also appreciate.

SPEAKER_01 3:18

That's right. And obviously these this data has been moving around between our systems for for years, but it's been a manual process which takes time and is error-prone. And now we have an API integration that allows us to move the data back and forth. I like to use the word automagically, not really a word, but it moves back and forth automatically and now has removed a whole bunch of friction so that people can focus more on getting the results out of their life.

SPEAKER_02 3:45

When we were doing our roadmap this past year, uh went with this theme of we said we will have an impactable roadmap. And an impactful roadmap is it has to have impact for our customer or respective partners. It is practical, actually, can can act on it. So you've got, you know, you know, that piece, actionable. And then we used magical. We said each one is contained inside of impactical. So we're on the same wavelength there.

SPEAKER_01 4:11

Well, we're gonna just keep making up words between the two of us. I I love the product on product pods. It's so much more fun than some of the other ones. So look, obviously, you know, we're here at CAN, and you know, this year, like the last couple of years, AI has been a huge topic, and everyone is almost trying to out AI each other. Um, but in some cases, you know, AI is an overcomplication, right? And sure, you know, in this particular case, you know, we could have looked at an AI type of integration, an MCP to MCP injected connection between our platforms. But you know, in this case, it was just much simpler to do an API, and we get that deterministic data flow, and we keep it we keep it simple. Now, obviously, on each end, there's going to be AI optimization and other things on both sides of the uh of the flow. But this one was simpler to keep, but it's just an API.

SPEAKER_02 5:07

100%. Well, you know, I the the lens that we will look at this through is first of all, AI, I've been saying for a while, you know, it's always included. We're always included, right? It's it's in every conversation we have. When I look at it, it's the undercurrent of what we're looking at, but it is right application for the right technology. This still starts with a business outcome. When we talk about the business outcome that we're looking for, where it's really is truly a sharing and sharing of data that's going in two directions, like we can accomplish what we need to through that. So you don't have to throw an AI or MCP layer on top of it. I I think that at the same time, what you can do with MCP, I it's turning integrations into actions or outcomes. So certainly there is opportunity for that, as well as what you said, Steve, that if you have agents living on both sides of it, like have at it. But I would always start and first start with what's the business outcome you're looking for? How is AI the tool for getting there? That gives me automation. That gives me optimization, that gives me some level of execution. That lets me take huge data sources and bring them together and associate them in ways that I could not. But it doesn't have to start with that. Not all automation is AI, and it's it's it's gonna continue to be our credo. We're not gonna do solution looking for problem. We're gonna solve for the problems that are up there. One product works into another.

SPEAKER_01 6:26

So we when you take a look at the talk tracks that we have in CAN, you know, AI is typically focused on a creative creative, which obviously with innovate, you know, media is involved with. But then also in this agenda buying sort of concept that the IB Tech Lab and ATCP are putting out there. I'm just wondering how you guys are thinking about those technologies and how we could plug in together to those types of buying workflows.

SPEAKER_02 6:57

Sure. To take it up a little bit, you can look at it individually. I'm gonna create an agent, I'm gonna create an agent. Really, when we've been looking at through both an Innivid but a media ocean lens, it's been very much around not individual agents, but orchestration of those agents. So we are investing in on both the Innivid and Prisma side on individual agents, but we're looking at it at a higher level. If you take an agent which is saying, I'm directed towards this outcome and I'm directed towards this, having an orchestration layer that lives above that is more than it's it's it's a must-have. And by layering that in, we have that. We also just announced NIVO, which is an intelligence layer that's living above it all. And that is really to inform intelligence, not just LLM or promptible, but really to be able to both have insights and actions. But it takes the shape of, okay, now I have this agent. Specific to Prisma, I knew I took a little while to get there, but specific to Prisma, we are really taking a very agnostic approach. So we have absolutely been working on Prisma buying agent. We have looked at it, the first lens is set simply about inventory and surfacing inventory. How can we do that in a way that's more fluid and more immediate than what you see today? So I'll use in the context of guideline. Guideline feeds us objectives. We, those objectives are then turned into, well, this is what I'm, this is my objective, this is what I'm looking for. That from a buying agent can go out to a seller agent with campaign details and other information that returns essentially this is inventory that can satisfy what you need, which is you know, taking a process of I'm reaching out to my rep, I'm doing this, I'm doing that, shortening that whole process, providing more accuracy and immediacy with respect to, when I said agnostic approach, whether it's ad CP, IAB, or simply MCP-enabled APIs, we we are not discriminant in that way. We want to work the way that publishers are looking to work. And we know that this is going to iterate over time. So we're we're not going down one direction. We are building initially to one, but that's just with the idea that this will expand out and we have to be open to it.

SPEAKER_01 9:00

I think that makes sense. Like from our perspective, we are building our own agents. We're also looking at MCP server technologies, which we've rolled out for both MediaTools and Lumina that allow us to enable other people's agents. But one of the things that I think is kind of interesting is that, you know, as we get this proliferation of agents and we get all this optimization, which is happening, you know, in real time, you know, and speed and at scale, you know, the system of record sort of concept that both, you know, Prisma and Guideline have been providing in the planning space for a while, I think it becomes more important, right? Because someone has to keep track of all the decisions that have been made and all the changes. And it feels like, you know, variations we're just getting way more variations now because of machine learning and now AI. So from my perspective, the the idea that you have a set of truth to look at becomes more valuable in this world because someone has to keep track of it all and make sure it was all spent and paid for correctly.

SPEAKER_02 10:05

Yeah, I mean, first of all, I can take you on the road to say that to everybody because at the end of the day, that's absolutely right. We're talking about a couple of a couple of things are happening. First of all, when in when you employ agents or you're employing AI, there's always going to be the human element and the human touch. At what point is someone decisioning to say, yes, I am pressing the button to say I approve what a, you know, what was done for me without me having to be as involved. And we know agents are working 24-7, but frankly, that workflow will work as fast as the person that kind of pre. But when that's done, 100% with scale of volume, the decisioning being made, the reason to have or the importance of the system of record is is highlighted or or increased tremendously. You need auditability, you need to account for every dollar spent, you need to account for measurement and outcomes. And and I look at the way that even the way surfaces are going, but ultimately it's hitting your platform or our platform as a system of record and being able to see that. And ultimately, when an agency or advertiser are looking to be sure that they got value for I entrusted you with X, I expected to see X, and all along the line, you have you can ascertain or you can be clear of what's happening in the supply chain, huge importance. So we look at it as like we continue to be the road and we continue to be the guardrails on the road uh the whole way through. It's people process process technology and data.

SPEAKER_01 11:30

Right.

SPEAKER_02 11:30

Every one of those components is impacted in AI or agents by in some way. And it's looking that ultimately, though, it comes back to a governed process.

SPEAKER_01 11:38

Well, you know, we talked about a lot of times this type of system of record stuff seems like less sexy than a lot of the AI. It's sexy right now. Yeah. Like what do you mean less sexy? You know, it's but it's it's it's so important. And I just again, like I feel like we're in a violent agreement that it's getting more important. And so I'm hoping, you know, that we get a little more buzz about system of records next year at Cam. Uh, or we can we we could have a whole series of system of record five guests. Sure. Sure. It's everyone's gonna be excited about that. It's system of record with Drew and Steve.

SPEAKER_02 12:13

I don't I might not go on the record with that, but uh I'm gonna tell you that that uh I don't know. For me, I actually scarily think that being able to tie things from beginning to end is is actually very, very cool in the sense of when when it comes down to what goes into that and what that looks like, what we're moving towards more we're there, right? Video. If I if we were doing this 15 years ago, we're talking about linear TV, we're talking about digital video, they're living in these two boxes. Talking into our dictaphones. That's right, not this fancy MC boom mic, whatever. But but the now when we're having, if I'm having a conversation, I'm not talking to a local linear buyer, I'm talking to someone that's in video investment, national video investment, local video investment, video investment, period. That is a a huge flip on on how we're looking at things and how what it means for buying, what it means for processing, what it means for measurement.

SPEAKER_01 13:05

Right. I think that we've seen those lines blur as the technology has increased, which makes sense, right? If people who are empowered to do more things, you've seen creative and media flow together more than it has in the past, but but still knowing what happened will always be important.

SPEAKER_02 13:23

Well, and we're in an interesting place. We have a we have an area where things have come more together and we've got a lot of fragmentation. And if we can use technologies, whether it's like I said, it don't default to AI, but when we look at technologies that bring this together in a way that's cohesive, and we talked in our integration, we talked a lot about efficiency, but a big focus of this is efficacy. Right. And and how do we best capitalize on the investments being made? How do we make sure the dollars aren't going down the road and being thrown away because of extra investment of time or extra, you know, just putting it where it shouldn't be? So so it all ties back to ultimately the the the game hasn't changed in the sense of it's still I have an outcome I'm trying to produce. I am spending or I'm putting a budget together to do that. How do we do that in a way that gets to that end goal, whether it be brand awareness, sales, you know, go down the line. I could say more. Like like we could start with why do we have advertising?

SPEAKER_01 14:18

What is this advertising you speak of, Drew? Yes. Yep. Anyway, Drew, thanks so much for coming on the pod. It's always great to hang out and hear in France or elsewhere. And hope to be hanging out with you sometime soon. Yeah, no, no. I can can hardly wait to get back together with you. Awesome.

SPEAKER_00 14:39

Thanks for listening. And don't forget to also subscribe and listen to our Wednesday podcast, Media Monitor, where we go deep on trends, look into the data, and try to understand where the media market's going.