The Role of AI in AdTech – Digital CxO Podcast EP119

Hello and welcome to the digital CXO podcast. I'm Amanda Ani, and with me today I have Mike Novacel. He is VP of Strategic Partnerships for Stack Adapt.

How are you doing today? Good. Great to meet you.

Thanks for having me. Happy to have you on the show. Can you share a little bit about Stack Adapt and what services do you provide?

Sure. Uh, stack Adapt is a marketing operating system. So our core business is providing a best in class platform for agencies and brands to run their advertising in.

Uh, and increasingly over time, through some of the work of my team and others within the organization, we're increasingly doing more to build interoperability with how brands and agencies can drive efficacy of their advertising spend through connections and integration with broader parts of their marketing stack as well. Um, such as their CRM systems, their customer data platforms, business intelligence platforms, creative tools, you name it. Um, to help them basically get as quickly and efficiently to an outcome as humanly possible with as few steps.

Wonderful. Well, the advertising industry has really taken on so many new technologies, and today we're talking about AI in the ad tech space. So from your experience, where does AI fall in the ad tech space right now?

Yeah, so this one's a little close to home for me because, uh, I was privy to kind of the early days of when people were conflating AI with machine learning and just basics of data science. So, um, for context, so I led new business sales at MediaMath back from 2014 to 2019. We were essentially the largest DSP in the market that was independent for quite some time, and we were very early at really applying a lot of the more advanced concepts of machine learning and predictive analytics to how campaigns ran.

Um, but I think the thing that even just taking a step back into programmatic nearly six years later I've watched come to fruition, is that a lot of the legacy tech vendors and partners in this space, they're still relying on old concepts and they weren't necessarily built to really adapt to larger data sets that are, uh, less defined and less discreet, uh, and require much more, I'd say intentional connectivity to help an advertiser predict what should happen next and where they should spend every dollar. So, um, one reason why Stack App's kind of at the forefront of that, in my opinion, is because we ironically serve the customers the benefit the most from ai. It's these agencies and brands that are trying to do more with less habitually week in, week out, uh, and are really trying to differentiate themselves in the market, right?

If you think about who is probably most inclined to be the benefactor of ai, it's these companies that are historically under-resourced and also really trying to, uh, grow their business at the end of the day and also grow the businesses of their clients. Um, so I think ethically that's been one thing that's really aligned very well with the way we approach, uh, application of AI here at StackAdapt. Um, and a couple quick nods we generally refer to when we talk about it, um, it expands or expands from everything from quality of life of the agency.

So how do we use AI to reduce as many steps as possible between them building a forecast to executing a campaign, uh, to making changes to their campaigns. So saying I wanna automatically, for example, update bids across thousands of campaigns when I'm an agency of five people and I don't have the time to do that manually, uh, to behind the scenes through the way we look at different data sets across verticals to help small budgets work harder, uh, and change models on the fly. Uh, so it's really baked into our DNA and it's largely been, you know, carrying through because of who we serve at the end of the day as our core customer base.

And so from your experience, is it more cost effective to utilize AI technology rather than hiring on more employees? Uh, at least from my own limited perspective, yes. Uh, so my team specifically at sap, we have a unique role, um, and position within the company in that we're generally trying to work with technology companies and ads businesses to power some aspect of, uh, what they're trying to build.

Um, so I have everything ranging from integrations and alliances within my team. Uh, so for example, we built product offerings with HubSpot, with Shopify, with Snowflake, with Salesforce, et cetera, with the role in those integrations is to reduce the burden of getting to an outcome by using a system that I've already signed up for and pay money such as CRM platform or commerce platform, and then getting data into Stacked app to run effective advertising, uh, to ads businesses like publishers or retail media networks or tech companies where they provide a service or an offering to a core customer base. So think of, for example, a MarTech company that sells into local SMBs and gives them one place to do everything from managing email to sponsored listings, to Google my business reviews, to advertising, um, where we are the backend of that.

And in all those situations, people are very intentional about how do I do more with less? And what role can AI play in getting me more quickly and efficiently to an outcome? Uh, so yeah, so in that regard, I definitely see the benefit of it, uh, because it helps also our partners hit their revenue goals more quickly, which they're able to get to by extension of going and giving their customers something that performs better and gets them to a better outcome faster.

Mm-hmm. And you mentioned earlier data, and of course data is gold for most companies, so, and there's just more and more data. So how is AI assisting in, um, harnessing that data drawing insights, management of the data?

Can you give some use case examples? Yeah, I mean, I generally like to refer to the fact that, um, time's the most valuable asset, right? So, you know, people commonly draw the distinction between AI and machine learning as, uh, you know, kind of being an extension of giving a very specific prompt versus telling something to answer a broader question that's looking at, uh, much less clean uniform data sets and then arriving upon a predicted outcome.

So I think in that regard, uh, you know, that really connects to advertising and that so much plays a role now in figuring out where's the right piece or right, where's the right place, excuse me, to deliver the right message to a consumer. Um, so when that comes to that brand, for example, our agency on behalf of a brand that is starting from scratch, right? They only have so many sales, they only have so many observations of an end consumer, but they have a hypothesis of who's their brand's gonna resonate with.

Uh, instead of them guessing and having to walk in and launching a campaign very explicitly saying, here is who I wanna reach, here's why, et cetera. You can flip the script, right? You can start with your goals and your objectives and say, I need to go hit this amount of sales.

I am in this category, I am this sort of vertical or profile of advertiser here generally is what I think my hypothesis is and my direct right in that assumption based off of all the data that you've seen across your platform in terms of performance, where should I start? Right? And that's usually been commonly referred to as kinda the cold start problem, even with like the early days of data science and machine learning being applied in, uh, programmatic advertising, let alone any marketing concept, is that if you didn't have a critical mass of your own data to start with, or the platform, you don't really have a great place to start.

So you ended up having to exhaust a lot of time and resources into pressure testing till a model could be developed, right? Um, and that really has flipped on its head because of the fact that AI can help preemptively get ahead of that problem, um, look at a broader and less structured, uh, series of data sets to recommend before someone has to get to market, which therein helps 'em be more effective in the process. So as AI gets integrated in more and more processes, how important is the human strategy?

Uh, I think it's very important. So at the end of the day, the, you know, I think the thing we try to focus in on is that AI helps liberate people from doing the mundane, and it helps refocus and reinvest that time into critical thinking in areas where subject matter expertise and experience will always trump, uh, what a machine obviously cannot fill the void of. Uh, so I think in that regard, you know, the, I'll pick like one specific example creative.

Um, so I for example, uh, you know, have quite a decent amount of experience getting exposed to what do creative or creative marketers have to go through, right? What do creative agencies think about every day? And if you just pick that as one example, a lot of the creative agency market has historically been operating off of a very like, you know, I'll say legacy billable hour manual work sort of model.

Um, what does AI empower is it empowers that person that wet to Pratt Institute to no longer have to resize 300 by two fifties for hours in their day anymore. They can refocus that time in applying what they went to school for and applying what they ultimately, you know, brought them passion and joy and led them to sign up for the occupation to begin with, right? Creative thinking, how do I build a storyboard?

How do I tell and really embody the values of this brand when I go create original content? Uh, so a lot of, I'd say the things that historically people were unintentionally pulled into having to plug the gap of technology, not filling the void of now AI solves for, so they can spend their time on much more impactful, meaningful work in the process. So that's where I, you know, typically think there's a very blatantly obvious beneficial example to people in marketing As AI evolves.

Where do you see AI in the ad tech industry two or three years from now down the road? Oh man. Uh, I think it's really gonna, you're gonna notice it front and center the moment you log into platforms.

Um, so everything from, you know, I Amanda log into a platform and it immediately welcomes me the same way that, you know, uh, airline loyalty program when I'm a diamond status member recognizes my preferences and knows what I like to do day in, day out. Like that experience will increasingly become more tailored relative to the user of the end tools and platforms that they use in advertising. I think that'll transcend into knowing also who are the brands that they're typically acting on behalf of.

What are those brands typically caring about? Uh, what publicly referenceable data sources, uh, for those brands are available and how does that play a role in recommending things, right? So imagine for example, you're an agency and you're working on behalf of a large Fortune 100 brand.

Don't you think it'd be great if the platform you're about to use to go run their campaigns knew where they were pacing against their sales numbers for the quarter before you had to go dig it up, and then by extension of that already gave you a recommendation of where you should invest their capital. That's probably a very logical example of where, uh, advertising technology can take AI going forward in the planning and forecasting process, right? Um, and then I think in the, in the middle of campaign execution, even after as well as where I see a big opportunity, um, a lot of people always refer to, uh, you know, kind of the a flywheel of tests, observe, learn, adapt, test again over and over.

But many platforms have not really embodied that in terms of the way they operate. Um, so I think on the back end of it, as campaigns are active and live, you're gonna notice a lot of platforms orienting themselves, probably more about using AI to elevate insights very quickly and giving the cons the human, the ability to opt into recommendations more frequently and quickly, uh, and course correct and remove that manual labor that goes into what historically has been the case as well. Um, and then I think the final, you know, the last mile of it back to that, that closed loop is, you know, what do you do next, right?

So, um, if you think about the typical like rote process that many people go through from budget planning to uh, media mix recommendations to what are we gonna do for the quarter, right? That model flips on its head. Like, what's the need for big blocks of financial planning and media planning when you can figure those things out on the fly much more efficiently and effectively.

Uh, so I think it, that's gonna be one dynamic that really changes quite a bit as well as platforms improve their utilization of ai. Wonderful. Well, if there was one key takeaway you could leave our audience with today, what would that be?

Oh man. Uh, that's a loaded one. Uh, think really hard about where you spend the most time in your day putting in too much manual effort and push your partners to address that for you.

Um, 'cause I think one key ingredient in helping the overall ecosystem get better is really bringing attention to where the partners you work with maybe you're unaware of. Uh, you having been confronted with too much of a burden to get to an outcome. Uh, and the best source of getting to that is the customer, which is you.

So, uh, I'd encourage any of us on the vendor side to ask our partners that and on the agency and brand side for us to be more attentive to speaking up, up about it because feedback goes a long way and making this whole industry better. Alright, well thank you so much for coming on the show and sharing your insights. Yeah, thank you for having me.

Really appreciate it. And thank you to our audience. Stay tuned.

There's more.