SmartBug Media® Named 2024 HubSpot North American Partner of the Year.    Learn More

Skip to content
Let's Talk

From Guesswork to Greatness: Paid Media in the AI Era

Paul Schmidt

VP of Marketing

SmartBug Media

Louis-Claude Martin

Paid Media Expert

SmartBug Media

woman looking at phone with coffee-1

Subscribe Now

Back to All Episodes

AI-Powered Paid Media: Faster Launches, Smarter Targeting, and Real Results

 

 From Guesswork to Greatness: Paid Media in the AI Era
2025-05-22  37 min
From Guesswork to Greatness: Paid Media in the AI Era
SmartBug on Tap
Play

 

Join Paul Schmidt, VP of Marketing at SmartBug, and Louis-Claude Martin, a seasoned paid media expert, as they unpack the real impact of AI on campaign strategy, targeting, and performance. From the power of first-party data to the evolving role of media managers, this episode reveals how to shift from manual guesswork to data-backed greatness in the age of AI.

📚 What You’ll Learn

  • How AI transforms paid media—from bidding and targeting to testing and scale
  • Why first-party data from CRMs like HubSpot is a goldmine for ad platforms
  • Where AI still needs human direction—and what that means for paid media teams
  • The time-saving benefits of AI tools for campaign buildout and creative testing
  • Whether AI agents will replace media managers or just make them stronger

🔑 Key Highlights & Timestamps

[00:20] Intro: The rise of AI in paid media—what’s changed and why it matters
[01:45] What makes first-party CRM data (like HubSpot’s) so powerful in ad platforms
[03:30] A real-world example: feeding actual customer data into Google AI for mortgage ads
[05:15] How AI-driven platforms like Performance Max decide when to bid higher
[07:05] Targeting evolution: why Google knows more than you think (without PII)
[09:00] How AI helps scale creative testing with speed and precision
[11:25] The limitations of AI in campaign strategy—why human oversight is still key
[13:45] Garbage in, garbage out: why AI is only as smart as your input
[15:50] Will AI replace paid media managers? A breakdown of evolving roles
[18:10] How AI reduces time spent on repetitive campaign buildout
[21:05] Multivariate testing: from 20 hours a week to just 3
[24:00] Targeting vs. context: the importance of user behavior history in bidding
[26:30] Creative development at scale—what AI can (and can’t) do
[29:40] Why your CRM audience size matters when feeding platforms
[32:15] Navigating privacy while leveraging customer data in ads
[35:00] The future of paid media: hybrid AI-human decision-making
[38:45] Final thoughts: what marketers should do today to stay ahead

🔗 Want to Connect?

👉 Contact SmartBug: www.smartbugmedia.com/contact-us
👉 Contact Paul: pschmidt@smartbugmedia.com

Relevant Resources

Paul Schmidt (00:14)
Hey everybody and welcome to another episode of SmartBug OnTap Today we're diving into one of the more exciting shifts that we're seeing in digital advertising today, which is the rise of AI in paid media. so if you ever launched a campaign, paid media campaign and cross your fingers hoping for great results and you know, you're not alone. For years, paid media involved a lot of guesswork related to targeting, to budget, to what?

messaging and creative to use. And thanks to AI and some of the advancements that we were seeing, a lot of our guesswork is being replaced with more precision and being able to go faster with launching campaigns. And so for marketers here that are looking to just be more transformative with some of their paid media, you're in the right place today. And so with that, we'll jump into today's episode, all about paid media in the AI era. So before we jump in today,

Don't forget to subscribe so you don't miss another episode of SmartBug on tap. Joining me today is someone who's been in the paid media world for a long time and is now helping our clients rethink their approach to digital advertising in the age of AI. Louis Martin, a paid media sales executive here at SmartBug. Louis, thanks and welcome to the show. Tell a little bit about, tell our listeners a little bit about yourself, your background and what you're working on these days.

Louis Martin (01:37)
Thanks, Paul. Hello, everybody. Yeah, so, well, I've been working in paid media for a long time, about 17 years now. I started almost by accident. I used to work for a company designing websites. And one day, I was at one of my clients, and he said, you know, I've paid a big chunk of money to a company to advertise on Google. Back then, was starting Google advertising. It was only like two or three years old.

And Google was not two or three years old, but the advertising on Google was kind of new. And so I took the weekend off to look into this, and I came back to my client the next Monday, and I said, I think I can help you. And that's how I started my online advertising journey. These days, my main focus is to help my client grow. And recently, I switched from

the production side of things because I used to manage campaigns and consult our clients on their strategy. Nowadays, and that's very recent, I just started on my sales journey so I'm now a sales executive at SmartBug.

Paul Schmidt (02:39)
Great. So let's dive right in with some of these questions. First one just kind of related to AI. So you're with ChatGPT coming online, you being in production for a mass market in 2023. You know, more folks are experimenting and testing it and testing it for different use cases. When what are you seeing are some of the biggest limitations of using and running using AI to help run paid media campaigns?

that you're seeing in the market today.

Louis Martin (03:10)
So one of the biggest limitations is that, or the biggest caveat if I can say, is that the AI is smart, but it's not that smart. It doesn't know the context necessarily of your company. It doesn't know the culture necessarily. So you still have to babysit it in a way. you need to give it as much context as possible. And the AI is only the

is only going to be as good as the prompt you're going to give it. So the more context, the more data you give it, the better results you're going to get. But don't get me wrong, AI is a huge revolution as far as I'm concerned. It's probably the biggest revolution in the history of the world. Not only in advertising, but at large.

Paul Schmidt (03:54)
Yeah, no, a hundred percent. you talked a lot about context and we'll get into that a little bit, like making sure that the AI has enough context to do a good job. completely agree with you. If you give it a, a very short prompt, don't give it enough information. You're going to get generic results. So I agree with you on that. When you think about the pre AI era, like before 2023, how do marketers typically approach targeting and.

budgeting and optimization within their paid media campaigns.

Louis Martin (04:19)
It was a lot of trial error back in the days up until maybe I'd say like three, four years ago when they started introducing a different AI and then this kind of campaign like Performance Max on Google and Advantage Plus on Facebook. But prior to that, it was like I said, a lot of trial and error. needed to have

have some very specific plan on what kind of audience you want to target and you need to target like create like 10 different ad groups or ad set with 10 different audiences separately and at the end of the day you will know which one ad performed the best and the same thing goes for the ad creatives. You need to test like 10, 100, 10, 50, 100 ad creatives. The more the merrier as long as you have the budget for. Nowadays you just like

Pitch in what you got, your text, your audiences, and the AI is gonna do the work in the background, and it's gonna come up with the best combinations to give you the best results. So nowadays, you give an objective to the AI, and it's gonna take care of the segmentation, finding the right audience, the right combinations of text images, it's, you still need to have somebody driving the

the platform but the AIs is you know taking care of what's what used to be more tedious.

Paul Schmidt (05:41)
Got it. Yeah. And it seems, you know, pre AI era, so many, so many campaigns that I would see clients running, were built based on the, the paid media manager's gut instinct or the client's gut instinct on what they feel is going to work. And, I mean, intuition is really important and, and, know, using your experience on it, but, many times, if this is like your first time running, paid media campaign for your company, like you may not know the keywords to run.

to target, you may not know the type of creative that's really resonant with your audience. What are some of the signs that a paid media strategy is just overly reliant on gut instinct versus being backed in data?

Louis Martin (06:18)
So, well, one of the signs is that when the campaign is not consistently performing correctly, like one month it's going to be performing well and the next for some reason it's going to underperform by a lot. So that's one of the signs you can know. And of course with AI now it's much more consistent the results you're going to get, but you need to train the AI. So when you start a campaign, don't expect it to

to have like amazing results after five minutes. need to give it, Google says four to six weeks. In reality it's less than that, but it's clearly not one day. It needs to learn and there's a learning phase. In fact, in the platforms, you'll see what stage your campaign is in the beginning and it will stay still in the learning phase. So you gotta be careful to let it run its show and let it do its testing.

before you make large changes, otherwise it's gonna stay in an endless learning phase.

Paul Schmidt (07:13)
There's that window of time that you can't just turn paid media ads on and one day expect to have clear learnings from it. We've definitely seen clients and different folks that we've worked with get a little impatient that, I'm not seeing leads come in the door right away. And they get frustrated that there is that period of learning. I think that what you said makes a lot of sense.

Do you have any like examples or anything like that that you could share that's kind of related to what you were just talking about?

Louis Martin (07:45)
One of them is we have a client of ours in the home service franchise business and we introduced AI campaigns and since we did introduce that, it helped narrow down which zip codes because it's a regional...

business so you know they're not advertising to the whole world they're really like very narrow geographically targeting like in an area in the city and so with the AI we are now able to target the correct zip codes and letting the AI find which are the correct zip codes to advertise to and allow a bigger budget to the zip codes that are performing better.

as opposed to the zip codes that aren't. That's one example. We also have a client, it's two hairdressers, I would have never thought that five years ago. But they started a little website five years ago here in Quebec. And we're running ads for them for three years now.

And on the Google side of things, once we introduced performance max campaigns, which are highly driven by AI, they started getting amazing results. And anybody in that field will know that the 20x ROAS is good. Normally, anything under 10, over 10, or even like over 5, 6, depending on your margin, is good. They have consistently 20x results. And even on their brands, their

their returns are 150-200X which means for every dollar they spend they sell for 150-200 dollars.

Paul Schmidt (09:24)
Wow. so let's switch gears a little bit and talk about how AI is actually being used today by our clients. We work with dozens of franchise organizations. work in health care. We work with SAS. We work with financial services organizations. So there are broad types of industries and tactics that we're using.

Can you talk about the most impactful ways that AI is being used in paid media today?

Louis Martin (09:50)
Well, once again, with the campaigns, for example, again, like performance max campaigns, it's going to test the audience. It's going to use predictive models to decide, am I going to bid higher for that person that's searching a certain keyword? Am I going to bid higher for that person, depending on the context of his search and also what that person has done in the past? Because Google, although with all the cookie and the privacy,

is not allowed to like know the name of the person it still knows a lot of things and possibly the Somehow the the browser history of the person so it's gonna know the context and it's gonna decide yeah this person is really looking for that product or not, so it's gonna bid more for for that particular person and Therefore give you a better chance to get the click and drive results

So that's one way that the AI is helping, but also, and this is like for more like the bidding and the audience targeting, but also anything related to building ads, better ads and testing better ads. And I'm talking about the creative here. It's gonna be able to test creative way more efficiently than we used to because back in the days it was like very manual.

Nowadays, like I said, previously you dump in a bunch of ads and the AI can also help you create ads. And it's gonna test which one's performing the best, the better. So it's a lot. It has increased our capacity to run ads by, I wouldn't say 10-folds, but it makes us much more efficient and much quicker to adapt and react and pivot when things are going the we want.

Paul Schmidt (11:27)
Got it. So, know, the beginning of that last question, you know, you talked a little bit about Google performance, Max and Meta's new functionality. Can you just dive into a little bit more about those specifically as we kind of move away from first party cookies? I'm just, I'd be curious, how those tools are leveraging AI to be more effective.

Louis Martin (11:47)
Yeah, so, PerformanceMax is for Google, Advantage Plus is for Meta, and they kind of act the same way. It's kind of a black box, okay? Back in the days, you would need to say, these are my keywords, these are my headlines, titles, descriptions, and you would need to separate them by keywords and by audiences. Nowadays, you throw this all in, PerformanceMax, for example,

and it's gonna search for the best audience for the client according to the context of your website, of all the input you gave to PerformanceMax. You can add audience signals also, so you can enhance your campaign results by giving first party data. So what is first party data? Well, first party data is everything the company has

as far as information in regards to your clients like name, phone number, telephone. so with that, if you inject that into the Performance Max campaign, it's gonna help Google find the clients, like a lookalike clients, like people that look like your clients, much more easily and once again, enhance your performance.

Paul Schmidt (12:53)
Yeah, that makes sense. like the combination of tools allowing you to insert your own first party data, whether that's stuff from your CRM or other analytics platforms to be able to just further strengthen some of those campaigns. I like how those things come together. Can you talk a little bit about AI's role within campaign optimization and targeting and creative development?

Louis Martin (13:14)
So as far as targeting, like I said, it's gonna know is this person a match for my advertiser. The person just searched a broad keyword, for example. It might be a match, it might not be a match, but Google is gonna know. Okay, well this person has been to these kinds of website, has made these kinds of searches prior to making that search.

it's going to optimize and it's going to bid more efficiently. It's going to bid higher if it thinks, well, this person has a high probability to match what the company is selling as far as what that person is searching, if that makes sense. So for the targeting, it's really going to help putting the ad in front of the right client and also bidding more efficiently when there's a fit. And for the creative, it's going to help us scale

in a way that we couldn't do before just because we're to be able to test more creative, more efficiently, more quickly and have a high level of confidence which creative is performing the best at any given time.

Paul Schmidt (14:15)
Okay. Yeah. And kind of a side question to that, creative development side of things, you know, and building out some, meta campaigns I was recently working on, you know, they have the AI functionality inside there that will take an image and be able to like create new variations of that image just right within the advertising platform itself. I'd be curious, you know, like that, that uses AI specifically to, come up with new, creative, but

Do you feel like that specific functionality that's like native inside of Meta and coming up with those new images and creative, is that more effective or would you say like manually, you know, using Canva or Photoshop or things like that to come up with variations? Is that a better route to go to come up with that?

Louis Martin (14:58)
I'd like to say yeah, AI is like the one shop, one size fits all right now, but sometimes, and that's why you still need a human behind the wheel, sometimes a human is gonna make a difference because it's gonna know more about the context, the culture. So the AI is really a tool to help, but it's a tool. It's still not like as smart as a human, so you're still needing a human to help.

Give more context to the AI so then it's is gonna give you better results But yeah Canva can still be good the best way to use it is maybe in conjunction with the AI to help create more variations so like you can create like the first draft and then you can send it to your AI tool to create variation of that much quicker than a human would but the first draft might be highly dependent

of the creator.

Paul Schmidt (15:50)
Yeah. Yeah. And using those tools too, it felt like there was some degree of hallucination that was going on where you, you know, you think about some of those early AI images where someone has like seven fingers and things like that. So I, I think, you know, you definitely have to have that human coming in until that AI is so good where it's much more realistic and accurate. So let's switch gears to talk about agents, agentic AI.

We're starting to see more agents across HubSpot and other platforms, folks using AI agents for multiple different types of

what's the value of using agents like from a paid media context? Like how are they helping teams be more effective and eliminate manual tasks?

Louis Martin (16:34)
It's really going to help the managers automate what used to be repetitive. So it's going to take that work off their hands. So they can really concentrate on what's important and strategizing,

Paul Schmidt (16:47)
I would add to that too, you know, there's inside of HubSpot, which a lot many of our clients also use for hosting landing pages and host in managing their ads and everything like that. Content agent is really what I've seen a lot of activity around, especially around just like landing page development. I think so often, you know, marketers, you know, get set, they have new goals in front of them or campaigns they want to launch in, you know, creating a brand new landing page from scratch is cumbersome, takes a long time.

And I think that's where we're seeing a lot of value and a lot of being able to move a lot more quickly, just being able to spin up a landing page based on a prompt versus you having to like dig out a template and create it from scratch every single time. So that's really where I'm seeing a lot of agentic stuff happening just inside of HubSpot specifically. Do you think that longer term, like AI agents are gonna replace the day-to-day tasks for paid media managers?

Or do you think it just helps strengthen paid media managers' decisions?

Louis Martin (17:45)
in the short term, I don't think they're going to replace managers. They're just going to help us. They're going to do the repetitive task much better than a person could do. So in the end, it's still going to need some direction from the managers. But it's going to analyze data much quicker than we could. it's really going to be a

It's like having a genius next to you to help you take decisions.

Paul Schmidt (18:13)
I like that a genius or a PhD level intern, something like that. So thinking about AI, think about like pre 2023 and where we are today. What are some things that were really time consuming back then that you feel like we could, moving like 10 X more quickly with these days leveraging AI.

Louis Martin (18:35)
Clearly one of the biggest time gain is in regards to a multi-variant creative testing and also multi-variant audience testing. I used to work for a company that was selling couches online and we could take 20 hours a week just creating campaigns. Now the same task would probably take me with the help of an AI three or four hours. So it really has helped us.

streamline much much quicker and putting out campaigns much much quicker and getting results much quicker and finding winners again much quicker

Paul Schmidt (19:10)
Yeah. So let's talk about how humans are actually being used then within paid media. We're talking about, we talked so much about AI today, but like AI can do a lot of that heavy lifting now. It can do a lot of that manual work we talked about. where should marketers be spending their time and energy if they have all this extra time now that AI is doing a lot of manual work.

Louis Martin (19:30)
to answer your question Paul, we should double down on what the machine can do. So, creativity, I know AI can do creativity, but it's still in direction and strategy also. We need to have like a big picture and a 10,000 feet view of what's happening

and then the AI is going to help us produce results. But once again, I think the brain of the human to have the big picture so far or right now, at least for the next couple of years, is still much, much needed.

Paul Schmidt (19:57)
Yeah. You know, it seems to me that, you know, historically without the use of AI, it's like marketers, you know, they have limited budgets, limited time to be able to run campaigns and they maybe focus on like a specific part of a funnel. And it seems that, with advances in AI and being able to move more quickly, it's not, it's like, you can not only just focus on a single part of the funnel, you can focus on other parts of the, of the buyer's journey, other parts of the funnel. You could also focus on other segments of the business. You could also.

just allocate capital a little bit more strategically across different parts of the business versus, you know, only being able to just push it all into like one segment and one part of the funnel at a single time.

So let's jump in next to, you know, thinking about first party data and CRM integration. You talked a little bit about this already. How important is making sure you're pushing that first party data, CRM data into your ad platforms?

Louis Martin (20:48)
It's quintessential. In the sense that the more you feed the platform with good data, and nobody has better data than the company themselves, so the more you can feed that with your CRM, for example, with HubSpot, can literally feed your audience and your list of clients to Google, with all privacy respected, of course.

Doing so will help the platforms and the advertising platforms reach your potential client with a much higher level of efficiency and accuracy than if you were not feeding them that first party data. So first party data is essentially who's your customer and you feed that to the platform and they're gonna give you a much better targeting.

and better results and a better bang for your bucks if you do so. I must add though that you need to have enough data to feed to the platform. If you have like, I don't know, 10 customers in your database, it's not gonna work. But like, our clients usually have like a huge database, so by feeding that to Google and Facebook, it's gonna help them have better results, much better

Paul Schmidt (21:57)
brings up a good question, which is for clients that don't have enough data or have dirty data or disparate data, things like that. Do you have any advice for those types of companies that want to take advantage of first party data, but maybe they're either a small company or a very narrow company. think about some of these.

government contractors that maybe can only work with two or three different companies or four companies or accounts at once. Any advice just in terms of those types of instances where they don't have thousands of customers to pull in data from?

Louis Martin (22:34)
Well, you're gonna need to set up some test if you wanna get better results in those case. And you need to segment the kinds of actions you wanna track and give a value to those actions. For example, maybe you have two types of form on your website that your potential customer can fill. And one of those form is a request a demo, for example. And the other form is contact us, okay? Request a demo.

potentially has a much higher value for you because you know that if somebody is requesting a demo, they're much further down the funnel and much closer to a buy to the final step. So you could allow that request a demo form to have a certain value and feed that to Google or Facebook. And by doing so, you're gonna help feed the algorithm.

and letting the algorithm know, listen, the people that fill those requested demo forms have a much higher value, therefore I want you to look for those kind of folks as opposed to somebody fills a contactors form, this is just an example of course, but let's say somebody fills a contactors form, his value is 50 but the requested demo is 100, so it's gonna try to balance to give you the most client that looks like somebody would request a demo.

Paul Schmidt (23:48)
And I think as long as you're tracking that data, whether it's you're using HubSpot or GA or any type of platform, you can really see what that conversion rate is from like demo to customer or contact us form to customer. And you can start to work out some of those calculations. So that makes sense. Let's jump into where marketers go wrong with AI. They turn their agents on or they sort of set it and forget it with some of their AI tools. What do you think marketers are doing wrong today with the AI?

Louis Martin (24:15)
Well, I think you kind of set it in your question. The worst thing you can do is just turn it on and forget about it. This is not a fire and forget solution. You need to monitor what's going on and give context and more feedback. The more feedback you're going to give to your campaign and to the AI agent, the better results you're going to get.

One of the biggest mistakes would clearly be to just let it run without monitoring it

Paul Schmidt (24:42)
let's go to the last question in this segment. So what's an example of a smart strategic decision you've seen paired with AI execution?

Louis Martin (24:51)
I think I kind of talked about it in the last question, but let me be more precise. if you use multiple conversion actions, each with its own conversion value, and then you let the AI optimize, it's gonna give you the best conversion value. And the more data points you have, the better it is. You need the traffic also. yeah, so if you have...

a bunch of conversions, the worst thing you can do is allow them to have the same value. No conversion should have the same value normally. They should have different values for you because it's not true that they're not all created equal.

Paul Schmidt (25:24)
Yeah, yeah, that makes sense. All right, so let's jump next into a case study client example one. Can you share a recent client example where AI driven paid media made a measurable impact?

Louis Martin (25:38)
Yeah, so we launched a campaign for one of our clients in the financial business and two things we did. First of all, we remarketed to their audiences, but we also feed them with their first party data. And you might say, well, isn't it the same thing? Well, no, because when you remarket to somebody, can just, for example, remarket to somebody who visited you.

visited your website, but in that case, that's what we did. We started a campaign and part of the audience were people who visited the website, also people who filled some forms, okay? But also, and this data, we don't have it as an advertiser, they also feed us with their data from actual clients that bought in the end. we don't have, the platform doesn't have this information unless we feed that data.

and doing so improve their performance and their cost per lead dropped by 66%.

Paul Schmidt (26:35)
Wow. What would you say made that campaign successful? Was it the creative that we're using? Was it the targeting? Was it that first party data? Was it budget allocation? Tell me a little bit more. Why was it so successful?

Louis Martin (26:48)
Well, mainly because we gave the AI the data, like the first party data, and we let it adjust the budget according to the context of every search the person were doing. And in doing so, it really drove down the cost per acquisition by 66%, once again.

Paul Schmidt (27:08)
So how specifically did SmartBug's team help leverage tools like HubSpot or any first party CRM data in that effort?

Louis Martin (27:18)
So what really helped is that we fed the campaign with the very good first party data from the customer. in that case, they're selling mortgage. I'm not gonna divulge the name of the client, but they're selling mortgages. And so the client fed us people who actually signed mortgages with them.

and we were able to feed that to the AI, in that case, the Google campaign. And of course, this is the best data you can have, because it's a thing to feed somebody who looks like a lead, but now we're feeding actual customers so we can tell Google, look, this is our customer, find data points that you can link those customers, and then go find me some customers like that that are in the market.

for mortgage right now that are looking at Google. So that's very powerful and that's what the AI of Google allows us to do.

Paul Schmidt (28:09)
Great. Well, let's move forward and just thinking about like the future of AI, how this is going to impact paid media. It's hard to, you we don't have a crystal ball obviously, but if we were to look sort of two, three years out from now, what do you think AI is going to do? How is it going to impact paid media? How is it going to impact some of our clients? How do you think it's going to help everybody along the way?

Louis Martin (28:36)
It's going to reduce the heavy lifting. It's going to remove the heavy lifting from the shoulders of the campaign manager. And it's going to allow the campaign manager really to have more time to strategize and plan at a higher level. And the tedious task like testing different ad variations, different text, titles, descriptions, images, that's going to

you know really gonna be passed on to the AI and the campaign managers are really more gonna focus on the high level strategy to to the they're kind of gonna become more like a a chef d'orchestre like we say in French like the the guy who managed the orchestra I don't know how to say like a conductor yeah

Paul Schmidt (29:14)
Hmm. Conductor.

Got it. Yeah. So, and then what about AI agents? So agentic AI is, there's a big wave of agentic AI happening in the market right now. Everybody's talking about agents and everything like that. How do you see the role of those evolving with paid media? Do you think we'll get to a point where agents are just able to run campaigns end to end without any human intervention or what's your take on how, like how much is the AI going to take over?

Louis Martin (29:44)
Eventually and I cannot know for sure but I think sooner than later AI is definitely gonna be able to run a campaign from end to end When is that gonna happen? We don't know my guess is as good as yours, but I I don't think I'm far off like three to five years They should be able to run campaigns from from it to it

Paul Schmidt (30:07)
So you just essentially stick in your credit card and say, Hey, this is the campaign I want to run. And then it'll just start giving you leads. Is that, is that what you think it'll happen?

Louis Martin (30:16)
Well, that's what Mark Zuckerberg wants. I was watching an interview yesterday actually with him and that's what he says. Their long term goal is like, you're a small business, you don't have like a big agency working with you and you don't have the budget to work with a big agency, well, give us your credit card and we'll do the work for you. Is it gonna be optimal? I don't know, we'll have to see but eventually, I guess that's where we're heading.

Paul Schmidt (30:19)
Yeah

Yeah. Well, let's talk a little bit in kind of closing today about the skills that marketers need to grow to be more effective in an era where AI is being leveraged for a lot of manual work. What are some of the skills that you think marketers need to improve upon and be thinking about to be able to make sure that, you know, they're not out of a job in a few years and AI is taking all over their work. What should they focus on?

Louis Martin (31:11)
anything like related to strategic thinking and high level creative thinking is going to help you stay in the game. Now between you and me long term I don't know I think AI is going to take over the world but like short term in the next couple of years like the more higher creative thinking you are the more strategic you are to help guide the AI well the better your chances are to keep your job.

in this ever changing world but yeah definitely AI is really advancing at the speed of light right now. We don't know where it's gonna go for sure but we know it's gonna help a lot with the most difficult tasks.

Paul Schmidt (31:55)
Yeah, it seems like in a short term, a couple of the areas that I've been talking with my team about is prompt engineering. That's one of them. It's like becoming better at creating the types of prompts essentially that will help you get the best outputs. And what do need to include within that prompt so that it's not hallucinating and that's getting you the desired output? think so many people that start leveraging AI at the very beginning, they...

They see just a sort of a blank box and they ask it a generic question and they get generic answers and they kind of give up pretty quickly. so I think that that's one area just in terms of like short-term training is like getting really good at prompt engineering so that you can just be better and be able to move more quickly with AI.

What investments would you say that marketing leaders need to make now to prepare for this next wave of change?

Louis Martin (32:46)
They should invest in three key areas. The first is data infrastructure. So the more connected, clean, and actionable your first party data is, the better your AI tools are gonna perform. like, once again, I don't know what the expression is, like garbage in, garbage out. So the better the information you the AI,

the better the results are gonna be. That's the first thing. The second thing is talent development. So you need to have like, very quickly I think, strategy is gonna be one of the most important things to have within your team. So people with strong analytic and strategy and creative thinking are gonna be key. And the third is a good tech stack, a flexible tech stack.

So tools that can evolve with AI. So platforms like HubSpot, GA4, and systems with open AI and AI capabilities. So you want to be in a position to pivot quickly. HubSpot is of course clearly one of the platforms to help you do that. You're on the cutting edge of what's happening right now.

Paul Schmidt (33:49)
Yeah. And I think that last one, you know, just mentioning like, what, know, when marketers and, leaders and organizations are thinking about like tech investments right now, it's like they're, you know, when they're thinking about vetting what tools they're going to buy, it's like they looking at the roadmap and what sort of AI development that that's on that roadmap, I think is so critical for leaders to be thinking about. I think more than, more than any other features said, that's what we're seeing within the HubSpot platform.

I would say across every single hub, in multiple different apps within each hub, there's some level of AI that's getting baked in. I think that's investing in your AI future when you invest in the tech stack that is thinking about that within their future.

Louis, before we wrap up, I just have a final question. What's one piece of advice that you would give to marketers who are looking to improve their paid media strategy in this age of AI? What would that be?

Louis Martin (34:39)
you need to be very, very specific with your goals. That's key, that's the foundation. And you have to get your data clean, get your data in order. once again, garbage in, garbage out. The more precise the data you give to the tools, the AI, the platforms, the better result they're gonna give you. So you need to make sure that your data is clean. So you need to focus on aligning your goals

with strong first party data and tracking that will matter and feed that into your campaign.

Paul Schmidt (35:11)
Awesome. Well, I think that's a great last insight to end on all about goals. And thank you so much for joining today, Louis. For everyone that's listening, as you know, a big takeaway here is AI is changing the game. it's not replacing our strategy by any means. It's allowing paid media teams to move more quickly, come up with better creative, be more effective with how they're allocating their capital.

If you've enjoyed today's episode, we definitely want you to subscribe so you don't miss all the conversations that we have coming up with other experts here at SmartBug. If you have any questions or would like to deep dive into any other things that we talked about today, Louie and I would love to hear from you. We'll put our emails in the description below and feel free to reach out at any time. Louie, thanks again and thank you for joining today's SmartBug on tap and we'll see you next time.