Skip to main content

Gumball’s AI-Powered Campaign Builder

Article by Gumball Gumball

April 2, 2024

Gumball CEO Marty Michael joins Bryan Barletta to debut the new Campaign Builder Gumball has released, using AI-powered tech to assist advertising campaign setup. Below is a transcript of their conversation.

Bryan Barletta  00:50

Hi, everyone, Brian Barletta here from Sounds Profitable, and we’re gonna be doing a special product DeepDive live, because Marty here from Gumball has a brand new feature to introduce. So Marty, why don’t you introduce yourself, Gumball, and a little bit about this new product?

Marty Michael  01:05

Great. I’m Marty, I’m the CEO and co founder of Gumball. Gumball is a platform that helps automate the process of buying a host read ad on podcasting. So what we do is we connect brands and creators for that process.

Bryan Barletta  01:20

Awesome. And this feature brings it a little bit further.

Marty Michael  01:22

Yeah. So what we announced today was the campaign builder using AI technology.

Bryan Barletta  01:28

So let’s walk through that. How does that work? 

Marty Michael  01:30

Yeah, so typically, a brand would come onto the platform, they could enter in some criteria, the demographics, they’re looking for, maybe a content category, and it would filter out the shows that meet that criteria. This takes that one step further, let’s say you’re really busy. And you want help putting that together. What we designed and built was a product that can comb your website. So you enter in your brand, URL, some budget parameters, some dates that are worked working for your campaign. And it will go through and find all the shows that meet that criteria based on a bunch of different searches that it’s making on the back end.

Bryan Barletta  02:08

So it’s using the website. So whatever my website is live for my specific brand, it’s pulling all the data in from that as many signals as possible, and trying to distill that down into things that will match against the podcasts in your platform.

Marty Michael  02:20

That’s right. So when you enter in your brand URL, the AI is going to work. It’s scanning the webpage for brand indicators, it’s matching that against a database of brand profiles as well and spitting out a complete robust profile for your brand. And then what it’s doing is it’s combing through the shows that are on the platform to find the best matches. And the way that it’s doing that is doing several 100 things, but coming to show descriptions, and then also taking into account shows that have seen success with similar brands on the platform, just combining all of that to spit out an entire campaign.

Bryan Barletta  02:52

Now that campaign that gets spit out, can I modify it still? Or do I have to take it as is? No,

Marty Michael  02:57

You can definitely modify it. So what it’s doing is it’s going to show you every single spot from every single show that it’s recommending to meet the budget parameters that you put in. But from there, you can simply uncheck certain boxes, if there’s a show that doesn’t, that you feel isn’t going to be a good fit. Maybe there’s a crater you don’t want to work with, you just uncheck those spots. And then at the very bottom, you say move to cart, and that’s it’ll move everything remaining into your cart.

Bryan Barletta  03:19

One of the things that excites me the most about AI and especially in advertising technology is helping make more informed decisions. So right now, what you have here is taking some solid information from a website and the visible information from a podcast. As this grows, where do you see it expanding on either side? Where do you see it on taking into account? How many campaigns are running through Gumball overall and performance in aspects like that?

Marty Michael  03:44

It’s a really good question. For right now we’re only working within the context of our own shows, I think part of the magic of this tool is that we could work with shows outside the platform, make recommendations on shows that would be a great fit for your brand, not only the ones that we’re working with. So that is a tricky hurdle to overcome, as we don’t have as much contextual as far as show performance with certain brands to match with. But I think there’s some secret there that you’d be able to come up with a instant result for what could be a good fit, and then use your own and your own data intelligence to find and comb through those shows for your campaign. When

Bryan Barletta  04:26

We talked about the data signals that a podcaster can provide when they sign up for Gumball, these are all the same signals, right? The title, the show the episode details, all that their podcast categories, their IAB categories. Are you guys starting to dig into and explore the value of like actually transcribing individual shows and getting deeper into the context there? Or are you finding that the surface level is enough to merit a test campaign?

Marty Michael  04:50

Yeah, without giving too much away? For sure. I think there’s way more deep digging that we can do and this is definitely our our first shot at bringing AI technology to campaign building. And the process of finding good shows to work with. But yeah, there, we’re already working on transcribing a lot of the advertising that’s going through the platform and the shows. And that will certainly start to help us make more informed decisions as we move forward.

Bryan Barletta  05:14

Now, what does it cost? What’s the minimum I can do for a campaign for this tool.

Marty Michael  05:19

So for the tool itself, we have different budget levels that are already built in, I believe, it starts at 50, and goes up to 200, every 50,000 increment. But again, once you get the show results, spit out from the builder, you can then edit that cart. So really, there is no minimum. So

Bryan Barletta  05:36

it comes down to like, what the smallest show you have and the one add by on them, but I can I can really distill it down to that if I wanted to. That’s correct. The reason you guys are doing this, like, what’s the main draw for it?

Marty Michael  05:47

Well, okay, I think, you know, part of the platform, the power of the platform is how easy it is to work with creators, like if you’re a brand you can come in, you can instantly see all the shows that meet the criteria of what you’re looking for. But that process is what we’re trying to improve every single day. And this campaign builder is really the next step in improving that process. So instead of taking the, you know, five minutes that would have taken to find all those shows, and do that, you can now rely on a technology AI system to build that for you instantly. So it’s just another way to get better and more powerful, informed decisions. When you’re when it comes to creating that campaign.

Bryan Barletta  06:25

I like that, because it means that you kind of mix it up right random and all these solutions to get like a random subset if me and you did, like a random playlists on older music apps, or podcasts, apps or whatever, we might end up with basically the same things, just different orders, right. And so utilizing AI to do this, and the unique differences from my side and unique differences from each podcast, even if we had competing brands, if we put it into this tool, or even the same brand, and you run it a slightly different way you update your site, you update some of your messaging there. The goal is ideally get them new things to try out, because this is to get them to expand what they’re buying on. 

Marty Michael  07:01

Yeah, I think it’s, it’s another incentive for brands to be able to get more information about what might be a good fit. But also it you know, it’s our job at the end of the day to bring as many creators as possible to the forefront of what should be a good fit for that brand. And one of the impossibilities of a content platform that doesn’t utilize technology is that it’s really hard for brands to find those shows because they’re relying on the human aspect of providing the recommendations, right. This takes that human aspect away and allows actual data and technology to provide those results, which I think is a really powerful tool when it comes to the ease of letting brands be able to see in reality, what the shows that are gonna be the best fit rather than asking. 

Bryan Barletta  07:49

Yeah, and I think that sounds profitable, we worked really hard to try and make the tech and everything accessible. I spent too many years in ad tech overall. So like I understand a little bit more of it. But like, just so everybody’s aware, it’s not in your best interest for a brand to use this tool, find it to be non valuable an inbound side of podcasting, right, like this is built to show them that podcast advertising works. 

Marty Michael  08:12

Yeah, and one of the cool things about the tool specifically is that you can we have, we do work with a majority of podcast creators, we also have video creators and simulcast creators. And you can actually choose which type of creator you want to work with or choose all and it will make recommendations for creators that are on the platform outside of podcasting, as well. So this is a, it’s a great way to find just the creators that are gonna resonate with your brand the most. 

Bryan Barletta  08:38

So, we are probably an hour and a half from when you made the post on LinkedIn, about this announcement. I definitely ambush you on this, but I wanted to get the story out there on it. This is a killer announcement. This is great timing for it. But what’s next, like where do you expect to bring this product in the near future?

Marty Michael  08:55

Yeah, I mean, I mentioned this earlier, I think there’s a real opportunity here to work outside of our ecosystem right now. It’s only providing recommendations on the craters that we work with. I do think that this is something that we need to bring to a wider audience. And I think it’s a it’s a really, really helpful tool. So why limit it right? Like if we can provide recommendations on the entire industry utilizing this tool? That is something that I would be very interested in. 

Bryan Barletta  09:23

Do you think that will lead to more partnerships between Gumball and other platforms and ad tech solutions and podcasting or more helping expand podcasting outside of just audio and even simulcast but into influencer?

Marty Michael  09:35

I think anything we can build that helps buyers get more information on the Creator slash make more informed decisions. It only helps the industry as a whole. Yeah, I’m not concerned about building this to try and increase our footprint or anything like that. This is really a tool which I find very helpful, very fun to use, and I’d love to be able to use it globally for podcasting.

Bryan Barletta  10:02

And my last question here, how long is the team been using it internally before you guys decided to bring it public? 

Marty Michael  10:07

Yeah, this is something that we’ve been working on for three to four months. It’s a pretty tight, quick turnaround. It’s something that we’ve been talking about for a while, like how do we make the buying process easier? That’s, you know, probably the quote that’s up on everybody’s whiteboard, right. And this was something that we felt was a great utilization of AI specifically to start with. So starting here, there’s plenty more AI tools that we’re working on now. And I’m excited to announce in the coming months and weeks, but this was definitely where we wanted to start.

Bryan Barletta  10:37

Well, this is awesome. So if I was a publisher wanted to learn more about this, or buyer wanted to learn more about this, how would I sign up? How would I find out more?

Marty Michael  10:45

Yeah, if you’re a brand, you should reach out to So you can reach out to our sales team, and they’ll help you set up with an account. As of this morning, our engineering team actually turned it on. So anybody who has a brand profile on Gumball can log in and they can actually play with the tool themselves. Anybody else who’s not on the platform should reach out to the sales team. Awesome. And for publishers same thing, same thing. Yeah. Creators, that Gumball that FM. They should they can also give you a demo of it as well.

Bryan Barletta

Fantastic. Yeah. Well, thank you so much for joining us for a live product deep dive with Marty from gumball. Thank you for having me. This is the first time I’ve recorded live in a studio and thank you to Nomono for the audio recording.