Utilizing AI to Propel Podcasting with Pete Birsinger - CEO and Founder of Podscribe

Our special guest today is Pete Burser, the founder and CEO of Podscribe. In this episode, we explore topics at the heart of podcasting as we explore Pete's fascinating career journey, We'll be delving into the actual value of transcription and its impact on the industry, and exploring the significance and limitations of attribution. Pete will also share his valuable insights on the ever evolving landscape of dynamic ad insertion DAI versus host red baked in ads. But that's not all.

We also discuss AI's exciting role in revolutionizing podcasting, taking a broader view of the subject. The episode will also cover important topics like brand and host safety. So get ready to be inspired as we journey through the world of podcasting with one of its prominent trail blazers, Pete first Singer. Hi, how are you.

I'm doing great. Thank you for having me, Marla. I appreciate it. So for purposes of my sanity and our listeners, can you please state your name?

Pete Burstsinger. Yeah, I probably would have mangled it. I I'm sorry. Mostly now most people they just say birst Singer, which I don't know how anybody would know it's I mean, my wife dad, I don't know if she says it right.

Actually no, I hear. Yeah, people pronounce ASA incorrectly. So we've got the same issue. So tell me something exciting from the past week.

Anything you want to talk about, something that was cool to you. Cool. We we're here in Austin and we got our polar plunge set up. So it's very consistently in Austin been for the past probab month hasn't rained it drop.

Then the loaves have been seventy eight and the highs typically like one oh two, one oh three pretty much every day. But now it's nice because we've had the polar plunge. This chest freezer in our in our garage set up, so I'll go there mid day. I have it down to thirty two point one.

It's a chest freezer. We fill with water and then I'll come out of that, you know, purified, and then I'll go work outside for an hour or two. All right, let me ask you a question, what is this plunge thing? What is that?

This cold water? Oh? Okay, yeah, this cold cold water. So I think people talk about some physical benefits like reducing inflammation.

Yeah, and I like it just because it's a way to to reset myself for a little bit in the middle of the day. So yeah, here you okay, I just keep cool. I had not heard that turn before. But yes, drinking water I get illed at all the time.

That's important to hydrate. So first podcast you ever listened to? I was driving with a friend to Tahoe maybe twenty thirteen, twenty fourteen in California, and they it was a three to four hour drive from the Bay Area, and then my friend is like, oh, you all listen to a podcast. I was like, I don't know what those really are, but he's like, oh, let's just try.

It's a lot. Maybe he did want to talk to me, but so he put on this American Life of course, right, so that was what it was kind of interesting. It's about some life on a Navy ship, so it was interesting. I think the first one I listened to on my own or so was Tim Ferris, so that still a lot of his so yeah, he's awesome.

So but overall, any favorite podcast genre. I listen a lot out of business podcasts. So two of them that I listened to now are the all in podcast. I think that's growing in popularity with some Silicon Valley VC types.

And then sometimes I listened to this Weekend Startups by one of the hosts too. Of all in podcasts, it's a pretty good one. I'll listen to alex Or Moosey has a podcast that's another business one. So a lot of business ones.

Usually maybe I need to diversify. Yeah, maybe something fun. But so do you find that there's enough I'm just curious enough diversity in the in these business podcasts and what I mean by that diversity of content, like different kinds of content. I do find a lot of the time I'm out of podcasts that I want to listen to, So there could be more.

Certainly, maybe I need a search a bit more. And then when that happens, usually what I do when I've listened to a couple of podcasts I listened to, I cut to my audiobook. So I so I've got a shower speaking and all listen, I'll listen in there. So I think I think one I could probably try to look harder for more business podcasts.

I'm sure there's a lot, but I would say even if there were, I usually yeah, when I I could have more listening if there was more that I was that I had found right, So I have to admit that I listened to like living your life as a Frenchwoman. I am not a frenchwoman, but it's so far afield from what I do day to day, so it's interesting. I probably should listen more to business podcast here. It's a good point.

I use it as real escapism. So okay, so let's talk about your background. You're from the Bay Area. So what was it like growing up as a millennial in this tech industry, startup, crazy venture capital it seems of the world.

I think I maybe caught the tail end of the peak. I went to UC Berkeley two thousand and nine to twenty thirteen, where and then pretty much I studied software engineering there, so I was well immersed in the Bay Area software engineering culture and world. And a lot of my friends would go work at Facebook or Google and these tech startups afterwards, and then a lot of them after that would then work there for a couple of years and get maybe a little disillusioned and try to start their own thing or get out of software engineering, or some keep some kept at it.

Certainly, I would say it was a good place, a good place to be, certainly. I moved to New York and twenty fifteen to Sea because I grew up in SF, went to school in the Bay Area, then moved back to SF to work in a tech start up there, which which was bought by Cisco. So I think it was not bad. I came on as they're getting acquired, so it's really no benefits and the acts of getting to work with Cisco people, so I think it was It was an exciting time.

I think. I think SUP is in a bit of a low point now, but I think it's coming back. I've heard of a number of AI startups taking actual route there and getting offices downtown. I think that may start to bring some people back there.

I moved to Austin becactly. I like the energy of the city there. It feels like it's a bit more of the place to be now. The low taxes and son I think help, but it's really the energy is what I liked and drew me most to Austin.

Oh interesting, Yeah, my nephew lives in Austin. He's got a start up too, but an energy would start up and they love living there, So I hear you So you're the CEO and founder of pod scribe, and we're going to talk more about that in a bit, but it's not your first rodeo. So tell me about pod Bible, which he co founded at around the same time as pod Scribe. Do I have that right?

More more or less? We actually so, it was. It was a weird name, podible. We had a couple of names for it, but we would call it podible more or less.

But oh my god, ever, got did I spell it right? Oh? Yeah, it's at pod bible. It's podible.

No, no, no. Whenever we started the name, we always regretted because even when we said podible, half the people would spell it p O D A b Ellie. So we we started because I was working in Atteck in New York and my boss, actually I was I became a sales engineer. He was saying he took a data science boot camp and was analyzing podcasts the topics that are in them, and he was working on making a recommendation engine to help people discover new podcasts even back then.

Every I mean, I think it's still true now, but discovery of new podcasts is tough. So I with a bit more of a software engineer back and I said I could help him, So then we began working on it. We built out a recommendation engine, and we made as part of a podcast player actually a listening a listening app. And we quit and raised a couple hundred thousand for the listening app, and he was going all right, but it was just a listening apps is a tough thing to build, or tough thing it's not easy to want to build.

First of all, there's a lot of bells and whistles that people take for granted. And then too, it's hard to get people to go and go to your podcast listening app these days unless you have a lot of exclusive content, which is why Spotify, I Amazon, they all recognize why Spotify was buying a lot of talent, and so we started making some of our own podcasts. Caller Daddy who was actually one of the ones we started on our own. It didn't work out.

There were some contracts Snap, so they did four episodes with us. We got a couple of thousand listeners. Then we wanted them to be exclusive on our app for a year. They were like a day max and then they left, you know, reservice to bar Stool a bit later and the rest the rest happened.

So it was it was a real challenge and they learned. I think it was a good idea. We wanted to have a podcast listening app that provided great discovery. People would come to it for that purpose, and then what we wanted to do was dynamically stitching ads for listeners based on the content they were listening to.

So if I was listening to a bird watching podcast, or you're listening to a maybe a Living Life as a Frenchwoman podcast, maybe Croissant Company might want to stitch in an ad right right and therefore, so we wanted to do that. But the issue we sort of found there was there were so many hurdles we had to get through. The first which build a good player, then get people on it and build the a. There were so many hurdles and we just fell off early and then we ran out of money and I had to go back to the drawing board.

So I just want to know if I win the prize for the most interesting pronunciation of this business, I think it it gives us a I like the Yeah, I think a certain certain sexts of America may really you know, the Christian or sex. I really bought it. Maybe we should, maybe we would have been more successful with that, so it could have been Yeah, yeah, I was really impressed. I'm like, wow, they're like pod Bible pretty so obviously I was not reading carefully.

Yeah, our marketing must have missed you back then. Yeah, yeah, probably somehow. So pod scribe, whatever the hell you call your first business podible it's dead now. It's dead now.

So they're platforms in the podcasting industry, Like, when did you decide, yeah, you know, this is a great industry. I want to be here. Truthfully, I I kind of just got drawn into it from the beginning. I was sort of questioning how into it I really was at the start with with podible Bible, I was.

I wasn't actually that I was the CTO then and I was. I felt like I was. I was kind of into it, you know, I like the idea of building something, but I didn't, you know, know, how into it I really was. But I'd say towards the failure and then the transition to to pod scribe, I even even then, you know, I wasn't.

I just had a lot of experience and in the podcast world, so I kept plugging away at it, But I would say increasingly over time I have grown to become more attached and like more and more the industry, particularly was going to events like podcasts, who went to see people in person, to work with many of the you know, larger figures in the space who I really enjoy working closely together. So I would say at first, I just sort of randomly happened into it because my boss, you know, because I just wanted we just wanted to want it to do a start up.

I didn't really hit know what to do. But increasingly, as I meet many of the people in the industry, become much more familiar with the industry's challenges and opportunities where I think it could go next, and build personal relationships. I definitely find the industry more and more attracting. My wife, who was selling HVAC in New York as a sales engineer before, always remarks when she meets podcasts industry people, how much nicer they are than the contractors in New York she always have to deal with.

So yeah, I mean, that's definitely. I think we all find that it's a pretty even though it's a competitive environment, it's very cool angel So I think people are always willing to support each other. But you said one thing that I think is really important, and this I know about you speaking to a member on your team, is about your focus on customer needs in terms of how you think about creating product. Can you talk about that a bit, because I think it's the right way to go.

Yeah, I would say so. One piece of context about us that podscribe is we've never we're not a VC backed company. We've raised I think like two hundred thousand over a couple of years from boos from angels, So all of our funding and livelihood is more or less always come from from customers, even from the very get go. So you know, we kind of didn't start out with a two million dollars VC check to then say Okay, we're going to go build this crazy thing, and like you know, it can be a bit more divorced from reality what the customers want, but you know, we'll make it work in the year.

Don't worry. You know, we'll make it work. We haven't quite had the luxury of that separation, but I think I think it's been quite healthy for us to have our livelihood be fully dependent upon serving customers as well as we can. And one thing, I think it's a bit of a spectrum that I think you some people could argue and perhaps do that we maybe going to too much.

But I think it's a trade off between what maybe we have is our grander plan to march forward on and then how much along the way do you a customer wants this, a customer wants that, or is it pulling? And so I think it's in truth you need both. I think, especially at the beginning, we were we were like, we'll just do anything, you know, like, oh you you can need air checks? Yeah, like that, Okay, maybe that's what we would we would focus more on that, but I think so, especially at first, we would.

I wouldn't say I had there was no plan or anything, but it was the plan the waiting of, like the where I thought we were going was much more malleable based on customer impact. And I think over time, I think we've sort of developed a bit more of a point of view on where we can and should go. But I would say that's largely informed by summarizing and hearing customers. So but yeah, usually I think we are focused too is talk to the customers, get their feedback and this build is absolutely as quickly as possible and then go back to them because they're always going to want something more.

So it and I do think most of our I do think it stands out with most of our customers. I would say one challenge thinking about a lot now is as our team grows, right at nine or ten people, now is how do we make sure we do not lose that especially as you know, more people come in and it's things get we're a little more diluted. How do we maintain that? So that's one thing I think a lot about because I agree with you.

I think it's essential. And when companies get a bit detached from what the customers want, right, that's sort of when I think they begin to die or lose ground, and so they can get back on track. So we have a technology platform, excuse me, and we created a product roadmap, so yeah, we listen to our customers. We've made a ton of adjustments along the way.

But to your point, we sort of have a base of knowledge as to where we want to go with this thing. So as much as we want to be responsive to our customers, we also have to make sure that we are, you know, building towards what we ultimately believe will be the right course of action. But it's really challenging. It's just a very delicate balance.

It's a it's a real tradeoff. It is in one thing I think that that can can help minimize that. The tradeoff is just doing everything faster. You know, it's easy.

Of course, it's obviously maybe easier said than done. If it's like, oh do I do this roadmap thing or just customer request, Well the ideal answer is both, so if you really can. So that is another big thing. I think that's that can allow you to seem and be much more responsive and customer centric while still going towards a bigger plan.

It's just moving faster. So we've developed in a number of things. I think that that help us move faster. But this speed I think can fix a lot of things, a lot of things moving really So yeah, yeah, I think it's a big component.

I can attest that you guys are really quick and very responsive, So thank you. So you know you're growing this company and as you're going through it, any like overall advice about your journey as you are growing and developing the company, or any sort of anecdotes or just things you've noticed. I would say the company startups only die when the founder gives up. I would say, and there's definitely times I think to give up, but I think a lot of it is just you keep trying things and improving.

Something's gonna something's gonna stick eventually. And I think too what I think one thing is I didn't make this up, but I would read and hear people say it earlier. Is but I think it's helpful as you need to have realistic expectations for the pace of things. Because if you think I'm gonna go create a startup and then in six months I'm gonna release this products, I'm gonna be making like one hundred grand a month and everybody's gonna be loving it.

Maybe maybe I'm not going to say it couldn't happen, and it certainly could, but I would say odds are it's going to take a lot longer and be a lot more difficult than you think it will be. And if you have the expectation that you need to hit these one hundred k a month in a year and you don't, and then you're just demoralized and you give up and think you're a failure. I think it's easy to based on seeing and talking to other people, maybe you know, like here, maybe I'm kind of it sounds like the highlights are also great.

But the reality is, if I've been doing this for like six years, we didn't make any money until the first like three or four. You know, if I was like, well, we need to make one hundred k in the first year, I would have given up like five times. Not to say that was my expectation, but I think for a lot of people who maybe wanted to do it, I think that having the expectation that it's going to be so maybe even ten ten times, just pick a realistic expectation you think it'll be and then downsize it by ten or something something like that, you know, And then because the trouble is, I think if you have these expectations and then you give up too soon because your expectations weren't real because you had this warped view of reality that isn't actually true for most people, then that maybe is really what sabotaging you.

And if you just kept plugging away at it. So in a lot of cases, I think it's just what h what what you really want and your in your priority. So I say that's the biggest thing, the realistic expectations, and keep keep plugging at it, and that you know, if you if you want it bad enough and work out it enough over time, I feel like in most scenarios it can work out eventually. Is one just thought on on that that that I've seen at least in my in my and myself, and oh, everything is always slower than we want always, like the products, customers signing on adoption, the industry, like everything always just takes way longer, even when we think we account for it.

Is sort of like the paradox that that I see. I'm like, you know, but also I think because the industry is so young, I think we're all trying to get our bearings on which way to that we want to you know, drive this industry. It's true, I think in relative to other industries out there where advertisers are used to buying on, I think they come to podcasting and they feel like they have a third of the tools they need or a third of the data. And I think in some ways that's actually an opportunity for buyers and publishers and toolmakers like myself.

But I do think the industry is it is quite young, and it's sort of a peculiar one because it's an open source one largely, so that creates challenges because nobody can come in and just dictate change rapidly. You know, Spotify and Amazon want to and they try, but they get pushed back when they get too much of them. You know, when Spotify starts getting too powerful, people are like, I'm gonna, you know, pull back a little bit because nobody really wants that, which I think is good for the ecosystems to remain open like that, but it does keep the progress out a bit more of a limited rate.

We have bodies like the ib or sort of where we're sort of this collective group deciding things. But I don't know if anybody's ever heard of like a committee that moves quickly on anything or it's really that effective. Yeah, that's an oxymoron. Yeah no, no, no, no, all right, So let's focus on pot scribe, which is a wonderful product.

So can you talk about podscribe. Where'd you get the idea to do what you're doing and the features and benefits of podscribe. Yeah, So we initially with at Podible, we had started out transcribing podcasts to make a recommendation engine to listeners based on the content. So we always were doing transcripts of podcasts and that's sort of where the scribe the name is.

And at first we tried, I was like, who would really need podcast transcripts? You know, after the whole app thing didn't work out, I thought, okay, maybe deaf people, Like how did they listen to podcasts? Like a there'sn't a public thing of transcripts. So we started down that path, but it turned out I couldn't find or you know, I could are not doing people really who were deaf wanted to pay a subscription for transcripts.

I think it was also illegal too, because we were monetizing other people's I didn't realize. Still later I think it was also illegal and nobody wanted it. That was like the starting point after that, and then from there went to transcribing. I thought, okay, well maybe if I, you know, I did the thing people aren't you're not really supposed to do, like you get emails in the RSS feed and then like you send a mass email out like we did that a couple too.

After the death. I didn't really know, I wasn't supposed to do that either. It was like back in twenty nineteen or something. And then I thought, you know, maybe if I email a couple hundred or a thousand people, some podcaster will want like their transcripts for their own podcasts and they'll pay subscription.

So did that and started getting a couple of podcasters who wanted to pay fifteen twenty a month for the for monthly subscriptions, and I was mind blow. I was like, wow, people are like paying for something online. And that worked, you know, got up to a couple hundred. But the trouble was, is this difficult to get new customers and podcasters they churned, you know, they stopped their podcasts and it kind of hit a wall where the percent churning, we're equaling our growth, so we couldn't keep growing.

And I was like this, you know, was not very high, you know, a couple of grand a month, but I always thought it was great. So then we started catering more towards advertisers to find the ads in the podcast that they needed, so air checks because that so that was actually a much better need because advertisers really needed that. The publishers too, certainly just to automatically find those ads there. So that was a big unlocked because advertisers they're constantly, you know, they're buying on dozens, hundreds agencies, maybe thousands of shows a year, so that was a lot of transcerts and they really needed them.

And often these clients would be growing too, so it wouldn't be you know, they wouldn't go out of business too often at least, so that was a much better switch. So that was really the start of our enterprise world. The air checks we made a bit of a competitive Intel product, So I would say the air checks, finding the ad and reporting back to whether it ran correctly, if there were certain issues, maybe it was too short, the mid role was purchased. It somehow got put in the beginning of the episode.

It's near a competitor of the promo code was wrong. We began reporting on that because agencies would tell us, they would show us. They'd say, look, we got like a team of twenty interns in a basement listening to these ads, like going through every single one, like checking these things, and I was like, it's a common everyone you have to So we're like, okay, you know, we can automate some of these pieces like some you know, not everything, but we can automate a lot of these these checks.

And then we added you know, check by check to automate, so we that that, i'd say, is really the foundation of what podscribe does. The verification, finding, finding the ads, making sure they run correctly, and reporting back to the advertiser so they can know and then also the publishers they can serve you better inventory, get ahead of issues, fix them sooner. And that that's really the foundation for everything past that, I'd say, what is more recent piece in the past year and a half is the attribution side of things.

So not only how did I buy run correctly? Did I get what I what I paid for? But then also did it work? So naturally these two go together because in some cases, when you see the performance, your next logical question might be, well, how can I improve it?

And if we have flags saying oh, well, your ad was behind three other ads, or is near a competitor, or the impressions were being served to a small percent of listeners, or isn't the wrong GEO, we can have a whole list of flags and often those can be a blueprint for how it can improve. So did it run correctly? And how did it perform? And then kind of repeating the loop, I would say, is the podscribe primary offering right now?

So it's interesting on attribution because when I started to think about, you know, pulling us together, which is around twenty nineteen. Yeah, in my background is corporate, so I would ask the question, well, how do you know if an ad works? Like how do you know you're spending money? And this concept of attribution was pretty new in the industry, which I found astonishing.

So not too many players at that point and still now really, I mean I think you're you're there are there are a few others, but I find it really hard to wrap my head around it because to me, I would think, if you're spending money, you want to know what the ROI is. So I know that's super important to USA, to ASA, But I have a question about AI. So how does AI come into it? Like did you teach yourself AI?

I know you went to school for engineering, but like where, like where does this AI stuff come in? How do you do that? Or not specifically, but I don't know how you do it? Yeah, it's yeah, Yeah, it's certainly been a hot topic year with which chat GPT as people I think it's been getting more brought into people's attention as they may tinker around which chat gpt and see, wow, this thing can like talk back to me, you know, maybe hallucinates sometimes, but it can provide reasonable dialogue back and be helpful for for something.

So we use it in a couple of ways. We've we've we've been tinkuring out tinkering in a lot. I mean, i'd say the first and most obvious way as we use it to convert the podcast from audio to text, so just to transcribe them. We do have our own internal transcription engine right now.

We probably rely primarily though on a third party for that just because there's is like slightly more accurate, although decently more expensive. But we do have our own for that. But that that is one area. I would say.

The next area is auto detecting where the ads are in the transcript. So with with that piece, we it is a mixture of it's the AI of there is definitely not perfect, so we do have a mixture of virtual assistance who will also help clean up the exact start an end. But that is one area where we're using it and looking to use it even even more another area which could be its own whole subcategory, but under verifica, under verification, did my ad run correctly? Sort of how I viewed a subcategory of brand safety and suitability.

You know, if I'm an advertiser who said I don't want to be on political podcasts hypothetically, and it ran on a political podcast, we would say we would flag that and say that didn't run as or it didn't run as you intended it to run. So we do. We use chat GPT to assign brand safety and suitability labels according to the GARM framework. So the GARM framework it will have I think there's eleven twelve different categories from profanity, drugs, sex crime, hate speech, terrorism, political topics.

So there is a number of different categories and then in each one will score it from low risk to medium to high risk, not necessarily depending on if the content is there, but depending on the how the content is being discussed. So, for instance, say there's a the category is sex hypothetically, it's a biological one talking about you know, biological parts and how it happens. It's an educational video that should be a low risk piece of piece of content because it's in an educational and informative manner versus something that could be maybe as far more graphic and descriptive, that could be on the higher end of the spectrum.

So one I think big area where we are focusing on with the AI using chat GPTs. I think what I think the hardest category is and that's the political and controversial sensitive topics category, because I think many advertisers, perhaps rightfully so, say they don't want to buy on news podcasts these days because they're too risky. And maybe they are too risky in a lot of cases, but that makes it tough. Men.

If you're a news podcaster and you want to put out responsible content that people want to listen to, how are you going to make that sustainable? So with the AI there, it does a decent job at distinguishing more fact based new news podcast versus more opinion based one. So I don't think it's perfect, but it does a pretty decent job and it'll give a reason for why it assigned every risk level. So brand safety is another big one.

Excuse me, I would say there's a couple more topics like finding similar shows, how are you up? AI? So there's a number of ways were I would say chat Gypt changed a lot of things, and we're revivating more or less everything. But some of these cases are better than others for it.

But the space changes, I think quite quickly. Some things are always have been the same for a while, but some things are rapidly changing on what's available. So I have a question, have you got the requests for host safety? We have, and we have started.

We actually had a version of that out earlier this year, but we were using Google's barred for that, which is sort of similar to yeah, and that's the trouble. We had it out for a bit, but then we started noticing pekiliar things where it would just hallucinate and just make up something totally and then we're like, okay, well now we need to tell it to put in put in links, like you only include something if you if you have a link for it, So then it would start including links.

But then what we noticed is none of the links would be valid. They would all be fake links, and then it kind of just became like a little bit too much. I think we need to wait for that tool or another one to become better. But the reason we use barred for that is because Bard supposedly has more current data in it, so you know, something changes with the host in the past week, you might want to know about it, But which I should be t it's you know, it's not as updated.

So I think we're gonna there are ways we could you do that more on our own, but for now it's a bit less of our priority. It's it's also a tricky, a tricky thing too, you know, with the with the continent. It's one thing I think we we do want it because it's definitely a thing advertisers want to go into. I'm sure maybe publishers to a degree, you kind of want if you're bringing on a new show to the network, or if you're an advertiser buying on it and that host does something crazy and you get in trouble, your boss is not going to be happy with you, so you might need a or somebody's not gonna you know, somebody might not be happy and so you may have wanted that information before him.

So it's certainly understand the rationale for it. So what we do is we uh curate our podcasters that come on our network, and I won't put on anyone that does haste speech or focusing on guns or stuff like that. So that's one way that I sort of, you know, do my own host safety. But it's really hard to say, you know, there could be just a lovely podcast and then you find out, oh my god, what what the hell does she say?

You know, on Instagram? So I understand it's probably it's probably incredibly complicated, but I bet you're going to get, you know, ongoing more and more requests for it. I have a feeling, I think I think you're right, And you know what. The tricky thing is with putting out the content rating in the host safety is an internal debate we have where should our host controversy and brand safety ratings match what the advertisers think, you know what, what what they think it should be, or do we you know, because maybe somebody says something about a topic that is sort of a hot topic these days, but maybe it should maybe be more spoken about.

And so our dilemma is do we just sort of go off what the you know, quote unquote status quo is today, which is a lot of times right, but a lot of times may have flaws. So do we anchor the ratings on what everybody expects in the current status quo is or do we shift it slightly in one direction and have a bit more of a point of view. So that's a very tricky thing that we're thinking about, and I don't know if we have the you know, the right answer for it. And I think every Branson the company kind of has to have a point of view.

And so it's one thing we're we're thinking about more. I don't you know that that to me is going to be just so darn hard. So yeah, good goes to you for even trying to figure it out. But it's so subjective, even more so than brand safety.

But let me ask you a question about getting back to brand safety. Are you getting a sense from advertisers that there's an increase concern about brod safety or is it sort of like stabilized. I think it ebbs and flows. I think depending on like what if there's a big thing for Joe Rogan, it may spike up, or if there's a certain controversy, you know, I think it ebbs and flows to a degree.

My perception is brand safety. Nobody would say it's not important. I don't. I don't think, but I wouldn't.

I don't think from because we one knock on us, I think is we have a lot of products, so we have a lot of things we could talk to people about, and so we kind of get a sense of their priorities based on it. And nobody would say brand safety is not important, But I generally feel like it's the thing people want to just have working and then like not think about it so much. They just wanted to like more or less work, and I want to look at performance in the shows is generally more how I think about it and how I think people think about it.

So time to time, I think it's you know, we're checking into and it's not Nobody would say it's not important, but I don't think it's something everybody wants to be dabbling in every day. You know, they want to like they want to be looking at, you know, performance in the shows much more so. So that's why I would never I wouldn't say we're a brand safety company. I would say it's a component of our verification that we that we do.

But I think more important is did my body run correctly? And then how did it perform? So yeah, on that topic, in terms of attribution, how how did it go? You know, did my add generate revenue.

Are you so, is there more work that you guys are doing with your attribution capabilities? You don't have to tell me specifically, but like, is there more? Are there more? Is there more refining on?

Yeah? Like what's up with We're done? It's perfect, it's perfect? Oh yeah, okay, great, Okay, nothing's done?

Okay, no, no, no. I would definitely not say so. I would say the attribution I think. I think a lot of advertisers don't believe any attribution in the space for good reason.

Now that I know how it's built, I would agree with them. In a lot of cases, I think they're Pixel based attribution came in hot a couple of years ago and pot sides you know, introduced it. But the thing is, I think increasingly brands have become more wary of it. And how I always explain it to the brands is like we can pixel based attribution is is useful, but it's also so are promos and survey data.

It's all of a piece. It can help triangulate your the final outcome. So promos are incomplete, you know, maybe half the people use a promo or maybe less the surveys not everybody fills it out. You know, pixel data, There is modeling involved, and there are issues, So I think it.

I wouldn't say if somebody says it a magic bullet, definitely not. I would say can it help provide unique insights and triangulate buys? Yes, for sure. And I think if I was an advertiser, you know, why would you want to have less data not more?

Especially in a lot of cases where can provide blind It can cover blind spots. But I would say, one on one key piece, I think that we're working more on to develop what I believe will be attribution that I think people will have a lot fewer qualms with is more incrementality based attribution. Where it's because right now how the attribution will work is let's say you, Marlo, listened to Tim Ferris here an Athletic Greens ad in it, and then within thirty days you go purchase Athletic Greens on the website.

We would say, oh, great, Marla purchase we saw it. She heard the ad, then went to the website and purchase conversion within the window the thirty days. But the trouble is with that is that you might have gone and purchase anyways you might you might you might have already you might be purchasing athletic greens on your own every six months, you know, going and buying a bulk shipment. It could have just spend that time of the year for you.

And so this is a problem that we see with some of the bigger brands, some of the smaller ones who know that many people are buying on or maybe they're not that into podcasts yet it's not that big of a thing because they're baseline is closer to zero. But for some brands where maybe a large portion of the audience is already buying or they're quite popular elsewhere, it can inflate the numbers. So we have ways to increase, you know, pear down, reduce the number of conversions that are attributed where baby people came from Facebook or they purchased before.

There's a number of cut the conversion. There's a number of ways you can tighten it to the more confident conversions. But I think the the flip that we are working on and have going for a few clients is more of a comparison where we say, take Marla and then let's find a exact twin of Marla somewhere in the universe. The only difference.

The only difference is the other twin did not hear the Athletic Greens add on Tim ferris On it now did, And then we say, okay, did the twin go and purchase Athletic Greens too? Or did the twin not? So if the twin went and purchased Athletic Greens as well, and you did, we'd say, okay, well that she wouldn't have done it anyways, probably don't count it. But if you did in the twin didn't, and the only difference was the impression, then we can say it with a bit more confident.

So I think that's where we're focusing a lot more and we're yeah, we're definitely excited for it because I the attribution I think is. I think it can be held in a lot more confidence by almost almost all perhaps all buyers, with the new approach, which I think in turn will help podcasting secure more more dollars. It can be proved quite confidently, and I think if it done right, I don't think we're quite there yet, but I think if it done right, I think I could go in a room and present to anybody like their buys and if it was working, I could pretty confident.

I'm pretty sure I could convince them that it was working based on the results and there wouldn't be doubt sold. I mean, yeah, that's that to me is incredibly important. It really is just to be able to you know, because we are essentially an agency, so to be able to say to an advertise or like, really, seriously, this did X and we are We're confident to that. So I look forward to all the tweaks and the evolution of attribution.

But I never could understand why this wasn't one of the first things that was dealt with industry trick, It's true, it's yeah, it's just tricky. I mean, in promos are are solid, there are a solid way to do it, but they have a lot of holes and issues too, and I think people know that, but there's still a solid way. But yeah, I just think with current pixel based attribution, we don't hear anybody who says like they love their attribution provider. I mean, I'd like to thank us.

You know, people are more confident in but I know still there are there are doubts. But I think we can get to a place where the doubts are much much lower with with this new approach. So that's that's really where the brunt, i'd say, of our focus and brain power is going. So when you say, like, you know, how important is brand safety, it's not not important.

It's not like you know, it's not the thing, like the biggest thing. It's like this is, I think is much more important. I agree, at least to us. Well, that's good to hear all the work that you guys are doing on it.

That makes me happy. So last question a little crystal bowlish ish, So what one change would you like to see in the industry? And it's kind of a vague question, but there is there anything that kind of maybe bothers you, you kind of wish it was done a little differently. M hmm.

That's a good question. I maybe the audience might I think this is a stage because I think it it. Also, you guys are doing this pretty pretty well. But from we work, maybe I'm biased too because of our client tell about.

Most of our clients are performance based advertisers or agencies, so we hear a lot of gripes with podcasting going to DAI an impression based and is becoming more just a endless stream of radio ads that everybody tunes out to. And I guess I can see, you know, from a publisher perspective, how maybe if it streamlines operations and there's certain things that can be easier about it, I guess I could certainly see the appeal, but at least from all our clients. I I don't think there's a single exception for our d our clients who would who wouldn't say I just want to baked in host tread ad over anything else.

And I do think in theory DAI, I don't think anything in theory there's a reason why d I has to perform worse. But I do think there are a number of challenges with d AI ad serving that causes DI campaigns to perform worse, like the number of our we see advertisers well by you know, faked in campaigns, which are fine in theory, where the advertiser buys a single episode, but instead of the ad being truly embedded in the audio, it's dynamically stitched in each download, which a lot of cases that works totally fine, is good and maybe gives the publisher more control.

But then what we see happen to a lot of the time is the ad will not be inserted into every download of the episode, and the advertiser publisher won't won't realize that, and then they have like two thirds of the downloads and they're like, why do my campaign. This happened to one major publisher recently where they switched from baked into faked in and then almost like a half dozen of our advertisers, we're saying, like, why are we sing performance? This publisher drops so much, like have you seen anything pete And then we're like, yeah, we're not seeing the ad inserted on like a third or half the impressions for it.

So that's just one example of several where I think in theory dais, you know, has a lot more potential and can do everything Baked in camp. But I think there are certain issues that advertisers are not always aware of that happened that in that impair DAI campaigns, causing them to just perform worse for whatever reason. There's several of them that I think caused the industry to be limited because then new buyers come in, it doesn't really work so well for them. They don't get performance they need.

So I think these so I'm not saying I necessarily have a you know, a total light eyed approach. Everything needs to go back purely to baked It and host read, but I think they're in some cases the industry may slow down its own growth with advertiser adoption if it's more of a churn and burning thing where advertises come in that doesn't work, and then they're like, try podcasts, they don't. They don't work for me, So I think, yeah, I think ad capping is should be mandatory, and not all hosting companies offer that.

So, you know, we had a situation where it looked like a podcast was going through blowing through the budget to fast. So we're able to pull her back and fix that. But that can be a problem because if you're listening to a podcast religiously and you're hearing the same ad like a zillion times, that's just not fun. Or like the same ad in a bunch of times within an episode that's not fun either.

That makes me crazy when I see it on cable. So anyway, Pete, this was really interesting. I could like geek on this for quite a while because this is sort of really you know, this is a black hole for me. But the technology is so interesting and I think it's really going to change our overall industry.

So I'm glad we're working together. I should say that to everyone and look forward to meeting you in real life. And thank you for appearing today on the Mind of a Mentor. Appreciate it well.

Thank you for having me, Marla had a great time, and thank you. And yeah, we created Mind a Mentor to help you along your podcast journey. To stay up to date with the latest episodes, hit the follow button wherever you are listening, and follow us on social media using the links listed in the show notes. If you're a podcaster looking to monetize your podcast with advertising, go to ascollective dot com.

That's ossacollective dot com and apply to join over eighteen hundred podcasters that are part of our community. And if you are a brand looking to connect with a female audience, Yes, Wemen control eighty five percent of US household consumer spending, so follow the money again, go to as Collective dot com and click on the brand and advertiser link. And finally, if you enjoy today's episode, please let us know by leaving a review. Thank you and we'll see you next week.