← Marketecture: Get Smart. Fast.
Loading summary
Transcript28 lines
- [00:00]
AdTech God
This is adtech God, and I command you to listen to this house ad. So if you're listening to this show, just know that you've really stumbled upon a giant network of content across advertising, marketing, media, publishing, and of course, the people that work in this great advertising industry. So go to market, check out all of our brands. We have multitude of shows from the Brand Forum, the Advertising Forum, the Monopoly Report, the Ad Tech God Pod, the Market, and more. We are bringing more podcasts to our network. We are consistently and constantly bringing on new shows. So check it out. Market or search for any of those brands in the app that you use to listen to this podcast. Enjoy the show and see you all soon.
- [00:59]
Ari Paparo
This is Ari Paparo. We're happy to present a conversation that took place at Marketecture Live this past March, 10th and 11th. Now we're in the podcasting business here at Marketecture, so this hits close to home. It is called the state of audio measurement and what the data reveals. And it's a conversation between Pete Berzinger, the founder and CEO of podscribe, and Matt Drengler, who is the head of partnerships of podscribe. They present quite a bit of data from their analysis of podcast audio listening should be super interesting for anyone in that industry. Hope you enjoy.
- [01:32]
Pete Berzinger
All right.
- [01:32]
Matt Drengler
A little daft punk. Am I right to get you, Little daft punk.
- [01:36]
Pete Berzinger
Hello, everybody. Thank you for for joining us here. And I guess we'll see you shortly for everybody walking out when you come back. So we're here today to talk about audio and audio measurement specifically. To start out, let's do a quick poll. Who here in the audience is currently buying streaming audio? Can you raise your hand if you're buying streaming audio? Not tv, but. Okay, a number. And now one more poll. How many here in the audience are actually buying podcast ads? Okay, we got a little Apple.
- [02:18]
Matt Drengler
Okay, all right.
- [02:19]
Pete Berzinger
A couple. A lot of overlap there. So. All right, well, maybe this will resonate to some degree with you, or maybe this will all be what you already know. So how to buy like a pro. And what does the data tell us? Quickly, who we are and why we're giving this presentation to you today. So we are podscribe. I'm the founder and CEO Peep. We measure ads for, I'd say over half of audio's top advertisers, and that covers both podcasting and streaming audio. Despite our name, we do measure outside of podcasting, although podcasting is our. It is our roots or where our roots come from. So we also do Measure streaming audio radio in a bit of CTV and display and video too. So and those are a couple of the brands we work with. And moving on. One thing that many of the audio buyers in here know today, maybe the rest of you know as well, is that audio today is digital. There is a perception that has lingered for a while and maybe was true at some point that audio specifically say podcasting is an offline channel. You have to use promo codes and surveys and you can't hook it up to your mta. You can't get impressions fed into your MMM properly. You have to do all these weird hurdles. I'm here to say that's not not true because we've seen it with hundreds of the brands we work with, how they can measure audio digitally like every other channel, for example, there's a real time dashboard and API. You can get user level attribution. There are some limitations with it. It is IP address based, but it still can be done with great success. We have always on incrementality in the platform. You can see as well which segments in your audience worked. So maybe you, you bought an ad on Joe Rogan and overall maybe it wasn't great, but perhaps you saw that plumbers in Missouri did well. You can actually see the results segmented and know which segment maybe did better than others to then adjust and optimize for your next buy. So and also, yeah, integrating with an mta, your internal MTA or MMM is a cinch. There are many brands that do it, so. And also there's global coverage too, not just in the US. GDPR is of course tricky, but it can be done in a privacy safe way.
- [04:48]
Matt Drengler
So perfect. So let's dive into some numbers so we can all kind of buy like a pro. Let's first at least set the stage a little bit about what's happening in podcast land. I'll give you a little bit of maybe context. I started focusing from a business perspective on podcasts about 13 years ago and the ad spend in podcasts in the US was $150 million annually and today we're at 4.7 billion. There's a significant amount of growth there. When you look at podcasts and streaming audio combined, it's in the 10 to 15 billion dollars range. We're going to go through a lot of our data that we pull and put together, but some industry stats that might be beneficial to kind of put things into perspective. So in 2025, almost 600 million people listened to podcasts worldwide. Pretty decent size, 73% of Americans watched or listened to a podcast, so pretty decent reach. When we talk about the performance of podcasts, we've seen about 44% of weekly podcast listeners have confessed that they have purchased after being exposed to an advertiser in podcast. That's just the people willing to say that they did it. And when we look at some of the disparity, right, when we talk about the amount of time that people spend listening to podcasts, you know, it's like 10% of their media consumption. And we're seeing it's about 2% of advertising spend. So there's a lot of opportunity there. And to showcase some of that opportunity, you know, part of our data set is we look at and we download and transcribe the top tens of thousands of podcasts out there. And as a part of that, we can spot the ads that are in there. And when you start spotting the ads, you can start calculating the ad load. So what does ad load really quickly? Maybe we already know this, but sometimes you got to explain the graph so they make sense. You've got 100 minute long episode of Smart Lists. Insert your favorite podcast here, and five minutes of that is ads. 5% ad load, and you start to see that trending upward a little bit. Right? So Q4, 20, 25, we have about 11 and a half percent ad load. Can we compare that to TV, which is what, like 19 minutes for every 16 minutes? 60 minutes of content. A lot of room to grow. Water's warm is what I like to say. So definitely something that is worthwhile paying attention to and drives significant performance for advertisers. Let's dive in on what that means with real numbers.
- [07:05]
Pete Berzinger
So you're saying, Matt, don't be alarmed by the rapidly alarm.
- [07:09]
Matt Drengler
A lot of room left. Yeah, yeah.
- [07:11]
Pete Berzinger
So we publish a report every quarter called the POD Scribe Performance Benchmark Report. And in it we analyze the campaigns, all the campaigns that we're tracking to try to provide to the industry as many insights as possible for how to optimize performance and buy better. Today, we're going to go through a couple quick insights, so hopefully you can walk away with something. Maybe some things may be transferable to other channels, maybe not. But to start out here in this table, we go through what the average conversion rates you might expect for if you have, say, a thousand impressions. So just really quickly on what is, say, a conversion rate, because that's probably the most important one there. If you have a 1% conversion rate, that means if you have 100 impressions, you can roughly expect to have one attributed conversion. So for podcasting compared to streaming, podcasting in the blue, streaming in the red, you can actually see podcasting actually has a more efficient conversion rate than streaming audio. We'll get into a little bit on why we, we think that is. But interestingly, when you go and then compare the cost per visitor and cost per acquisition, when you actually factor in the CPM for those, for those impressions that were paid, funny enough, the market, it shows that the market has basically balanced it out. They have pretty similar cost per acquisitions and cost per cost per visitor. So we also have in in our report too what you can expect per advertiser industry because these are just the general ones, of course. But yeah, these insights, at least in our last campaign were over across over 30 billion impressions, almost a thousand advertisers, about a hundred thousand campaigns and growing. So you can scan it if you want to get the report.
- [09:04]
Matt Drengler
Nice. Let's, let's dive in on some specific insights. Host red ads convert better and are more efficient in customer acquisition costs. So this is when an advertiser says I want to get into podcasts and my favorite podcast is insert your favorite podcast here. Is it Joe Rogan? Is it NPR up first? Is it smart list? Whatever it is, they want the host to read that ad. They want Jason Bateman to read that ad. And what we find is, makes sense. Yes, you should absolutely do that. When we look at purchase rates compared to host red ads versus let's say producer red ads that are typically dynamically inserted into a podcast, every download might have a different ad based on the targeting parameters. Ad servers like we're used to talking about here, hostred ads perform slightly better in terms of purchase rates or response rates. Looking good. When you throw the money into the equation as well, you see the cost per acquisition also a little bit better. So host red ads. Should advertisers do that? Yes, definitely. That's a great usage of the, of the spend. Let's dive in more.
- [10:02]
Pete Berzinger
Right. So long as the, the price is not too exorbitant. But yeah, if the price is roughly equal, it seems like the data definitely shows it makes sense. Another similar topic that's happening a lot in podcasting specifically is whether you should buy a single show or you should buy an audience across a bunch of shows and show specific buys where you say you buy just on the daily or up first by npr. Those have been the norm in podcasting I think for since the beginning. But with evolving ad tech, it is possible to buy across many shows to do audience Targeting to do geo targeting across many shows. So what we're comparing here as well, how do those buys compare when you buy a single show versus across many shows? And what we see here in the purchase rate, the efficiency per impression, the show specific buys definitely still seem to outperform the non show specific buys where they're buying across many shows, either programmatic or random network buy. But interestingly here actually the CPA for these show specific buys, the cost per acquisition is higher, at least per this data set, indicating that maybe actually people, when they buy programmatically podcasts through say a dsp, then actually the cost can be so much lower it's actually offsetting the advantage and then some that you could get from buying a direct show. So something to be mindful of that. Yeah, the host thread single show ads are the golden standard per impression, but it's worth making sure that the pricing makes sense for that performance. Excellent.
- [11:43]
Matt Drengler
Okay, let's talk about where in a podcast you would want to land your brand on to drive the best performance. And maybe not surprising earlier in the episode is better. You know, the way that typically buyers are purchasing this out in the marketplace, they're looking at buying like hey, I want to be a buy a pre roll or a mid roll or a post roll. And we don't really subscribe to that terminology here because the way that one defines a pre roll is different from one company to the next. We might define a pre roll as the ad has to hit within the first minute of the content, but a lot of shows are kind of changing that up. They have a minute or two or three minutes of content and then they insert an ad. And so to make things a little bit easier, let's talk about ad groups. So here in the ad groups, you know, we've got one through five. It is very much so possible that there are multiple ads within each one of these ad islands. All of that being said, you can see in here earlier the better. So we see that conversion rates stay higher farther left and the cost per acquired customers stays lower to the left as well. Now the interesting thing about podcasts is impressions equal downloads. That's an interesting concept because someone could download a podcast but not make it to your ad. And that's why it's so important that being upfront as proven by the data, seems to drive better performance for those more performance branded advertisers.
- [13:06]
Pete Berzinger
I wouldn't say it's necessarily a dirty secret in podcasting, but if you understand that podcast impressions are downloads, you're probably ahead of many or most podcast buyers out there. And it's not necessarily, it doesn't mean, you know, you can't use them. I mean, for performance marketers it doesn't necessarily matter. You're going off, you know, how many conversions and what, you know, the roas or incremental, you know, cost, cost for incremental acquisition was anyways. But it is definitely worth worth noting. So one thing that may be maybe somewhat unique to podcasting is the performance per ad length. And what we find in podcasting is actually the longer the ad generally, the better the performance.
- [13:51]
Matt Drengler
What?
- [13:52]
Pete Berzinger
So yeah, believe it or not, so both per impression, you can see the conversion rate is rising, going from 0.32, then kind of a dip all the way, you know, up to 0.36 for 90 seconds plus. But yeah, also the cost per acquisition sort of trend trends a bit, a bit downwards generally from 30 seconds on. And we think the reason for this is it goes back to the, you know, the host read by that if a host is reading an ad in podcasting, many times it's not, it doesn't necessarily feel like an ad to you. Maybe it can be quite, it can be quite engaging and draw you in if it's your favorite show. Whereas perhaps in other channels, you know, if there's a commercial droning on, you're maybe just you get annoyed and want to skip it and so on.
- [14:38]
Matt Drengler
I would also say too, you know, the, the hosts that are reading the ads oftentimes inject their own personal anecdotes and you might buy a 60 second ad, but you get a three minute ad, which is not uncommon. So that's, yeah, clearly the confounding variable in that this next stat is a no, duh. But we think it's important because it allows to give some credibility to the numbers that we're showing here. And if this came out differently, we would be worried. Here's the concept of what we're looking at. What we're looking at is what is the purchase index? So share of purchases divided by share of impressions based on wealth rating, 10th decile more affluent, first decile less affluent. And we start to see that these numbers are going up into the right as you get higher in the decile, almost as if people that are more affluent seem to purchase more, at least relative to the number of impressions that they're receiving. And so for those of us that don't speak in indexes, when a number is over 100, it means that it's above average. So we start to see that number go above 100 at about the fourth or fifth decile. So again, kind of one of those gut check things, which is like, if we didn't see this, we might be a little bit worried. So this allows us to kind of feel really confident in some of the insights that we're kind of bringing to you here. So we'll move on to the next one.
- [15:50]
Pete Berzinger
Or we wouldn't have picked it to put it up.
- [15:51]
Matt Drengler
Or. Yeah, we may not have.
- [15:52]
Pete Berzinger
Yeah.
- [15:52]
Matt Drengler
Or maybe just wouldn't have shown you that. Yeah.
- [15:54]
Pete Berzinger
So next we, we go into frequency here to. Because frequency is always a hot topic. How many times should you be serving listeners or consumers ads? And what we have here is a graph with the number of times a consumer in podcasting heard an ad per advertiser with the conversion rate. This is, this is across all our advertisers. And a lot of things, of course, could be hidden in averages, but generally, when, even when looking across individual advertisers, typically, you know, around five or, or even sooner, sometimes the conversion rate does start to, to really drop off. You can see it at 15 plus. It, it especially starts to tank. But generally in podcasting, I mean, this is, this is what we see. I think generally if people ask us what should we do, you know, what frequency cap should we have? And I typically, I say, just so long as you have some frequency cap, that's probably most important, just so you're not in the 15 category, which some brands just kind of inadvertently have a bunch of their impressions dumped there if they're not careful about it. So that, that's the takeaway, at least for me here.
- [17:05]
Matt Drengler
I would say too, the, the frequency cap, just to your point, just having a frequency cap, super important there, there was an issue when podcast land a few years ago where there were episodes that were being automatically downloaded on people's devices and it was being done in massive scale. And it typically would happen when, I don't know, let's say someone was subscribed to a show which, when you're subscribed to a show on the Apple podcast app, every new episode that drops, you auto download it. If you stop listening to that show after two weeks, it stops auto downloading. So there's that like two week gap thing going on. Well, there's also this other thing where let's say I stopped subscribing and then a year later I'm like, whoa, what's what happened in the Daily lately? And I go and I download the next episode, the Daily, and maybe like Oh, I liked it. And I subscribe it. And the app would automatically download every episode that you've missed since you were last subscribed. So just having a frequency cap. Just a frequency cap, even if it's at 15. Super beneficial. So anyway, I'll stop stealing the thunder on that slide. Let's talk a little bit about incrementality, because we just talked a lot about attribution. Raise your hand. Do you know the difference between attribution and incrementality? Everyone should know that by now, right? Maybe not. So, so here's the concept. I can say, where does the data overlap? Meaning, I have conversions that happen, I have exposures that happen. I have all these digital identifiers that connect the dots and that shows where they overlapped. That doesn't mean that some of those conversions didn't happen because of other channels influencing that audience or because of, you know, word of mouth or brand perception or other things that might be driving conversions. Insert incrementality. The concept of incrementality is we look at a control group. This is like group of people that have not been exposed in this instance to podcasts, and they're being exposed to everything else that a brand is doing, and there's some natural baseline of conversion rate that they're converting at. Great. Now we know that baseline conversion rate. Now let's measure the attribution for the exposed group. What's their conversion rate? What's the difference? And the lift is the incrementality. That's when you're a large advertiser, that's like the only thing that you really care about. Because if I'm. I don't know, insert the largest brand you can think of in your brain, there's plenty of them out there. McDonald's, Progressive, Pepsi, any of those that are out there, you think if they took it, all of the attributed conversions and all their attribution platforms and added them up, it would, like, quadruple their sales overnight. So that's why incrementality is so important. So we do this same level of benchmarking for incrementality. What we see in here, we break it out by advertiser industry. And again, we ask ourselves, does it make sense? So which industries are driving the highest incrementality in podcasts Right now? It's tobacco and marijuana. Is that weird? It's. To us, it's not weird. And the reason it's not weird is there's not many channels that those advertisers can advertise on. And in Podcasts, they can't. So the fact that they see so much incrementality, again, one of those gut checks kind of makes sense to us. There's a lot more here and we're about to get the yank. So I'm going to go ahead and move on.
- [19:51]
Pete Berzinger
When I first, when we started doing attribution a couple years ago, I didn't even, I didn't really know anything about incrementality, but I, when I realized something was wrong was when I forget I want. It was some big retail brand and we're doing attribution for them. And they had I think like a 78x roas on the campaign. And I was like, that doesn't, doesn't look right. And that was oh no, adx. Yeah, I mean it was a, probably a pretty good campaign, but that maybe, I don't know, many people have already had that moment where you see that for like in platform attribution it's just like what's, what's going on here. So yeah, that led us to also provide incrementality in the platform on, on this slide we've. When we break out advertisers by how big they are in terms of how many site visitors they have, we see a predictable decline in the incremental percent and incremental percent. On its own, it doesn't really mean anything. You multiply it by your in platform attribution or in platform roas to get your incremental roas. So on its own it doesn't, you know, lower doesn't necessarily mean bad because the lower 12 percenters might have, you know, 2020 x ROAS that they're multiplying it by. And predictably we see the bigger they are, the lower their incremental percents get. Because more of the population is already purchasing likely.
- [21:10]
Matt Drengler
Exactly. All right, really quickly, let's take a look at how the audiences are dispersed. So podcast ecosystem and streaming, streaming audio ecosystem very heavily indexed over indexing, I should say, in the top 25 DMAs across the United States. The this accounts for roughly, I want to say it's now I'm like fumbling over the numbers. Well anyway, podcasts over indexed on about a quarter of the DMAs and that's about half the population streaming under indexes that are over indexes on about a quarter of the DMAs. That's about a quarter of the half of the population, but it's not the same half. So the final thing here to kind of call out is those two different audiences have about 20% overlap. So if you're looking to drive significant reach in the digital audio space, podcasts and streaming together do a great job of doing that, and that takes us home.
- [21:59]
Pete Berzinger
All right, thank you all so much.
- [22:01]
Matt Drengler
Thank you, everybody. Thank you for subscribing to Market. New interviews are added every week at marketecture tv and your favorite podcasting app,