Loading summary
Transcript50 lines
- [00:07]
A
Most advertisers would say media quality matters, but what if we're still measuring it the wrong way? I'm Noorna Sear for Adtech and Filter, and today I'm joined by Sarah Miskevich, SVP of Market Enablement Verification at Protected by Media Ocean. We explore why the industry is shifting beyond brand safety towards brand suitability, where AI is fundamentally changing media verification and why advertisers may have more control over media waste than they realize. It's a conversation about what quality should actually mean and how that definition is continuing to evolve. Let's get into this episode with Sarah now.
- [00:41]
B
Sarah, I haven't had a conversation about media quality in a minute and I'm excited to kick it off with you today. Me too.
- [00:47]
C
Thank you for having me.
- [00:48]
B
As it relates to media quality, I want to get your take being somebody who spends your time and energy thinking about the topic. Maybe many brands say that they care a lot about media quality, but sometimes the actions suggest something else. Why does that disconnect exist?
- [01:01]
C
Yeah, I think it used to be kind of a buzzword. Right. I know that I need to be invested in quality media, but no one has necessarily taken a step back and actually defined what that means for their individual brand. The industry kind of guided them to, oh, it means less ivt, it means brand safe, but what does that mean to the brand? And then I think there's a lot of traditional KPIs that they are still oper with that don't actually equate to quality media. So reach, you know, scale, you know, those don't necessarily always equate to getting, you know, premium inventory or premium audiences.
- [01:38]
B
Yeah, I think about this too, that so often brand safety is the, whether people say it or not, it's the tier two objective and the tier one objective is efficiency and whatever type of roas you're going after. So it's almost like feels like I have ethical or moral clarity if I'm saying that brand safety is ultra important, but mysteriously, it falls a little bit
- [02:01]
C
onto the back burner a hundred percent.
- [02:02]
B
So a follow up here. The definition of brand safety, like you just mentioned, it's something that can dramatically change and changes from organization to organization, but I also think it's changed over the relatively recent past. What have you observed as holistic change when it comes to the definition of brand safety?
- [02:19]
C
Yeah, and I'll start with the actual definition of brand safety. I don't even use that word anymore. And I think the industry is already, you know, moving to brand suitability yes. They can be two different things, but brand safety traditionally is just very binary. Is this safe or is this not? And that's really just not how the digital media ecosystem works anymore. There's context, there's sentiment, emotion, you know, and things like that that are going to determine whether it's suitable for a brand, even suitable for a campaign, not necessarily for, you know, just the advertiser themselves at the holistic level. So brand safe is just a little bit too binary, and it doesn't actually work with how the ecosystem is made, but it also doesn't work with the unique brand values.
- [03:03]
B
You just mentioned a bunch of different signals that are out there that are a part of the suitability definition that perhaps haven't shown up with brand safety. What does that evaluation or consideration look like? Because we also. The previous question was how has it evolved over time? It sounds like the evolution also has to do with the fact that we can capture more signals for that consideration.
- [03:24]
C
Absolutely. Yep. So it's no longer, you know, keyword, you know, driven or domain driven. Not just, you know, typical inventory that you can buy or not buy. It's literally what is the context on each impression? Right. What does the meaning of that? Is that still aligning with a brand value? So it's evolved kind of on both sides. Right. For the brand as well as the technology.
- [03:46]
B
Here's a holistic question for you. If you were auditing a Fortune 500 advertiser for their media quality program today, what are the first three things that you would look at?
- [03:57]
C
So I would first start, what does quality mean to them? If they don't know, which is often the case, it's still traditional, you know, pillars of verification, looking at IVT, making sure we're doing, you know, 100% of the data hygiene. Right. Best practice, still looking at kind of the gray area of media. So this is, you know, what I kind of consider supply intelligence or supply quality. Just because it might still be reached by humans doesn't necessarily mean that it is a quality, you know, impression or buy for that brand. Things like MFAs, you know, fall into that. And there is still, you know, brand suitability. You know, does it align with where they want their brand to appear? You know, does that hold true for their values?
- [04:39]
B
Sir, I also want to ask you, how long have you been with protected?
- [04:42]
C
I've been with protective for two years.
- [04:44]
B
Okay, so you've been in the media quality, slash brand safety suitability space for at least two years. I would assume that perhaps some advertisers, they Hit these inflection points where they care a lot about media quality and IVT and what have you, and then maybe they fall into a bit of a valley and then back to a piece peak depending upon what they're getting signaled by their clients. Is that what you've observed on your.
- [05:05]
C
Yeah, definitely. I would say from more of an IVT is kind of that foundational hygiene. So as long as they're hitting their industry thresholds, kind of their rates, that's an always on not really paying attention to that as much unless there's a new emerging, you know, tactic or a threat, you'll see, you know, kind of peaks. When it comes to brand suitability. That's a. And we'll say it brand safety, completely different. That is very seasonal or, you know, aligned with political season, elections, you know, some natural disaster or something like that happened. That's when it, you know, peaks and valleys for sure.
- [05:36]
B
Well, I also need to ask you a couple of questions specifically about protective because we're talking about this evolution and what it is that people can do. And so often it feels like a tired conversation that if you talk to a series of vendors that work in the space that they're offering more of the same and the hope is that people want to address media quality, brand safety, brand suitability in new ways. It doesn't always feel like the measurement options are there. So my question to you is, what is it about existing verification solutions to date that haven't allowed us to address the problems that we were hoping to address?
- [06:11]
C
Yeah, the honest answer is that they've been around for a really long time. The verification model and infrastructures has not changed in a long time. So it's just legacy methodologies, legacy infrastructures. And that goes back to some of my earlier commentary around brand safety moving to brand suitability. It's just binary decisions. They're built off of keywords, exclusions, inclusions. And that's just not how context and natural human language and media context work anymore. So it's definitely evolving in terms of the modernized infrastructure and where the industry needs to go.
- [06:44]
B
What have you observed from customers as far as getting to an aha moment when they start using a solution like
- [06:52]
A
what you offer at part time?
- [06:53]
B
Yeah.
- [06:54]
C
From a marketing perspective, I think, you know, we all marketed ourselves as, you know, reducing media waste, reducing creative waste. Funny enough, the aha moment is for the advertisers when they're saying, oh, I can actually act on this data, I understand what my media waste is, but I'm not actually going back and reclaiming that spend. And so it's that aha moment of like this is actionable data. I can actually be doing, you know, in a kind of an ongoing optimization basis.
- [07:20]
B
Is there either a metric or a signal at Protected that really deserves more attention from media buyers?
- [07:27]
C
Yeah, and I promise you, no pun intended, I'm going to have to say our attention offering, that's something that is very unique to us. And I'm not saying attention, you know, in general is unique to us, but it's unique to us in the fact that we have a really unique integration with the Innovate ad server. Typically, the attention solution providers out in the market are really only looking at media quality because they don't have the technology or the visibility into the creative itself. So ours does still look at media quality. The exposure. How is the placement being presented to the user? That's a big part of it. Attention. But so is the creative. What about the creative is resonating with the audience? You know, is it a hero image, is it a feed carousel, those types of things? What environment is it is what call to action? And we have that visibility down to that creative version level and it can be auto optimized within the ad server, which is very unique. And that is a manual process that usually has to happen for advertisers.
- [08:22]
B
We've spent some time talking about ivt. At least we've mentioned it. And now you've told us a little bit about some of what's going on at Protected. I want to speak to another topic that's been probably the hottest media quality topic of the recent past, which is MFA or Made for Advertising Sites. The industry has spent the last few years thinking about it a lot. Has the conversation truly matured around mfa or are we still treating symptoms instead of the underlying disease?
- [08:48]
C
Yeah, I think we're still treating the symptoms. It's easy to spot, there's lots of different signals, so measuring and detecting an MFA is not hard. We're entering MFA 2.0 with AI slop, so obviously new tactics just like anything. But I don't see the industry necessarily doing anything about it and preventing MFAs from being included right into that media buy. I think we have a lot more work to do there.
- [09:13]
B
Of course you just brought up AI slop, so you got to take a. Just think about that. Because the volume of I think MFA sites, they just have to be on a level 10x100x. I have no like actual quantitative data on how big it is, but I imagine it's just beyond the scope of anything. I can even imagine just knowing where AI slop is showing up intentionally in the advertising space. So where is it happening from bad actors? I can't even really speak on it. So another question I want to ask to that point, how much waste still exists in the average programmatic campaign? Where do you see it hiding the most? And maybe it's related to some of what we just discussed.
- [09:51]
C
Yeah, yeah, definitely. And again, there's lots of categories within, you know, IVT industry standard is still around that 3%. But I will say, and this is something that we're very sensitive to, we're very forensic. So at protected, we started in cyber security. So when we're doing our detection, it's actually about higher. It's. It's between 5 and 7%, which I think is still too high. It's usually hidden. Programmatic is just a vulnerable channel. Right. You're. You're going for reach and, you know, kind of scale there. So it's. It's definitely higher in programmatic than premium and direct buys, but hidden in hops. Right. Everything kind of being diluted down the chain as well as, you know, less quality. Very much what we just talked about with, with MFAs, and less, I'll say, quality content and inventory and publishers.
- [10:41]
B
Do you think that there are any spaces where the average advertiser or media buyer is really underestimating the financial impact of your campaigns? Being exposed to all the things that we're discussing.
- [10:54]
C
Right, yeah, yeah. Several conversations that I have often surprised me. When do, you know, proactive insight sessions. This is your IVT re. It is climbing. You know, it's across this certain inventory. What's surprising is the level of acceptance I think the industry has adopted to. It's just part of, oh, it comes with, you know, the nature of programmatic or that comes with reach. Right. Or, you know, whatever my campaign goals might be. So I think the financial impact is often misunderstood and kind of truly left on the table. Like I said earlier, they're not even going back and kind of holding their partners accountable and reclaiming that spend. There's a lot of waste left on the table.
- [11:33]
B
You just mentioned. Sometimes we just accept that there's sizable issues and it's like, oh, it comes with the territory. You just have to accept that that's a part of what happens. And it makes me want to go back to the question I asked earlier around when does a client have that aha moment? Are you ever having a conversation with an advertiser or a media buyer that Initially says that they're far less concerned. And I've had those over the years where they're like, we're not it for cheap reach and we have these goals and like, don't bother me with all this other stuff. Have you been able to convert or turn some folks around to think about things differently that they previously weren't so concerned about?
- [12:09]
C
Yeah, 100%. It's fascinating. The media buyers out there don't actually know that you can hold the partners accountable. You can put terms in the iOS, you can, you know, some of a lot of the media partners and leading buying platforms have reclaim programs. They don't necessarily do their homework and hold them accountable to go and reclaim those. So it's interesting yet more aha moments where we're saying, hey, you know, you can do this. There's a reason why verification solutions are MRC accredited. Right. This is a single source of truth. We are kind of that verification layer to verify the data and say this is wrong, you know, make it right.
- [12:45]
B
You just mentioned the MRC and that made me want to ask you also about just industry regulation or industry suggestions around regulation because that's really how it shakes out, right? Like at the end of the day, unless things are baked into law and oftentimes we react slowly or we don't react at all. So I'm curious, how do you all advocate for yourselves and the work that you're choosing to do? Do you go and talk to the other industry regulated bodies or what does that look like for you all?
- [13:12]
C
Definitely, I think the best, best practice we have and rule of thumb kind of in our DNA is industry peers, collaborations, open sources, you know, anything where we can learn more, you know, vet our solutions. We're not 100% experts and you know. Right. And we're not doing everything in a vacuum. That is absolutely key. There's also an aspect of not every solution or metric needs to be MRC accredited. The supply, intelligence, supply quality that I talked about, that is going to be detection that is going to evolve very rapidly because of the emerging threats. So to go through an MRC accreditation, it's a year long cycle. So it just doesn't make sense. So it's okay that not everything is also MRC accredited, but there's, you know, special foundations like the Internet Watch foundation, things like that that it's very important to, in our opinion and philosophy to be part of because it's what's doing right for the industry, not just advertisers. It needs to be on the supply side. And in the whole industry.
- [14:10]
B
Yeah. There's really not enough action taking place and things move very slowly. As far as true change coming to light, it's going to be an exercise in patience.
- [14:20]
C
Exactly.
- [14:21]
B
So I also have to ask one or two more questions about AI. I know you briefly mentioned AI slop, but there are things that go beyond the slop and there are a lot of AI generated websites out there. They're not all falling into the category of MFA or fake websites or ad fraud. But going back to the concept of media quality, what does the evaluation from a broader standpoint look like for you all?
- [14:44]
C
Yeah, I'm actually so glad you asked this because I'm very passionate about it. Just because it was generated by AI does not make it bad, does not mean that it's less quality. I know several of us, you know, are using it ourselves. We've already seen it in our detection, like major credible publishers and inventory, you know, is using AI. So it's. It just goes back to that same conversation in your same strategy. Is it quality for the brand or not? It doesn't necessarily matter if it was generated by AI. So I think we need to again, very quickly start shifting the way we're thinking about that. It doesn't necessarily mean that it's bad.
- [15:18]
B
Yeah, I've seen a series of websites out there, perfectly legitimate websites. They clearly are heavily developed. Leveraging AI doesn't mean that they're bad. I think they're distasteful in their appearance, but it doesn't mean that they absolutely don't deserve to be ad supported or that there isn't legitimate traffic going to those spaces. This is just kind of the era that we're in right now where people are offering things that may not be up to the standards that they previously were. There can still be traffic that's going in those.
- [15:46]
C
Yeah. And I think it'll even out.
- [15:48]
B
Do you see AI becoming just as important for identifying bad inventory as it is for creating advertising?
- [15:54]
C
Yeah, absolutely. I mean, it's very key. We started on a modernized infrastructure. And so when I say that, and I talked previously about legacy methodologies, legacy infrastructure, in order to process the amount of data and do it in real time, we are talking about analyzing content, meaning, you know, HTML, CSS selectors, things like that, URLs, content, sentiment, semantics, language, everything. Video files, audio transcripts, like the amount of data. So even using AI and machine learning and everything for that is absolutely necessary in order to keep up with everything in real time, which is fundamentally what's broken Right in the system, we are doing things based off stale historical data. And again, that's just not how, how this works anymore.
- [16:38]
B
As we move towards a future where we hope that more people will be thinking about some of the different topics that we've discussed today, especially as it falls into the brand suitability category, what are you excited about and where do you anticipate that there will be healthy change along the lines of brand suitability?
- [16:54]
C
Yeah, that's what I'm most excited about because I do think the industry, meaning on the buy side, the advertisers are starting to understand that they can have, have control again, they have a voice, you know, their unique brand values matter. It's not, again, whether their solution provider says it's safe or not. I'm seeing an interesting dynamic of brand suitability blending into, you know, contextual targeting and contextual alignment. The two are, you know, kind of blurring together, which is actually, I think, a beautiful thing because it becomes a performance solution, not just a, you know, monitoring and measuring and kind of accountability solution.
- [17:32]
B
I really trust and believe that there is going to be a lot of advertisers out there that are going to gravitate, not just because of all the negative chaos that's out there like AI slop and all these websites that really need to be removed from the inventory that you would otherwise be buying, but also because the solutions that are out there are improving. So are there any final thoughts or words you want to share about what people should be thinking about if they want to explore a meaningful solution of change as they move into the future?
- [18:01]
C
Yeah, absolutely. It's kind of a hard ask, but I think going back to 101 ask how things are working, how is this being done, the industry? Something that's frustrating me because I am so excited about it, is that we're talking about moving to this suitability and brand alignment. But how are we actually doing it? You know, is the technology really there? It's not enough to just say that, oh, yep, we have AI and that'll take care of it. You know, what are the levers and mechanisms and what are the features or capabilities that are available to, you know, an advertiser or media buyer to leverage? We haven't kind of shifted into that tactical gear. I don't mean to put more work on, on the advertisers. Nobody needs more work on their plate, but they're going to have to do a little bit more due diligence.
- [18:43]
B
I think you all have a pretty solid methodology and you also have more levers that can be pulled. So there's something that is, there's something to bite onto for an advertiser who cares to find something new out there. So as I'm learning more about Protected as well and, and we'll just be in touch and see how things continue to resonate.
- [19:01]
C
Yeah, appreciate that. Thank you so much for having me. Again, this has been really great to kind of voice some of these.
- [19:06]
A
A special thanks to Sarah Miskevich for sharing her perspective on how media quality, brand suitability and AI are reshaping the future of verification, and for challenging us to think differently about what quality should really mean in today's advertising ecosystem. If you enjoyed this episode, subscribe wherever you get podcasts and leave us a rating or reveal. It really helps others discover the show. I'm Noor Nasir. Thanks for listening and we'll see you for another episode of Adtech Unfiltered real soon.