← AdExchanger
AdExchanger
The Power Of Creative Data
00:51:56
Creative is a strong performance lever. In fact, according to CreativeX, ad creative is responsible for nearly half of sales lift – which is more than reach, recency and targeting combined. So why doesn’t creative get the credit it deserves? Until...
Loading summary
Transcript75 lines
- [00:00]
Anastasia Lang
Foreign.
- [00:13]
Alison Schiff
Welcome to Ad Exchanger Talks, the podcast devoted to examining the issues and trends in advertising and marketing technology that matter most to you.
- [00:24]
Sarah Sluice
This episode is sponsored by iota, a trusted global provider of audience solutions for digital marketing. Ranked number one for global data privacy by Neutronian in 2023, IOTA empowers brands to enrich insights, enhance personalization, and transform omnichannel marketing. As part of the Dun and Bradstreet family, IOTA offers a full suite of solutions to make growth easy. Learn more@ieta.com that's e y e o t a dot com.
- [01:03]
Alison Schiff
This is Alison Schiff and you're listening to Ad Exchanger Talks. Thanks for doing that. My guest this week is Anastasia Lang, CEO and co founder of Creative X, a creative data platform that uses AI to help achieve creative excellence at scale. We'll talk about what that means exactly, and why Creative Data is a powerful data source for brands, one they should definitely be taking more advantage of. But before we get started, please allow me a quick plug and a request. Save the date for CTV Connect, taking place March 12th and 13th in New York City. And hey, don't just save the date, snag yourself a ticket, because this is a can't miss summit on all the key issues and opportunities in connected tv. You will want to be there. Learn more and register@ctvconnect.com Anastasia, welcome to the podcast.
- [01:55]
Anastasia Lang
Thank you.
- [01:56]
Alison Schiff
All right, what is one thing about you that not a lot of other people already know?
- [02:04]
Anastasia Lang
I always struggle with these because it takes you down sort of the humble bragging territory. So I was trying to think of something that not a lot of people know that's quite random. And the best I could come up with is that Gwyneth Paltrow once complimented me on my shoes. Now. Yet. The thing is, I didn't realize it was Gwyneth Paltrow until she walked away and someone had pointed out that it was, in fact Gwyneth Paltrow who'd complimented me on my shoes. But to make, you know, to make it worse, the shoes were purchased for me by my mother, who has, for the entirety of our relationship, always insisted that she has much better fashion taste than I do. Which is probably true.
- [02:41]
Alison Schiff
Oh, now, now she has proof.
- [02:43]
Anastasia Lang
She has. She's absolute and undeniable proof. No one has ever complimented me on anything I've purchased myself. So my mother definitely wins this one.
- [02:51]
Alison Schiff
Are those shoes still part of the repertoire? Are they in your closet? Are they the Gwyneth Paltrow shoes?
- [02:55]
Anastasia Lang
Sadly, no. But they Were they, were these, these lovely sort of half leather, half canvas boots and I wore them to absolute shreds. And when they were, they were, there was no more, you know, sort of saving them. I looked online to see if they still existed, looking down to sort of the SKU number and I could not find them. So sadly they had to go. But the memory, the memory lives on.
- [03:17]
Alison Schiff
The memory remains. Well, that was a nice stroll down memory lane and I'm going to take us even, well, I don't know, even further back. Probably further back. I do want to talk about Creative X. What it is, what it does. But first a bit about your background and how we got here. So I was looking at your LinkedIn. You started out with a few brief stints doing market research for Prada and MTV. And I saw that in 2005 you spent three months researching and advising on target markets for this new show, this then new show on MTVU which was the college student focused version of MTV called Fresh Produce that aired student produced clips and videos. And I went to YouTube because I didn't know what that was. I hadn't seen it back in the day. And so I watched some of the old clips from Fresh Produce and I fell directly down a rabbit hole of nostalgia for the early 2000s, which is when I was just out of college. So maybe that's why I wasn't exposed to it. But yeah, I can recommend do your research.
- [04:20]
Anastasia Lang
That is impressive. No one has brought that up in about 2 dec. The fact that you dumped this up and went down the rabbit hole says something about your journalistic integrity.
- [04:29]
Alison Schiff
Well, I thank you. But I can recommend that anyone going to YouTube and typing in Fresh Produce and MTVU to the search bar will not be disappointed, especially if they were in their late teens or early twenties in the early 2000s. Wow.
- [04:44]
Anastasia Lang
I think people single handedly sort of skew their data. But I'll be like, hey, we think Fresh Produce is coming back. We're getting all these like old hits on our.
- [04:50]
Alison Schiff
We're going to reboot it.
- [04:51]
Anastasia Lang
Yeah, we got to bring this thing back.
- [04:55]
Alison Schiff
But then you spent some time at the Daily Pennsylvanian, which is UPenn's student newspaper marketing and advertising manager. You're an executive board member there. So some sell side experience still in the college realm, which is interesting.
- [05:08]
Anastasia Lang
Gotta sell those ads. That's. I think everyone needs sales experience. And that is how I paid the bills in college, was selling advertising for the Daily Pennsylvanian.
- [05:17]
Alison Schiff
And then you did some consulting work after that, did some time at Google which makes it sound like prison when I say did some time. Product marketing manager. You were a for new business development in EMEA. And then that brings us to 2012 and Hatch Co, which is the startup that you founded. It's a curated marketplace for selling products from independent designers and professional makers, which is very cool. But what's the origin story there? How did you segue from all of that stuff into entrepreneurship?
- [05:47]
Anastasia Lang
So I'd been at Google for. For about five years, and to be honest, I just felt a little bit bored. And Google was a great company, you know, really clever people, big problems, et cetera. But as the company got bigger and bigger, I felt that I had less autonomy, I was able to take fewer risks. And so it just so happened that one of the projects that Google put me into was about starting an entrepreneurship center in London. I believe it was called Google for Startups at the time. And the building itself was called Google Campus. And a result of that project, I got exposed to lots and lots of founders and entrepreneurs, and I became very good friends with one of them. And I remember this very distinctly. We were out to dinner one night, we just closed a deal between him and Google, and he took me out to celebrate this big deal that he closed with Google. And during the course of this dinner, at one point he sort of sat back and he looked at me and he said, google is brilliant at hiring people like you. And I sort of perked up and thought, ooh, there's a compliment coming my way. And I said, what is? People like me? And he said, google is brilliant at hiring these underconfident overachievers. And I thought, well, I'm not underconfident. I am an overachiever, but I'm certainly not underconfident. And what he said is, he said, you know, look like you are missing that last bit of confidence to think that you can go do this for yourself. And, you know, I know it's perhaps a bit exaggerating to credit my move into the startup world from one comment, but that comment did have a disproportionate impact on the way I thought about myself and what I wanted to do. It was sort of like this bubble had been popped. And once that set in, I started to actively think about leaving Google. I went down all the traditional rabbit holes, you know, interviewing at venture capital firms, other startups, and nothing, nothing quite fit. And then eventually a friend of mine, also a colleague of mine at Google, was starting an E commerce company, and frankly, I just latched on and became his Co founder. And that was sort of my, my ticket out. My ticket out of.
- [07:50]
Alison Schiff
And then you realized that you could do it completely on your own because in 2015 you founded Creative X. And I'm just going to read what it says on the tin and then you can expand on it. So our purpose is to elevate creative expression through data and to build the global standard for creative analytics. So I'm going to use a business cliche, unpack that for me and explain what it means to elevate creative expression through, through data. And also what sparked the idea to create a company like Creative X. Yeah.
- [08:27]
Anastasia Lang
So the, the, you know, the, the thing that we noticed all the way back in 2015 in trying to build a patches e commerce company is that increasingly many of decisions that we made had something to do with the creative. Right. The images and visuals we were putting out there to communicate who we were and encourage people to become our customers. Yet of all the decisions we were making, decisions you made about the creative were completely devoid of data. We made decisions largely based on intuition and gut feel. And so we started thinking about, well, how do we create a degree of structure, a degree of objectivity, almost like a GPS for creativity that can help us better navigate how to make these creative decisions in a more informed way. And that's where Creative X was born. So when we talk about, you know, if you sort of try to strip all the, all the fancy speak out of it, what we're really trying to do using technology is provide a creative decision making infrastructure to help marketers make more informed creative decisions, to scale all of their advertising across all their markets, all their touchpoints, all their brands in a way that is not only aligned to driving better effectiveness but, but also consistent and in line with the brand. They're trying to forge the impression they're trying to leave with their consumers.
- [09:51]
Alison Schiff
So let's go a layer deeper. How does the technology work? What data points are you analyzing? What do you look at to determine whether a piece of creative is going to perform or not? Whatever performance means, yes.
- [10:04]
Anastasia Lang
So here I think it's helpful to look at things as two layers. Right. There's sort of like if you were looking at a tree, there's the stuff you see on the surface and then there are all the roots underneath. So I'll start talking about the roots first and then we'll sort of extract up. So what's under the roots is what we think of as creative data. And creative data is basically structured, objective measurement of all the various creative elements in Your ad. This can be everything from the logo placement, a text overlay, the kinds of people you cast in the ad, the cuts you have within the video itself. And historically, you know, all that creative was treated as an art. It was hard to quantify, it was hard to optimize. And the technology's ability to extract that data starts to change that. Right? So we can start to sort of measure everything that's happening across your content. So that's sort of the underlying layer. It's what's happening, you know, in the background. If you picture an assembly line, right, and there are images and videos coming down the assembly line, we're basically stamping them with lots of those creative data elements so we really understand what is everything that's happening in this piece of content. Where things get really interesting is where you take that data and you start to cluster it into themes that marketers can use to make decisions. That is, to what extent does this piece of creative represent my specific audience? To what extent is this piece of creative branded in line with my brand identity, my brand ethos? To what extent is this piece of creative representative or diverse?
- [11:35]
Sarah Sluice
Right.
- [11:35]
Anastasia Lang
So you can start to take this sort of underlying microdata and combine it into larger concepts. And that's interesting when you look at it in an ad level, but let's be honest, we can all kind of look at an ad ourselves and be like, yeah, that ad is representative, that ad is branded, et cetera. But most marketers today are creating millions of ads. And so the ability to be able to get this degree of vision across your entire creative ecosystem and extract insights and then ultimately make decisions based on that is where the power of the technology really comes into its own.
- [12:07]
Alison Schiff
I mean, that's such a compelling thing for performance marketers to be doing, analyzing that kind of data. And creative is really important for performance. I saw a stat, and I think it was on the Creative X website that Creative is responsible for almost half of sales lift, which is more than reach, more than recency, more than targeting combined. And that's wild because the ad tech industrial complex spends like, so, so so much time focusing on the targeting aspect. So if creative is such a key performance lever, and it's also a way to lower your CPMs, I would assume, because if you're more efficient and you're creating creative that performs better, you would have lower CPMs, why does it get sidelined? Like, what is up with that?
- [12:53]
Anastasia Lang
I think a lot of it are just the impact of the technology having never been advanced enough to really be able to do this until very recently. And so, you know, in the past we used to think media and targeting and all of that were things that should be done entirely through, through people. Right. We, we carried all the plumbing ourselves and as technology got better and better, we started to trust it to give us some signals to help us make those kinds of decisions. I think creative is just that, that last, that last element that where technology sort of fell behind for a long time. But largely thanks to advancements and things like computer vision, we can now see in a way that we couldn't see before. I think some of the greatest industry breakthroughs have come from this, this sort of type of vision. Right. If you think about the microscope or the telescope, all of them have given us access to see something in a way we weren't able to before. And I think that's what computer vision has done for creative. Right. Which in fact will transform advertising. Obviously I'm very biased here, but that is my belief.
- [13:59]
Alison Schiff
As I was asking my question and saying, oh, why is creative sidelined? I was thinking actually creative is adored and the big idea is venerated, but it's really that there's been such a separation between creative and performance people. More so than creative is necessarily sidelined.
- [14:19]
Anastasia Lang
That's right. And I think the unhelpful thing that we do in the industry is we tend to put ourselves in camps. We're either the creative people or we're the data people, but we're not both. And so when we have these conversations, people tend to take one side or the other. Instead of embracing the paradox mindset and thinking about how can I be a data driven creative or how can I be an insights person that really values creativity? So the way that our organization or industry, the advertising industrial complex, as you called it, the way it's set up, actually makes a problem, the problem worse. I think the other, the other big issue here, and you highlight a really important point, right, which is with any technology, a lot of the difficulty is the change management. It's sort of the soft stuff, is the hard stuff, if that makes sense. And so one of the issues, one of the issues that we've seen here is when people hear about data being infused in the creative process, the knee jerk reaction is machines will tell me what to create, machines will tell me if my content is good or not. Machines will take my job. And while there are probably some companies that are trying to work on this, that is certainly not how we approach creative data. We don't believe creative data should replace the Big idea. We don't believe creative data should, should replace the people who are coming up with the big ideas. We see it as an enabler to scaling those big creative ideas in a way that will help you maximize their impact and performance. Which is very different than, I think, the assumption most people have when they start to imagine a world where creativity and data live together.
- [15:50]
Alison Schiff
So when it comes to how to be an insights person that values creativity, I was watching this case study that you guys have up on your site with Bayer. It's a video with Patricia Corsi, the chief marketing and information officer at Bayer Consumer Health. And she asks an interesting question which is similar to how can an insights person value creativity? Which is how can brands measure creative excellence in a way that's not just about optimization and roi, but also about, like, resonance and making creative decisions that resonate? So, yeah, what, what can brands do? What's a good way to put this? I don't know, like, get out of their own way and make measurably better creative without getting, like, lost in the weeds of it all.
- [16:36]
Anastasia Lang
Yes. Well, let's start with, with the big question, which is if, if you have a brand vision or creative vision, to what extent are you confident that that vision is being adhered to in all of the millions of ads that you're executing to get that point across? And if we ask people, you know, how confident are you that your branding is correctly represented, that the key messages you're trying to get at or getting out across multiple agencies, you know, multiple markets, potentially dozens of platforms, dozens of different formats, people intuitively tense up because they know there is erosion of messaging. So before you even talk about optimizing the brand and getting better, let's think about to what extent can we start to understand whether or not what we think we're saying to our consumers is actually being delivered across those, those millions of ads that we're deploying. So one of the very first things we like to do is start to say, well, okay, in this case, Bayer, what does. What are some basic creative standards that you have for the hundreds of brands that are in your portfolio? And we start to codify those things for them. And the first bit of insight we deliver is really around helping them understand to what extent their creative vision or their creative standards are being adhered to. We're not talking about optimization. We're not talking about new creative insights. We're just saying, are you saying what you think you're saying? And how much opportunity is there to help you make sure that you're delivering your message in a way that is more and more aligned to what you initially set out to do.
- [18:07]
Alison Schiff
How often are brands not saying what they think they're saying?
- [18:12]
Anastasia Lang
In our research, about 80% of the time. So. And when I say 80% of the time, I'd be more precise. About 80% of the creatives that we measure, the ads that we measure when we start working with the brand, are not in line with some of the creative standards that the brand themselves have defined. Some of them are really basic things, like they're not putting a logo in their ads. Right. It sounds so simple. But yet again, in this advertising industrial complex on which we're on a treadmill of constant content creation, things slip because the systems we build to control that process, that content ecosystem, that content supply chain, were built before the age of huge media proliferation. They were also built before the age of, you know, Gen AI and Gen AI produced content. So we're not even talking about that yet. But, but the amount of content we've been used to create over the last couple of years, which again is a big step function change to what we did about 10 years ago, is about to get another exponential increase. So, so the, and that's actually the biggest, one of the biggest. Aha. Moments in, in working with, working with our brand partners is when they start working with us. There's a recognition of we have been moving so fast to create brilliant content, and yet in that final execution of that content, some things are slipping through the cracks. Right? We're, we're not putting our branding on it appropriately. We're not creating, putting that great idea in the right container so that it can perform well on the different platforms that it's going on. We're not featuring our product in there. You know, again, the things they're measuring will differ a little bit based on each partner and over time will actually differ quite significantly. But over, over about a million ads that we've analyzed and over a billion dollars of ad spend, we found over 80% of advertising does not adhere with the brand's own creative standards.
- [20:08]
Alison Schiff
And then there are, and I don't know if this is something Creative X would, would help with or it's just one of those things that happens. But do you remember the snafu by Mattel when they accidentally printed the URL for a porn site on packaging for dolls based on the Wicked Movie?
- [20:25]
Anastasia Lang
I do not.
- [20:26]
Alison Schiff
That is, yes, it was supposed to be Wicked movie dot com, but on, on, I believe it was Barbie dolls with characters from the Movie. They had Wicked.com which was a very different website. An adult entertainment company owns the web address and they realized it pretty quickly. You know, after printing all of the boxes, when people started to go to that website, maybe some children, that human error. But it's amazing that that can happen. Although not, not so crazy, right? I mean there are so many people involved in this process of producing and developing and things do fall from the crack. Sometimes it's your brand message, sometimes it, it's the fact that you're putting the URL for a porn site on a, on a box with a doll for a kid desert.
- [21:18]
Anastasia Lang
I guess this is a. I'm very curious now, but I instinct tells me I should not type this URL into my computer. My work. Not now, maybe, maybe once you check out at home. But no, that's exactly right. And again, I think this error is all too common. It is simply a symptom of us moving very fast because that is what the media landscape demands. But the problem here is that this is only going to get worse as AI applications really become mass adopted. So a lot of, you know, a lot of what we think about is how do we enable marketers to get that confidence from powering every creative decision with data, but really also future proofing the content they're creating in an AI first world. Right. Because increasingly I think we're going to be moving to the stage of what I think of as self driving advertising, where machines are going to be talking to machines not only for the purpose of content creation, content optimization, but also media decisioning. And in the same way that humans used to control the car, but now self driving cars are on the road on their own. It's completely black box. But even in that world they're on the road where there are regulations and standards and systems and advertising needs the same thing. If AI is in the driver's seat. Brands need systems of control and that layer of measurement and control is what we aim to provide.
- [22:42]
Alison Schiff
So we're going to take a quick break in a few minutes to hear from our sponsor, but I want to broach an important topic and change gears a little bit before we do that. And I know it's something that you're passionate about, which is dei diversity, equity and inclusion, including accessibility and representation. So Creative X's technology has features that help brands integrate representation into their creative production to like really understand their target audience. Create advertising that resonates because it represents, which sounds almost like very buzzwordy, but there's real, there's real, realness there. And you've also collaborated with the Gina Davis Institute on Gen Media to build a product that gives brands like concrete data they can use to act on their DEI commitments. So tell me a little bit about like how and why representation is good for roi because there's a business case for it. And you know how you help support that with technology?
- [23:43]
Anastasia Lang
Oh, this is, this is a topic that I, I, I both love and get very frustrated talking about because you know, to your point about being good for roi, I have yet to come across a single study that demonstrates that more representative advertising isn't good for business. We've seen that done demonstrated at the shareholder level and share price level, at the brand level, at the sales impact level, at digital cut through and engagement levels. More and more studies continue to come out that show you that representative advertising is good for business. Yet action doesn't really seem to change. And that I think is the sort of, the, the frustrating paradox in it is I don't think we need more research to tell us that this is important, but yet somehow action on this is being, is being stymied. And so what we set out to do and our thesis here was, well, maybe the reason why action isn't being taken is because of a lack of measurement. So we partnered with the Geena Davis Institute a couple years ago because they're again, they're the experts on diverse and inclusive advertising. They're the experts on representation. They've been doing this for years. Right? So we don't pretend we are, that we are the technology people. So we tried to work with them to think about, well, how can we start to understand to what extent your ads are representative? What if we could measure things like, you know, sex and skin tone and age range as well as situational settings in ads to help you understand how are you casting different people and to what extent are you featuring them in certain situations? For example, how often do you feature women in domestic settings versus professional settings when you have people of color in your ad, how often do you show them as in leadership positions or as professionals versus emphasizing their physical strengths and characteristics? And you know, the data says pretty much what I think you expected to say, which is we tend to be biased in the portrayals that we show out there. I think what we realized through this very lengthy experience is that measurement is part of the solution, but it's not the solution. And I think part of the problem remains to be that we as a society, especially behind sort of closed doors, are very uncomfortable having some of these conversations in a way that actually elicits change. Despite the fact that I think the intent is really good, people genuinely, deeply believe this and want to do better. But there is so much fear of making a mistake which actually paralyzes people from action. Which is why some of these numbers about representative advertising, despite all the positive performance association, just haven't moved very much.
- [26:36]
Alison Schiff
All right, well, there's a lot more to say on the topic of DEI and media advertising creative production. We're going to take our quick break, but stick with us and you'll hear about it. Foreign.
- [27:01]
Sarah Sluice
Hello, I'm Sarah Sluice, Executive Editor of Ad Exchanger, and I'm here today with Nate Carter, Vice President of Global Sales for iota. Welcome, Nate.
- [27:11]
Nate Carter
Thank you, Sarah.
- [27:12]
Sarah Sluice
So everyone is talking about AI. How do you see AI transforming the intersection of data privacy and personalization?
- [27:23]
Nate Carter
Yeah, it's helping in several ways. We see AI tools helping our team be more effective and more precise in what we're able to offer up to our clients in terms of audience recommendations and insights. At the same time, we're seeing our clients using AI in order to be much more selective in terms of how they're leveraging audience activation across media channels. They're looking at broader use cases of data and getting to more precise results with their campaigns.
- [27:58]
Sarah Sluice
There's also a continued shift towards a more and more cookieless ecosystem. What approaches is IOTA implementing to ensure marketers can maintain effective targeting through this transition?
- [28:13]
Nate Carter
We're working really closely with our partners in order to ensure that our data is transformed into any and every identifier and placed in every ecosystem that our brands need it to be in. This can look at transforming our data against things like ID 5 unified ID. We also have partnerships where we can take offline data sets and use them to map into contextual signal from a customer standpoint. We're seeing more and more where our data which resides offline being transformed and moved into environments like Google Cloud Platform, Snowflake Databricks and others where they can start to do offline work before moving it into the digital universe.
- [29:00]
Sarah Sluice
So you brought up your customers, so let's dig a bit deeper into that. How are your clients thinking about delivering meaningful personalization without compromising compliance?
- [29:11]
Nate Carter
They need to work with partners who they know are taking compliance as seriously as they are. And that's our job at Dun and Bradstreet and iota. And so we do that. We work in Europe, we know how to be GDPR compliant. We work state by state in the US at the same time we know that these law changes are going to diminish some of the depth of personal signals that you're going to get on a consumer in any market, and we think that that's a good thing. As we move towards a privacy first market, our customers still need to drive the same results. And so we're seeing them leverage a greater breadth of data, as we discussed, with AI being an enabler of that, while removing anything that could be sensitive both now and in the future. As we move towards a future that is increasingly privacy sensitive, it's important for a healthy ecosystem that we lead the way and we're doing that with our customers in being privacy first.
- [30:08]
Sarah Sluice
Thank you, Nate. Checking those boxes, but also thinking strategically about how to still maintain performance. We appreciate you becoming on the pod and thank you for supporting the podcast.
- [30:19]
Nate Carter
Thank you so much, Sarah.
- [30:38]
Alison Schiff
All right, we're back. And I think I'd be remiss if I didn't bring up the current political and cultural environment. So how. How should brands think about their DEI efforts in light of the world we live in? And have you noticed brands pulling back or getting touchier about dei like in the lead up to the election, post inauguration? Because, yeah, there's a fear of messing up, but there are other fears too.
- [31:05]
Anastasia Lang
Yes. So the short answer is absolutely. I think we've seen a lot of initiatives around diversity and inclusion and representative advertising deprioritized as initiatives. I think there's still some talk of it, but you can tell it's fallen out of a lot of. A lot of roadmaps and plans. You know, my personal view is that we have somehow politicized diversity, inclusion in advertising versus whereas I do not think this is or should be a political issue. If we take a step back and think about what brands are trying to do, or at least should be trying to do, is authentically representing what their consumers look like. And that is not a political statement or should be a political controversial point. It's simply about, you know, at the heart of marketing, I believe, is about really understanding who your audience is and what do they care about and can you speak to them in a way that will cut through? So if you take all that political stuff out of it, I think the questions a lot of brands should be asking, maybe we need to put a different label on it, is am I authentically representing my consumer in a way that cuts through? And if I don't know the answer to that, well, I should probably start with some measurement and then based on what that measurement shows, I should think about what that says about the potential of my brand and our growth and how consumers are going to react to this message if they themselves don't see, if they don't see themselves represented in the content that we're creating.
- [32:39]
Alison Schiff
Where do ads typically miss the mark from a DEI standpoint and maybe like an accessibility perspective and how much of that is a technology challenge and how much of it is like a human challenge? Like just maybe the solution is getting the right people in the room from the beginning.
- [32:58]
Anastasia Lang
So where things miss the mark, I think this is where it's all your usual suspects, right? I think advertising, to your point, is just a number of people sitting in room making decisions. And good people have biases, right? All the time. I have them, you have them, they're there. All we can do, sort of be aware that we have them and do our best to, to build systems to, to recognize when we're, when we're engaging in them. And so the biases we see in society, I. E. You know, women are, are more skewed towards playing domestic roles. When men play domestic roles, they are incompetent at it. Things like that, those things end up playing a role in advertising. Right? Advertising, sort of just a reflection, a mirror of reflection to the way we behave. I think the accessibility conversation is a little bit more nuanced in that here. I think because there is a gap between the majority of the folks who are making the ads and especially some of our, some of the older consumer segments. There's a lack of appreciation and empathy for how they will engage with the ad. So for example, one of the things that we found is that, you know, people above 60 tend to have, you know, more pronounced vision problems, right. Their eyesight, gates worse, etc. Yet most advertising has a font that's a size 12 or 14 and when you think about that on a small screen, it's basically unreadable. So we're putting out all these ads that, that people simply can't read the message on because it's not set up for an older audience. So there are things like that which, you know, can technology help? Absolutely. But I do think part of a bigger part of the problem is there has to be some concerted central effort to want to make this change. I think again, you have a lot of very well intentioned people who want to make great content, but they tend to make content in silos and everyone's sort of doing their own thing. And together, when you aggregate this all up is when you have a problem where no one's systematically thinking about how do I make content that is representative of my audience and inclusive and accessible to everyone. One counterexample I have to that though is Halion, which is, which is pioneering this concept of inclusive advertising using our technology. So what they did is they used our data, they analyzed hundreds of thousands of ads for some of these elements, right, for things like readability, inclusivity, et cetera. And based on what they found, they said, okay, we now have a system by which we can give every market and every brand an understanding of to what extent they are representing their consumer, who's going to be a slightly older consumer in a way that is inclusive, accessible, etc, and have now rolled that out as a system essentially as an infrastructure for making sure that going forward the concept of inclusivity and accessibility is standardized in the way they think about creative decision making.
- [36:05]
Alison Schiff
And there's also just unconscious bias. I was interviewing a few years ago now, the CMO of Bob Evans Farms, a woman named Time Hill, and she was meeting with an agency that was representing the talent selection process for this like light humor TV spot that was going to center around a black woman security guard character. And, and each of the three actresses were being directed to like scowl and look annoyed. And after the presentation, the cmo, Time Hill, she spoke up and she said something like, is there anything we can do to just like not fall in this trap of the angry black woman Persona, like when we select our character for this or our actress for this spot? And it wasn't deliberate, but you know, she had said to me, like, I do wonder what would have happened if I, if I wasn't in the room. And there's this of course, basic truth that AI, AI powered systems, they have a tendency to take human biases and amplify them. So it's a human thing and it's a technology thing. But yeah, what, what role does AI play in analyzing content for or bias. But then also the counterpoint, like, how do we address bias in AI, which is often a reflection of ourselves.
- [37:32]
Anastasia Lang
Yeah, yeah, I think that, you know, to your exact point, most AI systems are black box systems that are trained on the data we give them. And as the saying goes, you know, you put crap in, you get crap out. Or perhaps more constructively, AI is again going to be a reflection of all the things they learned from the way we've been doing things to date. And if we accept the fact that these are black box systems, again, sort of the self driving cars on the road, then the question is, how do we continue to have systems of control so that the brand can make sure that whatever the AI is creating is actually in line with the things that they care about. So this is where you can start to do things like say, okay, well, if I'm, if I'm brand X and I don't want to have tropes of angry black women in my advertising, this is where we can come in and analyze every ad that Jenny I is creating to flag those things up front and say, hey, actually this is against this brand's standards. It will differ per brand. Right. Every brand will have its own things that they really care about demonstrating or avoiding. But what we're trying to do is build this impartial layer that will allow them to create a system of control over any bit of content that's out there. My sense is AI systems itself will get better at this, but they will always be reactive to what we're finding. So it will take AI systems, they'll create some content that folks find offensive, and then someone will go in the back end and engineer the fact that that content should no longer be shown. So we will always be reacting. If we want to be proactive, we have to really build infrastructure that gives us that control over the AI systems.
- [39:21]
Alison Schiff
And how real are some of those things that you were talking about? Separate what's legit from what's kind of BS when it comes to advertising and marketing use cases for AI? The real practical application stuff you can do today, right now, and what's just like smoke and mirrors and marketing fluff?
- [39:39]
Anastasia Lang
I think it's, you know, I think a lot of it is. I think a lot of it is, is stuff that has the potential to be real. I think what is exaggerated is the extent to which people are using it and how much human input is required to use it. Well, right. So when someone says, oh, this ad is AI generated, fantastic. What they don't show you is the, you know, 45 minutes you had to spend prompting the system to get it to look exactly like what you wanted. And you might have then even downloaded it and then sort of fixed it yourself. So I, I think the exaggeration is less than the potential of what's possible. But the fact that these systems still require a tremendous amount of hand holding and, and maneuvering to get them to do what they. What we want them to do.
- [40:31]
Alison Schiff
What AI tools do you use in your daily life? I mean, putting aside the tools developed by the company that you founded?
- [40:39]
Anastasia Lang
Well, you know, I'll tell you, I'm. I'm a Huge. This is, this is not going to rock anyone's world, but I'm like a multi day user of ChatGPT and I have found myself using it in ways that have, I don't want to say they've transformed my life, but they've given me a degree of clarity and feedback that I think would have been much harder to get. So I'll give you some examples. We have an executive meeting once a week where I meet with our senior team. And during that meeting our job is to make decisions and my job as the CEO is to moderate those conversations. And so one of the things I started doing a couple months ago is I've started feeding the transcript of those meetings to ChatGPT and asking it to rate us as a team and our decision making ability and rate me as a moderator of, of our, you know, senior executives. And it's been really helpful. You know, it's, it's pointed out when, you know, I can do a better job in terms of pushing for clarity on a decision, it's pointed out like, hey, it's three weeks now and you're still kind of going in circles around this thing. So it's been really good for some of these applications that previously, you know, I kind of would walk out of the meeting and feel like, oh, I think I could have done a better job here, but didn't really know exactly why. And now I have a system that essentially gives me feedback. And look, not all feedback is valid, right? But it's, it's pretty good. I can't sit down with everyone on the executive team after every meeting, be like, hey, how do you think that went the other place where I do a lot of writing, I'm a big writer. I will never let Chad GPT write for me. But the hack that I've been using is I've been feeding it things that I'm writing and asking it what, what is the one main point you take away from this? To see how much that squares with the intent that I had. I also ask it to rate the things that I write for logical clarity. So, you know, again, it's there sort of an editor, right? I've always had editors throughout the company, but that took more time. So it significantly condensed my editing process, which I found tremendously valid.
- [42:45]
Alison Schiff
Valuable. You know, I've started using Perplexity for research and it, it is helpful. It's. I find it most helpful for stuff I already know. I just want it presented to me in nugget form quickly as opposed to having to trawl through a bunch of blog posts or do, you know, a whole research project that involves clicking on a million blue links. And I was definitely resistant. And it was very quickly that I saw the value there. Kind of to my chagrin because, I don't know, I'm a little bit of a take me to a cabin with no wi fi and a stack of books and I'll be fine kind of a person. But I found it really helpful.
- [43:27]
Anastasia Lang
Yes. And, you know, I think when, when ChatGPT first came out, you know, it felt like such a revolution for many people. I remember going down this rabbit hole and looking up what were the predictions that people made about email when it first came out, you know, in the 90s. And a lot of them are very similar to what we see now about AI. Right. It will render a lot of the workday obsolete. It will make a lot of people out of a job. You know, we'll. We'll move to sort of a universal, universal basic income because no one will need to work as much. And of course, the very opposite is true. And I've noticed those same things for myself where, you know, as I've gotten more efficient on certain tasks where it's cut the time down, it hasn't. Doesn't mean I get to go home earlier. I just kind of get to do more stuff. So, you know, I do think some of those predictions about how we'll change, what we're able to do are not as have been, have been sort of greatly exaggerated.
- [44:25]
Alison Schiff
I did have, and I've talked about it on this podcast before, a strange experience where I asked Perplexity to write an article in the, the style and the tone of Alison Schiff from Ad Exchanger. And I pride myself on being amusing. It's like my personal brand. I like to elicit chuckles and be a little jokey. And the resulting article was just like, dad jokes the whole way down, like, oh no, is this what I write? Like, this is.
- [44:55]
Anastasia Lang
I've had a similar experience because, you know, initially I thought, oh great, maybe I could get it, see how it writes for me and get it right for me. And I fed it a bunch of the articles that I'd written and I asked it. I had another thing I was writing and I said, okay, based on this, based on the stuff that I've written, please create an article. And, you know, it was fast, but it wasn't good at all, and it didn't sound like me. And again, this is where I am a purist that, you know, the team the team at Creative X will tell you how annoying it is to work with me, because they would. Can we just write for you? And I'm like, no. Like, my voice is all I have that is mine. I will never let anyone write for you. I don't care if it's a quote. I don't care if it's Presley's. I don't care if it's a social post. Like, if it. If it's got my name on it, I want to be the one to say it. And again, you know, when I. When I read through the stuff that ChatGPT had written for me, I was like, oh, this is. I would absolutely never give it the pen. But as an editor, it's a really good editor. It's an unforgiving editor.
- [45:50]
Alison Schiff
Well, I'm a very forgiving editor, I think, actually. But you'd have to ask the people that I edit.
- [45:58]
Anastasia Lang
Yeah. I wonder if their perception would be the same.
- [46:01]
Alison Schiff
Well, we're nearly out of time, and I wanted to ask you a question that has to do with some research, some other Creative X research that I came across that you released in Cannes. And one of the findings kind of blew my mind, which is that brands don't use nearly half of the creative assets that they produce, which is nuts. I mean, like, what a waste. It's like 45% of global creative assets that are produced are never activated in any capacity. So basically, they're just lost in the handoff between product and media teams or between global and, like, local teams. So, I mean, my question is just what's up with that? And, like, what can be done to fix it? Because it's just so wasteful.
- [46:45]
Anastasia Lang
Yeah, I think this is part of the. Part of this. One of the side effects of this content production treadmill we find ourselves on for years. Brands are saying, I need more content, I need more content. And they really started to ramp up the production, the production of new ads, and then those ads would go sit into a digital asset manager or maybe someone's drive or someone's email. And when we looked, and because we have access to both sides of it, we see all the content that your creative partners are producing, but we also see all the stuff that's running live that you're putting media against. So this gave us an interesting idea of like, hey, let's combine those two data sets and let's see, of the stuff that's been produced, how much of it has actually ended up running, you know, at some point later in time. And what we find is exactly what you said. About 45% of it never goes live. And some of the reason why it doesn't, some of the reasons why it doesn't go live, you know, they stem from, again, both technical issues. That is, people can't find the ads. They don't know those ads were made. There's a disorganized, decentralized sort of content storage system, et cetera. I think there are great, some great digital asset managers that are trying to solve that problem. But the other part of the issue is, as always, you have producers of content, right? And then you have the folks who will actually take that content, use it for their brand. And sometimes you have systems where centralized folks produce content, expecting local teams to then use it. Local teams don't always want to use the stuff that central produced. There are also, as always, some people, elements embedded in this. But I think part of the value of having this kind of data is first of all, you can start to isolate. You can say, well, look, we created these assets for this big global campaign for a brand that's live in, you know, 20 markets. And they can specifically say, well, here, here are markets that never activated it, right? Here are markets that never used any of the, of the campaigns that, that we created for them. The other thing that becomes kind of interesting is when you then compare those globally produced assets to the assets the market made themselves, they're actually very similar, right? So if you take, you know, if you take a, a beverage company, they'll put the bottle of the beverage on a beach, right? And it's. The local team might have produced it, or the global team might have produced it, but the ads fundamentally are not too dissimilar. And yet we have not only spent money in not utilizing the ads that were created, but we've then doubled down and created more ads that look similar without, which just compounds the whole problem. So as always, technology and people are both, both have a role to play in fixing this problem. But as marketers are continuously coming under scrutiny, I think this problem is getting a lot more attention because this content landfill of unused ads is just growing and growing.
- [49:40]
Alison Schiff
Oh my God, what an image. A content landfill of ads.
- [49:45]
Anastasia Lang
It's big. It's big. But you know, the thing that I was most surprised by is I remember we were working with the CMO of a global Fortune 500 organization. They were the ones who had initially put us onto this problem and said, look, you have access to both the stuff before it goes live and after it goes live, can you put this data together for us. And we went back and we did the analysis and the team came back to me and they said, hey, our research is showing that 80% of their content isn't being used. And I said, no, that's wrong. A very useful knee jerk co reaction. I said, that's wrong. Run it again. And they ran it again and they ran it a third time and they were like, look like these are the results we keep getting. We've tried everything. We don't know what the problem is. Can we go back to the team and see where we may be making a mistake in doing this analysis? So we went back to the CMO of this organization. We said, look, we think this data is wrong, but right now we're showing that 80% of content for this global brand is never being used, never being activated, never being shown to consumers. And this person said to us, that's exactly what we were expecting. So, you know, I think all CMOs understand that there's probably a ton of wastage in the system. They just haven't had the tools to really isolate it and do something about it.
- [51:00]
Alison Schiff
All right, well, I guess the tools exist now. So stop being so wasteful. It's time to recycle. Our planet needs you.
- [51:11]
Anastasia Lang
Foreign.
- [51:17]
Sarah Sluice
Today's episode was brought to you by iota, a Dun and Bradstreet company and trusted global provider of audience solutions for digital marketing. IOTA is your partner for quality certified audience data and transformative digital marketing solutions. With comprehensive consumer data capabilities, IOTA delivers the confidence and tools market marketers need to Thrive. Discover how IOTA can make growth easy at iota.com, that's u I e o t a dot com.