let's get started. Well, hello, everyone. Welcome. Thank you for joining us at this webinar that is titled the cookie conundrum. What happens to advertising in a world without third party cookies? What now? My name is trittico Parker. I am the head of Product Marketing at Quantcast and I will be one of your presenters today. The presentation will be roughly about 30 minutes long, we will leave ample time at the end for q&a. So please register your questions as the presentation progresses, we are looking forward to a rich discussion at the end. For those of you who might be new to Quantcast or who haven't had a chance to connect recently with our sales team. Quantcast is a global advertising technology company. We are in over 20 offices across 10 countries. We build products and services for brands, agencies and publishers to know and grow their audiences. Our flagship product is the Quantcast platform. for publishers. It provides audience and content insights as well as monetization capabilities and for marketers at brands and agencies. It provides end to end campaign planning, activation and measurement capabilities. I would also like to welcome Elena Morris, who is a lead Product Marketing Manager here at konkurs. She is going to be one of my co presenters today. Elena welcome and why don't you introduce yourself?
Thanks, Judy. Hey everyone, I'm Alina Morris. As Trudy mentioned, I'm on the product marketing team here at Comcast. I oversee Product Marketing across the global publisher business. So including the Comcast platform for publishers, and Quantcast, measure our free audience analytics solution and Quantcast choice, our consent management platform. I actually come from a publisher background prior to Comcast, I spent several years in the publisher world, first at Yelp, and then more recently at CBS interactive. And I'm going to apologize ahead of time for the lovely sounds of the city that I have in New York with me today. I'm sure you guys already heard some of the sirens behind me. So hopefully Fingers crossed on the streets notice stay quiet today. Okay, cool. So I'm going to get us started. Um, so what are we going to talk about today? Well, first, I want to explore some of the major trends in the industry right now, to understand how they're going to impact advertising. Then we'll look at emerging industry proposed solutions and how or where they might help our industry move forward. And finally, I'm going to hand it over to shirting. She's gonna take us through Quantcast unique approach to the new world and the role that AI and machine learning has in future proofing and your business, our business and the industry. So let's first touch on some major trends that are really impacting shaping and transforming the digital space today. There's a lot of different factors at play right now I'm all of them impacting our ecosystem. And it's really hard to talk about one of these without touching on the others. And so let me look at each of these a little bit closer. Of course, the reason we're all here, third party cookies are going away. Third Party cookies were introduced nearly 20 years ago across all browsers, and were and are actually still used in some browsers to associate a person from site to site. From a marketer and publisher perspective, third party cookies have helped us understand consumers behaviors and habits and helped us infer interests outside of the purely contextual content that those users are reading across the ad tech and Mark tech space. And they've helped us serve ads and effectively measure the impact of those ads with metrics like ROI, CPC, and many others. Now, we're just a few short months away from third party cookies being eliminated completely. And when I say completely, I mean eliminated from the last browser that still has them, which is Chrome is important for us, though to all recognize that we already live in a world without third party cookies were part of the internet, Safari and Firefox deprecated third party cookies more than a year ago, for Comcast and for our customers. And for other emerging solutions out there, it actually means that we just already have a viable place to test approaches without third party cookies, which is great. So let's also now look at the evolving consumer privacy landscape, which in many ways, is what kind of forced third party cookies to start being deprecated. So in response to mistrust of data and data usage, largely shaped by third party cookie use, and some data leakage scandals. In the last few years, we have seen broadened sort of corrections from a regulatory perspective. There's been a flurry of regional privacy laws being passed, most notably and really kicking off this major regulatory trend, which was GDPR and the EMEA market. And then here in the US, we've seen ccpa, now cpra, and we're seeing a lot more movement in this space in the US now, Virginia just passed a law in New York and Washington are passing laws and many other states around the country are looking at this too. And then globally, there's a lot of regions that are passing their own regulations as well. Um, notably, we see Canada evolving their current privacy law, which some may say we may be as strict as GDPR. And Brazil, Singapore, and several other major regions around the world are passing more regulations, that consumer privacy is not just about regulation that's being passed from the government perspective. There's also consumer privacy standards being created for our industry from within our industry that will guide emerging identity work. And we'll look at that in a moment. Overall, one of the things that we believe at Comcast is that identity and privacy are really two sides of the same coin. So any path forward in the future has to really include both. The next factor that can potentially impact the future is the fact that walled gardens are under a lot of scrutiny. Right now, both Google and Apple are really being looked at closely for anti competitive antitrust reasons related to what they're doing to the advertising industry for both digital but also in app with idfa. And for search as well. So while we don't know exactly how this is going to play out, what we do know is that it's creating even more noise and confusion, and certainly more uncertainty in an already hot topic area.
And finally, there's going to be a lot of movement and shake up in the ad tech loom escape, the third party cookies and the associated mash tables that map one cookie space to another, have enabled DSPs, dmps ssps, all the acronyms so on and so forth, to really interoperate. And all of that will be falling apart with the departure of the third party cookie. So we do anticipate that this will change and consolidate the tech that exists today. Of course, in response to all of these changes, and in an effort to help mitigate some of this risk, and some of the uncertainty of the future, there's a lot of emerging solutions that are coming out. But we can't really talk about the solutions until we first talked about the standards that are really being created to help guide these solutions forward. So the first group that I want to talk about and want to touch on is Ivy Tech labs project three arc. So much like the work that Diaby Europe did when the with the transparency and consent framework, which guides compliance for GDPR. I B's tech lab is creating standards of excellence for independent ID solutions. Many companies across the industry are involved in this group, including marketers, publishers, ad tech companies, including podcasts and others, we're still waiting to see exactly what those standards will include and how they'll be formalized and finalized. But I do think that we can expect to see consumer and consumer consent really at the heart of those standards. And so as emerging IDs come forward and emerging solutions come forward. And we do believe that consumer consents really going to be at that at the core of those Nexus prebid prebid helps guide the creation on the ID pipes to help connect the industry. So when we talk about the loom escape, and the connection between all of those companies, changing prevailed will effectively help uphold some of those connections in a new way. Comcast is actively of all involved in the conversations here, especially around the emerging proven modules. And then finally, we have w three C. This is the browser Consortium. And they're really leading the conversations around development and testing of the foc solutions. Again, Quantcast is actively participating and involved here from an engineering perspective with w three C to help define this and test out this cohort solution for the industry. Okay, great. So we know that standards are being created. But what are those standards really going to guide? Well, the answer is, is that there are a ton of independent identifiers and solutions emerging, actually about 80. According to this recent adexchanger article, we know that this can be really confusing. We hear it every day from our from our marketer and our publisher customers. And even internally, it's very challenging, as I'm sure people in the audience can relate to, it's challenging to navigate all of the new solutions that come out even on a day to day basis. So what we try to do is we really try to think about this in more of like a simplified fashion. And the way that we've we've thought about it is looking at these on like a spectrum of solutions. On the far left, we have authenticated solutions. These are based on people's true logged in deterministic data, like an email or phone number. This is something that like, if you have an email, that's Elena Morris at Comcast, you know, for sure, okay, this is only to Morris. On the far right, we have cohort based targeting. This is group based targeting of people who have the same inferred interests based on their online actions. And in the middle of something maybe in between these two, perhaps with the accuracy of authenticated but maybe with a little bit more scale that we anticipate seeing from karbach solutions. So let's look at these a little bit more closely.
I'm going to start with authenticated identifiers and unified ideas. A really great example of an Authenticated Identifier. These IDs are extremely accurate because as I mentioned, they're based on someone's true Id like an email or a phone number. And these identifiers are really going to be an important piece of the puzzle moving forward, because of that accuracy. And because of their, because of the fact that they're based on that deterministic data set. But I do want to touch on something that I find on personally very interesting, which is the scalability piece of this solution. And we've heard in market around authenticated solutions, and that perhaps as much as 80% of the internet's going to be logged in in the future. But I think we as an industry really needs to be honest with ourselves around this number. I come from the publisher world, I spent several years at CBS interactive. And today at podcasts, I work with some of the biggest publishers around the globe. And I can tell you that getting users to log in to something, even publishers, sites that they go to every day is really challenging. The biggest publishers in the world see five to maybe 15% of users logged in at best. And you do have some outliers that have maybe strong subscription models, or they have very high user trust, and and maybe they see up to 40, or 50%, or even maybe more of their users logged in. But then asking those publishers that share those emails is going to be a real challenge, just to ask them to share that valuable data they've worked so hard to get. So while we definitely believe that this is going to be an important part of the puzzle moving forward, I do think personally that we need to be honest about the scale and think about pairing this with something that maybe has a bit more reach as part of a holistic plan for the future. So then, let's look at the other side of the spectrum, more of the group or the cohort based solutions. In this solution. A browser is going to be collecting information about browsing history and habits, inferring users behavioral interest and using that information to assign a user to a cohort or group of no less than 1000 people. This group will also include other consumers with similar interests. the browser's will then share that cohort ID I'm indicating like what group that consumer belongs to with both websites and advertisers. And the concept here is that because it's not an individual identifier, it will respect consumer privacy. And if we think back to what we were discussing around industry standards, consumer consent is a real critical piece here. I think it's interesting to know, I'm as it relates to the fox solutions, in particular, that these are not currently being tested in the EMEA market due to GDPR standards. And from my perspective, it just goes to show that any emerging solution, even solutions coming from from Google really do need to have the consumer and consumer consent at the center of what they're doing. Either way, we do fully believe that flock will make the necessary changes to operate in the EMEA market and to respect consumer privacy at a global level. And we do think that foc will be an important piece of the puzzle as we move forward. So many are asking in these scenarios, who's going to win, there's so many identifiers out there, there's so many solutions. Some of these, we didn't even touch on like contextual from a publisher perspective, many publishers are lean a lot more heavily into their contextual strategy, and building up their first party datasets. So there's just a ton of options out there right now. And you know, one of the conversations that we have internally, and one of the conversations that we have with our customers is where should we put our chips? Are we placing our bandwidth UID? Are we only testing flok? Are we is contextual going to be the new gold standard. And we've been having this discussion for a long time. But actually in this webinar, we'd love to hear from you. We'd love to get the perspective of our audience of where the people attending today think that I'm identity is going to fall. So I think there'll be should have a poll be popping up.
Awesome.
So I'll give everyone just a few seconds, I would love to hear perspectives. And we have a wide array of people in the audience today, ranging from agencies to client direct to publishers to other ad tech companies. So it'll be interesting to see kind of the different answers that come out. Just give everyone maybe five or six more seconds.
Cool.
Let's go ahead and close the poll and take a look at what those results show us. Oh, interesting. Okay, so we're seeing actually the highest response for first party audience segments, which we didn't actually cover too much in this webinar. So that's really interesting. I'm curious to know, hopefully, we can take a look a little bit closer at I'm maybe some of the folks that responded, we can get some feedback on that. That's awesome. And well, either way, on our podcast, we don't really believe that one size is going to fit all. In fact, we think that all of these solutions will play an important role in the future. And so you can see we've put our chip kind of right in the middle of the board. we're placing our bet on on all of them, so to speak. So I'm going to hand it off to Sheree. Now to talk about this in a bit more detail. sruti over to you.
Yep. Thanks a lot, Elena. So you know, as Elena mentioned, the Quantcast approach. And technology is really grounded in this philosophy, that there is no silver bullet, there is no one size fits all. And we need to be prepared for a future where all these solutions will likely play a role. With that, you know, we like to say that our approach is, can be described with three eyes. The first eye is industry standards. As Elena mentioned, we've been participating with IB tech labs project we are with prebid, as well as w three C, to define the industry standards that will guide the future of advertising. The second eye is interoperability. So with interoperability, we have built a platform that can take in multiple different external ID signals, be it from UID, or live RAM, or Google sock, or something else entirely. The reason we believe interoperability is important is the same reason why as an example, we support different dmps. Today, it's because our customers use multiple different dmps. And we like to meet our customers where they are at. We do believe that in the future, there's a good chance that our customers will use multiple of these solutions. And hence, we are building and architecting the platform to make sure that we support that. Finally, the third eye is innovation. This is where all of these incoming ID signals will be combined with multiple other signals using AI and machine learning to statistically determine who the people are, who the audiences what their behaviors are intent and interest. So let's take a little bit of a deeper look at what these multiple signals are, and how we and machine learning can help. So let's just start on the on the left like first signal, I'd like to talk about this contextual. Now, of course, contextual, can be used for contextual targeting. But that's not what we mean here, we mean contextual as an important signal to be able to understand user behavior and intent. The second signal is first party, you will all have heard that, you know, and the poll showed that a lot of us think that first party curated segments, first party data is going to be more and more important. So this would be first party signals not just from our publisher partners, but also first party data that our marketer or partners might have. The third signal is consent. As Alina mentioned, consent is really center stage, we cannot use any of the information in any of these signals without consent. So we have to make it a part of our approach. Then, of course, we talked about identifiers and cohorts, Alena talked about those, these are new emerging solutions that are starting to become available, and then that will continue to evolve. And then finally, there's all bunch of other signals that are available like time language geolocation device. So there's, there's a multitude of signals that can be looked at to help us
provide addressable advertising. However, there are some notable things about all of these. One is that these signals are really complex. They have raw data that needs to be processed differently. So it's quite messy. The second thing is they are real time. So they are constantly changing from every minute to every second, they are constantly changing constantly in flux. The third thing is there's a constant rate of change. What we mean by that is, you know, we have UID 2.0 today, but it used to be UID 1.0, maybe there'll be a UID 3.0. And the same thing goes for other solutions like flock, it will probably go undergo multiple different evolutions. And then finally, the fourth thing to note is there is incomplete information and isolation. So just contextual, may not tell you everything you need to build a complete picture of your audience and to be able to then power addressable advertising. The same is true for something like geolocation or first party. So what what we really need is technology that can pull all these multiple signals together, combine them statistically, and then try to understand audience behavior, intent and interest. And that is where AI and machine learning comes in. So for us, it's our, our AI and machine learning engine that we've been developing basically, since we were founded. So our our takes in all of these signals, combines them statistically to make sense of them. So let's look at how all of these signals comes come together. Alright, so the first thing we need to do, as I mentioned earlier, is we need to check for consent because without We cannot use any information. This is where our expertise and and our experience with Quantcast choice really plays a role. Quantcast choice is a consent management platform that is widely deployed that we have built. So this is where it plays a role. Once we check for consent, we can input the incoming bid requests from the Ad Exchange or a first party tag from our publisher partners into our system. Once we input them, then we go to the stage which is called parse. In the parsing stage, what we do is we take apart these inputs and look for all the different signals that I just talked about. So one set of signals is what can be called website signals. This will be things like the website URL, time language geolocation device, so on and so forth. The second set of signals will be the cohorts. This as an example could be Google's flock, Microsoft, I here has something a proposal called para keyed. So there's obviously there could be different types of cohort IDs that once they become available, will be attached to a bid request, and we'll be able to parse them out. And then finally, there's identifiers. So this is things like you ID, which will also come attached with a bid request.
Once we've passed all these signals out, then we enter the enriched state. In the enrich stage, what we do is we really enrich the incoming signals with Quantcast technology and with the unique technology that we have. So with that, let's look at the first way in which we can enrich and that is contextual. So the website URL comes in, it comes in with some standard content tags. But Quantcast has unique contextual technology called araz topic map that we built by crawling trillions of different web URLs. So we can look up this web your incoming web URL against that topic map, to be able to understand what really is the context and to be able to add in that rich, contextual understanding, then the second way in which we can enrich is first party. For those of you who don't know, Quantcast measure we launched in 2006 is a is a huge product that helps publishers understand their audiences understand the content and engagement and things like that. And we have first party tags deployed across 100 million plus web destinations. And so what we can do is, we can look at the incoming web signals, try to see if there's a first party cookie match, again, this will be something that will still be available and in the future. So we can try to look for that match, we see the match, we can add in that understanding as well. And then finally, the stage is the combine stage where we take all of these signals, the enriched ones, the other ones, like cohorts and identifiers, and we combine them statistically using arra, which is the AI and machine learning engine. So I know this is a lot. So let's actually look at a real world example. And, you know, try to look at it from that lens. So let's think about an example where we want to run an ad campaign to be able to reach people who might be in the market for ski goggles. Now, with third party cookies, or in presence of third party cookies, this was easier to do. And the reason is, we could look at different events that the user undertook. connect them together. As an example, the user checks snow status somewhere, then the user went and read a blog about how to ski, then the user when maybe went and looked at something like a review of ski boots. And because of third party cookies, the third party cookies are like a breadcrumb trail, right. So they helped stitch all of these dis this different disparate events together. And then it gave us a good understanding of Oh, maybe we should serve this audience, an ad for ski goggles. In absence of third party cookies, we still see all of these different events come in, it's the being able to thread them together. That is the challenge. And that is where all of these different signals, combined with that statistical combination using AI and machine learning can really help us so we can combine and match against geolocation. First, see if there's a first party match, look, look for consent, of course, that is very important. That's at the very start. Now, again, you might wonder as to like, Well, okay, so every match against the geolocation. That doesn't really tell us anything, it could be two different users from the same geolocation. That's absolutely right. But statistically, even if it is by a very small amount, the chances that it is the same user go up just a little bit. But then it's not just that it's also being able to look at all these different signals and combine and match against all of them in a high dimensional space using statistics. Goal method methods. So that's sort of the approach that we are taking. Hopefully this example makes it a little bit clearer. I would now like to just talk a little bit about where we are today in in, in this approach. So where we've started is we started with measurement. And the reason is, you cannot optimize what you cannot measure. So to be able to provide really good planning activation capabilities, we need to first be able to measure so that's the reason we've started with measurement. What does that look like in practice? So in the Quantcast platform, now you can see a third party cookieless conversion report. What is that? Exactly? So what you see in front of you here is, is a snapshot of what that report or such
a report might look like. The black bars that you see are essentially the conversions that are in third party cookie spaces. So those are just the conversions in the Google Chrome space. And we all know how we are doing those right? ads, conversion happened, third party cookies, the breadcrumb trail connect those, we can count it as a conversion, that that's the that's the basic effect. It may not be as as simple as that. But that's the basic of it. In in ITP environments, like Safari, basically, third party cookies are not available. So the blue bar that you see in the graph there, those are the conversions in ITP spaces that we have been able to determine using the approach that I just outlined earlier. So let's just take a little bit of a deeper look at what I mean by that. So essentially, it's the same approach, essentially, it's taking the same signals in and combining them statistically. However, when it comes to this conversion report, we are getting two sets of signals, we are getting a set of signal which is denoted by that those orange, you know the orange river that you see, which is coming in from the ad serve event or the ad serve website. And then we are getting another set of signals, which is coming in from the conversion website, which is the blue, what we do with machine learning is take the signals in of course, the parsing and enriching needs to happen. And then we take these signals in combine them statistically and then try to see if there's a match between the orange and the blue. And if there is a match, and we feel confident, I mean, we as in the technology feels confident about it stick with a certain degree of statistical confidence, then we count it as a conversion. Now currently, our engineers are earning on the side of being more fair and more accurate, which is to say we are only taking credit and counting conversions that we are very, very confident about. But with time, we want to also bring in the scale which is keep the accuracy but increase the scale. We have engaged with a few different customers as as our alpha customers on this, and the initial response has been very, very encouraging. If any of you are interested in looking at something like this, please reach out to us please reach out to your sales team member to try to get a demo of the Quantcast platform and also this particular functionality. So with that, in summary, I just kind of wanted to summarize the Quantcast approach. Well, first of all, the Quantcast platform does not rely on third party data, it has never rely on third party data since its very conception, does that mean that we are not susceptible or the platform is not susceptible to this? You know, third party deprecation? Not at all, you know, we will also still be impacted, but to a lesser degree. The second thing, like I said, we have architected the platform to take in multiple signals and combine them using AI and machine learning. Now, of course, anyone can do that. But we are doing this really well. We believe for the reasons that we have experienced and expertise across not just AI and machine learning, but
also
across many of these signals that will be important in the future. Example first party with our Quantcast, measure footprint, consent with Quantcast choice experience, as well as the unique topic map that araw creates. And that allows us to add in this rich contextual understanding. And then finally, like I said, solutions we have available today, we have started the cookieless conversion report is just the beginning. We have looked at Safari as a model of the future and started testing in those environments. We are actually even starting to test some campaign planning and activation capabilities internally. So this is just the beginning. Please engage with us if you would like to see a demo of of this solution. And then finally, you know, like I think all of us are navigating this brands, agencies, publishers, advertising technology companies like us. So we would like to just put forward these key takeaways for is focused on first party data. I know everyone's saying this, and you guys have heard this ad nauseum. But really, the reason everyone repeats it is because it is that important. First party data has always been important because it's important to have a direct relationship with your audience with your consumers. But it's even more important now. So focus on that. The second thing is make consumer consent part of your strategy. You may feel like
oh, well,
there's no regulation and like, I don't need to worry about it right now. But as Elena mentioned, this really is here to stay. Because it makes sense, consumers should have a say and should be able to consent. So make that part of your strategy. The third thing is test different solutions. As we've seen, this is a space that is constantly evolving, before locking in on any particular ID solution or any particular partner, look at various different options, test them. And then finally, choose a tech partner with strong expertise in AI. We don't want to say this because Quantcast has strong expertise in AI, which it does. But it really is something that if we are hearing from our tech savvy customers, from industry experts, so it's just something we think everyone should be thinking about. So, you know, we would like to hear from you, because a lot of this is not new, a lot of this is probably something you've seen before. And we are really curious, which of these resonate with you the most. So there should be a poll coming up. And maybe just, you know, tell us what, which of these resonates with you the
most? All right, we'll
give it a couple more seconds. And then we'll see what the poll results say. I am really sorry for any background noise. That's the leaf blower outside my apartment complex. Unlike Alina, I am in sunny california. So it's it's different kinds of noises, but still have them.
Okay, focus on first party data. Once again, I think we saw this a very similar vote on the earlier poll made consent part of your strategy, test different solutions, find a tech partner with AI expertise. Interesting. It's it's quite varied, for sure. But first party data clearly clearly wins in both these polls. That's great. So with that, I would just like to thank all of you for attending this webinar for participating in these polls. And we would you know, now would love to open the floor for q&a. And so I'm looking at the questions here that have been entered. Please keep entering more if you have any.
So,
so the first question is, so how will affiliate links as you sees? How will affiliate links be able to track sales? This is something that, you know, is maybe something that will be better answered by our technology experts. And so we will be publishing a blog right at the end of this week to capture the q&a from this webinar, so be sure to answer that one. Thanks for the question, Andrew. I'm going to say that we will answer it in the blog. Can unified Id tell if an email or phone is a bot? I would think so I think unified ID is being governed by you know, again, like Elena mentioned, multiple industry partners are coming together to define the standard and one part of it is likely going to be authentication, like maybe some some kind of double factor authentication and verification, things like that. So I think there's a good chance that we will be able to or other UID will be able to tell if it's a phone or a bot. Brian said all of the above. It's It sounded like it was probably response to the who will win question that Elena had an absolutely right, Brian, I agree with you. We are we are kind of on the same lines. So let's move on. clock is only one component of Google's approach. How does Quantcast think about fledge that addresses interest groups and contextual targeting? Larissa, great question. I would have to say that this is also something that you know, it's best heard from from our technology experts, so please watch this space. For more information, we will definitely make it a point to answer this one. In fact, I would love to understand this in a little more detail myself. So thanks. Thank you for the question. Paul does not include full list of solutions, Swan is noticeably lacking. That's absolutely right, Joshua. And you know, it's essentially the nature of it is like, you know, we would have to maybe even create a poll with those 80 people. But Swan is definitely an interesting, interesting one. Our engineers I know, are definitely monitoring it. So your point is very well noted. Thank you for the comment. Three out of four solutions are focused on engagement, but do not address measurement or real optimization. This is Joshua, again, I think maybe this was, again, about the the four solutions that we had on on the sort of the poll, which were I think, first party consent, contextual, and, and flop. And, you know, to be fair, it is it is right, like not all of them can address measurement, or real time optimization. And this is where I would love to get you an answer, definitely from our technical team. But this is where we truly think that a combination of them using statistical techniques can help solve some of these measurement and real time optimization issues. As we mentioned, we have started with measurement. And we are starting to experiment with real time optimization, but happy to follow up with you, Joshua, in terms of hearing more from you, and even maybe connecting you to an expert here at Quantcast.
All right.
So the next question is, how can we scale out and still target audience segments we may currently be targeting using a DMP, and third party cookies if we have limited authenticated data. This is a question from john del Vito. Thanks a lot, john, for the question. And this is this really goes back to the point that Elena made, that it is great if we have authentic data data and to the extent that we have it to whatever extent we have it, it helps us increase the accuracy in our modeling. However, scale is something that we truly hear at Quantcast believe that will come from deploying AI and machine learning from deploying statistical methods. So in absence of third party cookies, where the scale can come from is by looking at multiple of these different signals, taking a top down approach by looking at them statistically. And then also adding the bottom up deterministic data that Elena talked about, which is authenticated. And in this sort of a statistical combination, even a smaller amount of authenticated data can go a long way in increasing the overall accuracy of your model. And then the scale of course comes from using those statistical top down techniques. Happy again to connect you with one of our experts here at Quantcast, we will address this question in the blog in a little more detail. Thanks once again for for the question. All right, let's keep this moving. If 70% of brands are concerned that consumers will decline consent to use their data for marketing according to a recent Forrester survey, why not just give users opt in control of display of advertising at the domain level on on all of their digital devices including linear Ott in exchange of their PII. This puts the onus on advertisers, the agencies and the digital publishing publishers to make advertising as engaging and relevant as the content it supports. That is a really, really interesting point and a note and a suggestion. I think to some extent, this is this is this is a really interesting topic right? Because consumer consent and give giving people consumer consumers the ability to say yes or no is the first step. But with it comes this whole thing of like well for the most part then they might say no without realizing the consequences, that advertising powers the open web powers the free and open internet and powers the access to education and entertainment and so on and so forth. So we here at Quantcast truly believe that consent is the start what needs to follow is consumer education. So that they will, they will understand the consequences and then they will choose to opt in with the advertisers and publishers that are showing them engaging and relevant content just like you said. So very interesting point. What I'd love to engage with you more on this. Rob, please email us we will share some email ids later in the in the webinar. So we'd love to continue this conversation further. Is the contextual enrichment. multilingual?
Hi,
I do not know I have, I want to say yes, but this is something that will definitely need to follow up with our tech experts. Alina, do you have a sense of whether or not the contextual enrichment is multilingual?
I think it probably is, to a certain extent, but we would have to confirm with our engineers.
Yeah, I think that's fair. And again, we'll definitely answer it in the blog that will follow. What is the confidence level? percentage? So this is this, Brian, this is a really interesting question. We in product marketing, get it all the time, we have to say we are not engineers. All I can say right now is he engineers are setting the threshold to be really, really high. One thing that I've heard from our engineers, as well as our CTO is that it's not when we think about these confidence levels and thresholds, they are not static, they are dynamic. And those thresholds themselves are learned. Because it's very hard to say that, oh, I want 99% confidence, well, maybe 98% would have been good enough. How do you know and the way you know is you put in a measurement strategy, see what's working and go adapt that threshold based on what you're seeing what's working. So it's not really a constant one number that can be given out. Again, it's it's a really complicated topic. So happy to engage you within with the technical expert here at Quantcast. But thank you for the question. Brian Fraser asks, Will we be able to access the presentation deck aside from the recording? Great question. We were not planning on making this deck externally available. However, we have a lot of content coming through in the next couple of weeks, we are publishing an industry perspective that that should capture a lot of this content, and in fact, in even greater detail. So we will look out for that, we will email you with that. We will email you that as well as some other content that we'll be putting out. And then you know, we will certainly take this request into account and and maybe we will make the decision to share this deck as well. So watch out for that. Alice asked, Can the PDF version of the deck be shared or just the recording? I think that's the same question. Thank you for the question. Alice. Matt asks, How will you support ingestion of the consent signal coming from other cmps Elina, do you want to take this one, please?
Yeah, absolutely. Um, so we already addressed in recognized concern signals from other cmps through the bitstream today, and typically how that's done, at least in the EMEA market is through the transparency and consent framework signal, also known as the TCF signal. So that to my knowledge gets attached that TCF signal gets attached to the bitstream and pass throughout the industry and we see other CMP signals that way.
Awesome. Thank you. Kevin asks, Will Quantcast adopt UID? 2.0 100%? Absolutely. Like we said, interoperability is one of our core principles as we approach the future. And so with that, we will support you ID 2.0, but also other solutions that our customers might choose to engage with, that they've seen success with. Daniel asks, What testing strategies and or methodologies is Quantcast utilizing with its clients? Can you share some initial findings, please? So Daniel recurrently, we are engaging with our clients on that cookieless conversion reports. There are a few ways for the clients to engage with us and actually help us help them improve the accuracy of those conversion reports on their own campaigns. That is, if you are interested, our product team would be happy to chat with you. The program is still in alpha. So we do encourage really close collaboration with the product team. But that is the first thing in terms of testing strategy, the cookieless conversion report, being able to look at IDP spaces and try to use this methodology to figure out those conversions is our first step. It is really a really good ground for us to test this approach to make sure that it's working for our alpha customers and then start scaling it across other reporting capabilities as well as like we said planning and activation capabilities. Maxine asks Thank you for the question. Daniel. Maxine asks, Is it necessary to work with Quantcast to use the RR tool? I think the answer there is yes. Allah is Our AI and machine learning engine. You know, unlike many other machine learning capabilities in other platforms, ROI is really woven into the fabric fundamentally integrated with the platform. It's almost synonymous to some extent with the Quantcast platform. So there is no you know, here's our, here's the podcast platform, it really is just woven together. In the very foundation, there's no, it's not a layer of AI that goes on top of a platform we have AI is in the foundation in very core in the very fabric. So, yes, it is necessary to work with Quantcast. But please reach out to our sales team, there are multiple ways in which you can engage with us, and the sales team will be more than happy to talk about those. Sarah asks for a smaller agency, what is our best first steps to get ready for the end of third party cookies? Um, I think that that's, that's a really great question, Sarah, you know, some of the things that we talked about is, is certainly something that that you should consider. First thing is trying out a few different solutions. Now, of course, as a smaller agency, you may not be in a position to be able to try out multiple different solutions, like too many of them that is, but it will still be advisable to try out a few of them. The second thing is, of course, work with your clients, educate them about the first party data so that you have you know, a set of clients who understand this issue. And they are, they are protecting themselves and in turn protecting sort of their business with you. The third thing I would say is like look at your stack. As Elena mentioned, the dmps, the DSPs. The SSP is all of them working together is because of third party cookies. You should be asking hard questions to the players to the tech providers who are part of your tech stack. And asking them about how do you plan to work with all of these other things? What is your plan, and to that end, as Elena mentioned, look for partners that have that have less fragmented offerings and more holistic offerings. And again, this is where something like I mean, the example I know is contrast. So I have to talk about it. Quantcast platform comes built in with the data comes built in with planning, activation, measurement, ad serving all of those capabilities. So you know, our dependencies on other platforms are a little bit less. So it's more integrated. It's it's much more consolidated already. So look for partners who can do that, and ask tough questions of your partners. So thanks, Sarah, for the question. sariah asks, could you explain the ad server and conversion part a little more? I know, do you want to take that question?
Sure. So as Shruti mentioned, as we were going through kind of the approach that we're taking to a poster party cookie world, we're looking at a variety of different signals to help us kind of map together with a certain degree of certainty that this person is indeed in the market. First, he calls to go back to our to our earlier example, when it comes to the conversion reports, we're instead of looking at this like as one big picture, we're looking at it as two kind of cohorts, which is the group that has the ad served and the the group of the conversion. So in both of those different cohorts, we're still looking at all the same signals that surety outlines contextual as part of that first party, cohort, ID, etc, etc. And what we're trying to do using AI and machine learning is say, with a certain degree of certainty, that we believe, with a certain degree of certainty that this person who saw this ad was also this person who converted. So that's what all of our testing and optimization is around right now. And as already mentioned, there, there are actually ways that arm marketers who work with us can help improve the accuracy of our which we can report on those conversions for their own campaigns. And we're happy to talk more about that, and with our alpha partners, if they're interested in that. So, um, but it all comes down to our multiple signal plus AI and machine learning approach.
Thanks. So you know, we will also be addressing that actually mentioned these questions in the FAQ. So please watch out for that. definitely get our technical experts to help us there and you will get a little more insight. Patrick asks thoughts on changing the focus to content versus the consumer for better targeting and relevancy. This is a this is a really interesting question. And you know, there's one class position on this is that contextual signals have always been Very important to understanding user behavior and user intent. Whether that means, you know, leaning all the way in to contextual targeting or not, is something that is up for debate, we haven't technically seen better performance by doing just contextual targeting, which is why that's not what we do here at Comcast, which is just focusing on the content and you know, making the ads be relevant to the content, and less focus on the consumer. So we don't think that that's how the future will pan out. Because even even with just contextual signals, there's a lot you can do. Even without third party cookies, as an example, with just one website, typically, a lot of these publisher websites have so many different content. And with the first party understanding, you can still understand what are the different articles they read? And then what does that mean holistically, so it doesn't have so just because someone is on a page that is, let's say, and review about a car, it doesn't have to be an hour ad about car parts or tires or something else, because maybe on that same website, let's say Forbes, they also read about hiking, like quite a bit. So there is you know, it's it's a, it's a fine balance, contextual is definitely an important signal to us, going all the way in and leaning all the way into contextual targeting. It really, you know, we haven't seen very good results necessarily with that. That being said, you know, we don't know how this all is going to evolve, what our consumers what our customers are going to want. So we again, we are staying prepared for all of it. But that is the Comcast perspective right now.
Actually, I would just add to that question, can check the way that we think about contextual as part of one of the signals that we're looking at is different than actual contextual targeting? And I think the question is related to like, if I'm understanding correctly, and I can't pull it up on my screen, but the question is also related to like, should we be leaning more into a world where we're relying more on contextual targeting, and my perspective is, is from a premium publisher perspective, like contextual always has been an important part of their offering, and it's still will be an important part of their offering moving forward. So I think contextual will still play an important part of the future. I think, the challenges that the industry, especially from the marketer side doesn't want to rely necessarily solely on contextual because that's not what drives the performance to shooties. Point. I think what we've seen on the demand side is that like, there's still high expectations of performance. And while a media plan might include contextual as part of it, it's also going to need to include something that drives more results. And so I think that's where like, it's going to be like a part of the plan, but it's not going to be the entire plan.
That's great. Thanks, Elena. The next question we have from Kevin is, what do you expect these changes and adaptations, will do to CPA in general? That's, again, a really good question. I think not just for us, but for pretty much everyone in the space, we will see some impact on performance. We will also see much more discrepancy when it comes to measurement. Like already in the industry today, we have had a tough time of aligning on measurement methodologies, and being able to see similar results across different measurement methodologies, that gap will likely only widen for the time being till till the time that you know this, this is more in flux, and everyone's figuring out their approach, eventually then sinking their approach with someone else's approach. So I think overall, I think, you know, we should, as an industry expect to see some impact on performance, and some impact on the consistency in measurement across various different solutions. But that is not to say that, you know, we won't get through this, we will get through this. And that's where again, those industry standards, the interoperability comes in. Thanks a lot for question.
Sorry, I just saw something Surely, that's also what we're seeing, like, on the publisher side, at least, like that's why people are starting to test different solutions now. Because it's, it's people want to be prepared, like, what is the impact of this going to be? Like, let me start testing now. So I can get involved and start understanding like, what will What will my targeting look like without third party cookies? What will my measurement look like? Like? How will I be able to understand audiences? Like let me get a sense of that now and test all of these different types of approaches that are coming out so that when the day comes that Chrome does officially deprecate third party cookies, um, people have a good understanding of what's working and what's not.
Yeah, and
part of building out the solution is not just like, not just like inactive bystanders, like we're like people are have been active participants, and that will get us to the best place moving forward, I think.
Yeah. Um, and it really is, you know, it's important that we are all sort of, you know, honest with ourselves. And each other about this. I drew leaving, like if there's anyone necessarily saying that this will not affect anything and everything is going to be like Joe said as it was right away, you know, beginning tomorrow. That's naivety, that's that's promising things that are not true. Anyway, back back to the questions. Ahmed asks, What is the exact solution that Quantcast is offering. So I'm it. Overall, the Quantcast platform, like I said, is an intelligent audience platform that offers for marketers, campaign planning, activation and measurement capabilities. With respect to third party cookies, the solution that we have available, available today is the cookieless conversion report, that basically gives you a glimpse into the conversions in the ITP spaces in Safari environments that we have been able to determine without the use of third party cookies. Happy to connect with you further, and give you maybe a demo of the platform and a demo of the solution. Thanks a lot I met. Annie is asking. authenticated identifiers seem to be more invasive than the cookie. Do you see the industry cracking down on that in the near future? That's a really interesting question. I think it's also you know, something. That's where consent plays a role. I would love for Elena to answer this one. Yeah, that's it.
That's exactly right. Sure, do you. I think there's two different kinds of consent, there's explicit consent, which is what kind of guides the laws in Europe. And then there's implicit consent, which is what we live by here in the US. And if I have if I give my email and Elena Morris, and someone then uses that email for all sorts of different things, like maybe I only gave an email for a newsletter, and then my emails being used for, you know, a variety of other things across the industry, like, absolutely, that's a problem. So and I think that's where like when we talk about standards, like that's kind of where the heads out of people who are deciding the industry standards, like in Project react, for example. So I do think like, there's going to be a level of consent moving or sorry, a level of standard around consent moving forward with which emails will need to have explicit consent tied to them that that this email can be used for x y&z but not a, b, and c or something like that. And I think we don't know what that's gonna look like today, because those standards are still being created. But I'm, I'm, I feel almost positive that that's where we can expect.
Yeah, um, thanks a lot, Ali. Now we literally at time, I was going to try to swap this, wrap this up rather in two minutes before. But thank you, everyone, so much for the amazing engagement. We have many questions that have been unanswered, we will answer them in the FAQ that follows. Well, sorry, the q&a blog that follows the webinar. Any other questions, please don't hesitate to email us at innovate@quantcast.com. We have an upcoming webinar on May 4, where our cmo indrid burden and Patrick call who is an AI expert from b&h dot AI will be talking about bias in AI and how it is going to affect marketing and marketers. And then most importantly, we have an industry summit that Quantcast is sponsoring the cookie conundrum a recipe for success. It's on the May 19 that will be hosted by the cube and it will have a panel of industry experts sharing their thoughts and opinions about this very topic. So please watch out this space for more details on that. Thank you once again for attending this. It has been such a great pleasure and good day