The Future of Marketing in a Post-Cookie World - Alex Langshur, Cardinal Path

The rise of data regulatory compliance and the imminent death of cookies is a perfect storm that will give way to a newly-empowered marketer - one who understands and invests in the value of 1st-party data. ‍

This session will cover dominant market forces in digital marketing including:

  • The evolution of digital media spend trends
  • Avoiding unnecessary risks in data privacy
  • The impending death of cookies
  • Why identity is the new gold
  • Crafting a future-proof first-party data strategy


Alex Langshur

Co-Founder & Co-CEO

Cardinal Path

Alex is a founder and co-CEO of Cardinal Path, a premier digital data analytics firm and member of Dentsu Aegis Network. He works with leading organizations to create, implement, and action advanced analytics strategies that drive bottom-line returns. He leads the key client group responsible for driving value data strategies that integrate and activate data from across client's technology stacks. Alex hosts the Google Partners Podcast, has keynoted and spoken at digital analytics and marketing conferences around the world, is the past-President and Director Emeritus of the Digital Analytics Association (DAA), and teaches digital marketing optimization at the University of British Columbia.


Hi everyone, and welcome to this session on the cookieless future, and specifically, I'm going to want to talk about how organizations need to get ready now today for that cookieless experience that's coming in the very near future. I want to start off by saying I'm Alex Langshur, I'm a founder and co-CEO at CardinalPath. And in that role I'm really thinking hard about what our clients have to do to really be prepared for the segue into this basically changed environment for reach and measurement online. And I also want to thank the team at ObservePoint for inviting us to be part of this session today.

So it's rare that we get these opportunities in business or in your personal life to understand what the future is going to hold. And yet that's something that we really tried to do in businesses, kind of get a sense of what the future is going to be. And I think what we're at is this amazing point right now, where we know today that the way that we've been using our tool sets in the MarTech and Adtech space, to be able to understand the reach of our message and measure against that is fundamentally going to change. And that's due to the full deprecation of cookies, which will be happening in relatively quick timeframe. So how do we turn this into an opportunity for us? And the thing about it that I want to just point out to you , is that when COVID hit back in in March of this year, if we had known in November this time, last year that we'd have COVID, what would we have done differently? Like how would we have planned differently? And there's a lot of things that we might've done differently. I probably would have bought a lot more toilet paper, for example, but it's that type of foresight that we really need in the business environment.

So I want to go over what's happening and why, explain what the implications are for marketers at a very tactical level, then get into why identity is going to be so incredibly important going forward, as well as the criticality of first party data, and then some key takeaways. So, it's a jam packed agenda. Look, there are many reasons why this is all happening, but I'm just going to outline four that I think are pretty big.

Number one is that we know the consumers are far more aware about privacy and concerned about, and I'll show you how I know this to be true in the next slide. For the first time, we now also have a regulatory environments, both in the EU and parts of the United States. Brazil is having one, Australia. There's a bunch of places all around the world where frameworks are setting up that penalize organizations in the way that they manage data if they don't have the consent to use the data from consumers that they've received it from, according to what they've said. So before we'd say "oh, just collect it." Now we have to be very mindful that the over collection of data without consent and the use of data analysis of data without the proper consent to do so, actually brings in risk to the organization.

Secondly, as I'll show you right now today, there's a lot of lumpiness in the data that you're currently collecting and which kind of giving you some skewed perspective about how well your message is reaching people and understand the impact of that. So that's important to understand.

And then finally, the large ecosystem vendors are, um, happy about this to a certain degree because under privacy rules, they no longer really need to share data out. There's a lot of reasons why that's no longer a good idea for them, but for us as marketers, there's a certain degree of platform lock-in that we have to be mindful about. And I'll explain what that looks like in a sec.

So how do I know that the consumer is a lot more concerned about privacy? On the left-hand side of your screen here is a report from Blockthrough, which is our 2020 ad-blocking report which came out in may of this year. And the key story here is that about three quarters of a billion devices currently right now today have some form of ad-blocking technology and what that means that we think we're reaching people, but we're not because they're basically blocking the ability for our cookies to be placed. We know that some of the large telcos in this case, Verizon has now launched a one search, which is a privacy safety. It doesn't track anything that you do means for doing searches online on the web. Verizon has also gone and built the Verizon media group. It's trying to build a large addressable audience. So they're, they're thinking about this and they're putting position themselves inside of the consumer.

And lastly, if you take a look at the growth stats for Duck Duck Go, again a privacy compliant search engine it's spectacular. They're really at the beginning, the kind of hockey stick stage of their growth. We're seeing 3 million searches per day per month increasing, and then last month, 4 million per day per month. So a lot of actions and activities by consumers, which are saying I don't want to be tracked or I'm conscious about being tracked and I don't quite like it. And so the, the, the barriers are getting higher. Now, Apple positioning itself on the side of the consumer launched ITP, which stands for intelligent tracking protocol in 2017, which basically block third party cookies.

I live in Boston, I go to, it places a cookie on my machine. That's a first party cookie from to me, it sells ads. Those ad spaces are sold out in real time to somebody who places an ad on that space. There's a third-party cookie that's delivered to my machine. And Apple said, you don't have a relationship with the people that are placed in that ad. So we're going to block that cookie. Mozilla has ETP, which basically does the same thing. I want to take your attention to the right hand side of this timeline. First of all the cat and mouse game of fingerprinting and kind of tech oneupmanship which Apple kind of saying blocking those things progressively. I don't think that's a survivable or a long-term strategy, and we certainly don't recommend against it, but right now under ITP 2.X, 2.2, even first party cookies will be deleted after seven days. All fingerprinting is basically blocked and is blocking any first party cookies that are coming from a cross domain as well.

But the real death knell in all of this is Google's decision in this year to deprecate cookies by 2022. And so anyway you look at it, that decision is the death knell for this piece of technology, which has been so fundamental to a lot of the ways that marketers have worked in the past. And we really now know what the timeline is. The timeline is H1 2022 cookies will be deprecated. And so if we don't start to act now in advance of that, we're going to be in for a world of hurt. And it's not just there. Apple, again, always trying to position itself on the side of the consumers is also now deprecating the ID for advertisers or the IDFA. This is also gonna have some pretty big impact for anybody that's got an app, and that wants to be able to use that ecosystem and the way that it could in the past, it won't be able to. Now you have to also grant permission to be tracked with iOS 14, which I believe is going to be now pushed out to 2021. But the point being is that we're seeing it happen, not just at the browser level, but also at the app level too.

So what are the implications for marketers? Let's get into it. Red, and pale green and green. And the point I'm trying to make here is that what's going to be really heavily impacted, are anything that's really cookie dependent. So DMPs, which we're really reliant on cookies are going to be heavily cookie dependent, some things viewed through advertising or multi-touch attribution, those things are just history, right? They're Deadpool, they're gone. But what's going to be low impact, and I think we want to focus on the positive because that's my general kind of feeling about things, is that email will be still strong so loyalty and all the things that we've done around loyalty marketing will still be very there. Analytics in the main will be okay as well. And CDPs or customer data platforms, which are really trying to regroup all the information around an individual will also be in pretty good stat, I believe. Let's let's keep going forward here. Let's take a look at the specific ways that this is manifesting itself in your data today. Cause I think it's important to understand that.

On the left-hand side is basically a user journey with day zero, visit one under pre ITP. You'd set a cookie on the machine, you'd be able to measure that and see what they did during that visit. They lead, they come back for the second visit. We see the cookie, we append all the new information to the preexisting sort of building up that user journey. They leave, they come back for the third visit on the eighth day, we still see the cookie. We now append all that new information to the existing information. And now we're really developing a robust picture of the user. Under ITP 2.X that first party cookie even from Google is gone after seven days and maybe deleted after one day too, but after seven days, for sure it's gone. And what that means is that when you return on the eighth day, there's no cookie. So you're not recognized as having been there before. And we see this action, the data.

So at Cardinal path, we did this little bit of analysis where we looked at a lot of data from across a large number of clients for the same month, over three years, July, 2018, 19 and 2020. And what we see here in orange are all other browsers in gray, is Safari browsers only. And you'll see, in 2018, there was a slight difference, but basically they're about the same at about 58, 56% of returning visitors. And then when we go to July, 2019, we see that Safari returning visitor metric has dropped substantially by about 10 points. And then in July, 2020, it's dropped again a further eight points. So there's no underlying change in the way that people are behaving between these two browsers. What's really happening in this case is that Apple is blowing out the first party cookie, which is therefore deprecating the ability of the metric to work. And this is really profound for us because we may be making really good decisions, but on bad data.

So what does that mean for things like re-marketing? Well, it used to be that if I had a thousand dollars to try and reach a hundred people across any of the different browsers I could do so, but after 2017 in ITP, we start to lose the ability to see those people. And Mozilla has the same thing too, with blowing up third-party cookies. And the resulting thing is if I had a thousand dollars for a hundred people, now I have a thousand dollars chasing 50 people. My CPA's go up, my CPMs go up and remember, everybody else has an ecosystem trying to find those same people too. What's really interesting is that Apple announced at the same developer conference where they announced the deprecation of the IDFA, that they were going to treat and manage cookies, not at the browser level, but at the OS level.

And so what that means for us as marketers is that anybody coming from any browser, but on an iOS based device, we'll have cookies that will be blown after seven days. So what does that mean? Your retargeting audiences will not last beyond seven days. Your exclusion audiences will also deprecate after seven days so if you're using this a lot, that's going to be problematic. Your lookalikes only have seven days of see data. And that's really important to have a large data set for your lookalikes, because as we'll see later on, it's important to preserve the efficiency of your spend. So these are real impacts happening right now today, certainly on Safari and Mozilla browsers, but soon across all browsers.

So what does that mean for us as marketers? Well, it means that we need to move to something new and that's basically identity and first party data. And if you leave with no other message from this session, what I really want to say to you in the strongest possible terms is that the clock is ticking. When Google deprecates third party cookies, what that means is that we know that by that timeframe, if we don't have our first party data meant and assembled appropriately, if we haven't started to enrich it with identity and through PII, then we are going to be in for a world of hurt. So what, what that basically means to me is that if we want to go for identity and I'll explain to you why that's the case, then the future of our systems are really predicated upon a few simple things. One is consent, the other is identity, and all of that's wrapped up with privacy.

So if we say that I want information that belongs to you, that's private, then I need your consent to get that information. And that enables me to develop identity. And if we don't kind of look at these things in that triad then the organization's gonna find itself at a disadvantage, those that are so one of the ways that we can do that is we can look for, I'm not talking about contextualized advertising, which again, everything old is new. Again, I'm talking specifically now about if I'm trying to work in the new world, in which identity really matters, we're going to have to understand what our ability is to create identity. And that's going to come from us, being able to provide value to the consumer through having them authenticate. I think authentication is going to be a really, really important thing for us going forward. When people authenticate, when you ask them for something and you have a clear give-get, you'll be able to resolve consent issues, and it's going to really build your, the richness of that first party data, which is your own data. And it tends to be pretty future-proofed because when people authenticate, they don't tend to change the underlying things would be username or email and password, those things tend to stay the same. So it might not necessarily be an easy lift for many organizations, but it's going to be really important. And so this has to be done in holistic way across all of our channels.

And the message that I want to give to everybody is the following. I live in Boston and I, and I chew gum and I buy gum. And when I go to CVS, which is our local, um drug store here, and I buy a pack of gum and I use my loyalty card, I do. So because, um CVS is going to print something like this, and it's going to get cut off a little bit with the background stuff, but it's basically, I'm trying to show you a three-foot long receipt. And I do that because I'm going to get some value and the value is in the form of a coupon, $2 off or something of that nature. And, and I know that they're getting where we're on buying, how often I'm buying, what I'm buying, how, what things I'm using to buy, what time I'm buying them. I know they're getting all that really rich information and in exchange for that. And I trust them because I get my drugs there. So I think they've got some pretty good systems. So I'm making that decision that I want the value I trust them, and so I'm going to give them this information.

And when I look at the coupons, the coupons are very much aligned to what my interests are. So it's not about signing to push something that I'm not interested in, It's prioritizes to what I need. And I think this is another major message that I want to give to everybody is that we know the future going forward is going to be in identity and first party data. And if we want to be able to collect that, then we have to create an engaging user experience, which is about the user and this has nothing to do with the tech stack. It has everything to do about thinking about our brand ethos and how we incrementally ask for information in a context appropriate way so that consumers feel that they're not being kind of raked over the coals to give stuff up they're doing so because it's very much in keeping with the with the kind of purpose and goals of my visit.

And when we get that information, what happens is that we're moving from an anonymous IDs where I really I don't know anything, we're enriching that with PII to create, um, identity. And that's so critical because the ability for an organization to have its own ID graph, and in fact, even its own device graph is super duper important because the direction that we're heading to in this whole space is addressability. And we're looking to find addressable markets where I can reach individuals as opposed to cookies. And the only way that that works is if I have enough of my own first party data, to be able to create these audiences, cause you know, creating an audience of five or 10 people, it doesn't work. You've got to have enough attributes to be able to understand that this is actually an audience that's worth going for, for a whole bunch of reasons.

So we need to collect as much first party data as we can. We need to get our consent management systems in place that we can then add that to that PII, which we store in our appropriate place. And then we then use that to start to build, develop our ID graph, and from that audiences. Now, how do I know this to be true? Like how do I know that this is what's happening? Well, you've got the Google roadmap for their Google marketing platform here on the right hand side of your screen. And I want to draw your attention to the green box, which I've highlighted here on the left-hand side of your screen. And what Google is basically telling us is that if you want to make its marketing platform really shine, if you really want to be able to do all you want across all of that massive ecosystem that it has, you're going to need to bring your own data to the party, right? And they're telling us, this is the upper left-hand corner of this roadmap we read from left to right. And from top to bottom in Western cultures. And so that's the most important thing that Google is telling us. It is saying, this is what you want. So it's important to pay attention to that.

Now in the Google ecosystem, you'd be bringing your first party data into something called Ads Data Hub. Now, Ads Data Hub, or ADH, is basically a data clean room. And so what is a data clean room? And I think it's important that we spend a second or two talking about that because it's going to be very big going forward. So a data clean room, if you go back to the beginning of my talk, I said that privacy and the regulatory frameworks that are evolving around privacy have in them penalties, which are financial in nature, which means that there is risk for you to onboard data that you don't know it's provenance. If you don't know the consent with the data that you've onboarded, you might find yourself, actually onboarding a pile of risk. And similarly, you don't want to share your data with anybody else because you don't want that to come back to you because they've done something that's inappropriate with that data. So the data clean room is essentially a way of saying I'm going to bring my data into a place which is a privacy safe Haven. You'll bring your data into that same place. We'll strip it of all identifying information prior to bringing it in. So all you're really seeing are attributes associated with individuals. And all I can see are attributes associated with individuals. And that enables me to say very much like lookalike modeling.

I've got somebody I'm looking for, people who are identify as men between the ages of 35 and 45 who live within 10 miles of Boston, and to have a college degree, and who work in this industry or whatever you want, right? Whatever attributes you want, you define what they are with your dataset. You then match that up to the larger data set within this data clean room. You create a new audience based on that. And here's where the magic happens. Magic happens is on the other side of that, there is an ID, and that ID for those audiences can then be used to map out across any addressable media that's out there. Now, the thing about that is if you're using Ads Data Hub for Google that'll happens within that environment, which is great, but you're only going to be able to export and find audiences that are within the Google environment because ADH works for the Google ecosystem. Similar, what happens in Facebook stays within Facebook. Suddenly what happens is Amazon stays within Amazon.

So what happens is you're then going to get into a bit of this inability to really do cross platform understanding. So there are other data providers or identity resolution providers out there like Epsilon and Axiom and Merkle and Neustar on live ramp. And they, their value proposition is that they'll work with any of these ecosystems out there because they have access to them. So there's kind of a bifurcation that's happening here where some may rely on the, on the data sets of the platforms that they're using and others may choose to use ones that go outside of those platforms and give them true cost platform capabilities.

I also want to share with you, if you look we've seen that, like Disney for example, yanked all its content off HBO. Why? Because it built Disney plus, why? Because it's building an addressable market, you log into it and knows your ID. You use it on different devices. It has your ID graph. Basically it now is creating an addressable marketplace that it can sell advertising on. And that's happening we're seeing in many, many different places. And I think we're going to see that really expand where all kinds of organizations are gonna start to build their own Addressable markets.

Now, I was talking about the need to get as much first party data as you can. And the reason that I say that is because if you take a look on the right-hand side here, you've got that little picture of Ronald McDonald, which you can pretty much tell it's Ronald McDonald, but when I blow it up, it gets pixelated. And now it looks like a Stephen King kind of clown, like kind of scary. It's the same thing, if your seed data set's too small, when you blow it up, what happens is it loses resolution and losing the resolution means that you get inefficiency in your spend because you're not able to do as good a matching to the large, large data set that exists out there. So we need to be mindful about that.

All right, just a couple more slides and they'll get some questions. So I want to bring this all together. Um, what we're saying in the cookieless world is simply, this is that it's all going to rest on a foundation of identity. And that means that we really need to be able to centralize our information and make sure that we have control over it and make sure that we have consent attached to each data point. And that that consent always follows the individual around. So we need to have good consent management practice in place. We need to have awesome governance in place. We need to make sure that our data quality is constantly being managed. And we're also going to have to bring in more data as it evolves. So our data onboarding systems need to be there, but when we have that in place, then we can get to the audience insights. 

That's the place where we can start to really like tease out and understand the behaviors and patterns within this data set and create these large or these audiences. And what I want to drive here is that to do that at scale is something that's going to require MarTech stack. That's just the way that it is. And I actually believe that one of the roles, job descriptions of the future will be audience curator within any marketing team, because you can imagine that the, the ability to really understand these audiences and what drives them goes to the next level, which is orchestration, i.e. once I know who they are and what drives them, then I want to create experiences that are aligned to them in such a way that they resonate. And then when I've done that, then I can push that out across any of my channels that I want through activation. That's part of the ad tech startup too. So this is kind of the whole marketers tool set that we just see going forward.

And so if I kind of give you like the top four takeaways, and I'll look at some questions in a sec, is that if I haven't already made it super clear, you've got 14 to 20 months maximum to really get your data house in order. If you are not thinking about this very specifically, then you may find yourself getting spanked because the larger the organization, you are the slower it moves. And the clock is ticking when Google deprecates them, it's done. Right. So how are you collecting your data? Where is it being collected? Do you have consent attached to all of it? How are you managing that consent? Do you have the systems in place if somebody says they want to get a delete, do you have all that governance system in place around it? So identify, and then what are your governance systems around that first party data and what are your consent management framework around that?

Secondly, Google announced GA4 cause you know, we live in an app and web world. There's a lot of nuances with the GA4, it's not your standard GA tracking. You need to understand what it is cause it's gonna open up new possibilities. It's going to close some old possibilities. So that's again a non-trivial migration and you're going to want to start thinking about that now cause again, this is terminating sometime in H1 2022. If you've been reporting things to to senior management around year over year metrics, we want to really encourage you to put an Asterisk beside them and educate them that post ITP is going to cause your metrics to kind of shift. And if you've got anything, for example, that's relying on the cookie recognition and return visits, that's going to be problematic. So we're encouraging clients to think about their measurement framework and understand that their measurement framework might have to be device dependent. It may certainly have to be device dependent for a certain while. It may have to shift out some of the old metrics that you're looking at for new ones and KPIs simply because the platforms are not treating the data all equally.

And then finally you know, just as with COVID when everybody went out to buy toilet paper and they couldn't find it anywhere, we think that once people wake up to this, they're going to be looking around for identity and identity resolution providers. There's not many out there they're not all the same. These are not quick decisions that one makes. So you're going to want to be looking at identity resolution providers as an option to help you get there faster. And we're just saying the time to start looking at that and talking to those people is really now. So that's the four big takeaways and I'll look at some some questions here.

So I'll just go to this. "The problem is that we'll get flooded by org ID requests from all companies and people can get bored and declined just because you're in..." Yeah. Yup, that is true. So the question is "the problem is that we'll get flooded by org ID requests from all companies and people can get bored and climb just because you're number 100 on the list," and I understand that. We're going to be in a new environment and it's still got to shake its way out. And just as I think the adoption of any new technology we have challenges around how people use that. And some people use it better than others. You know, for example I go to a website and within three seconds, I've never been to it before they throw up a light box with a "subscribed now". And my feeling is that's just really not nuanced because I can't even read and you're asking me to subscribe, right? So not really thinking it through, even though we've known that for awhile. So I think this is the same thing we're going to have to get used to it. And I'm certain there'll be approaches that will evolve.

I've got another question here. "I don't understand how first party data can solve for attribution. For example, if someone clicks a Google ad in third party cookies no longer exist, how do we determine the ROI on ad spends by focusing on first party, we must map a nonplus campaign ID to new first party cookie, and then send that back?" Right. So I'll just summarize Brian's question, which is "how does first party data help us understand ROI?" And the thing that's happening there is it's all about addressability again, right? So if I'm creating an audience and I'm pushing that audience out to the world, and I know it's under a campaign, so I've created this audience, let's say I'll stick with Google, I've gone through Add Data Hub and I've built their odds with them, and now I'm pushing my campaign across their ecosystem. I know what the campaign is, if somebody clicks back through that campaign, that's how I'm going to measure it because Google's not going to tell me WHO it exposed that ad to, It's going to say your ad there was 5,000 people in that audience and when we exposed your ad to 2,500 of them, so that's the way that that's going to work. And it's going to be a challenge for us to really make sure that we know the impact of that ad. And this is one of the benefits of working within the ecosystem, for example Google's ecosystem, is that they're gonna be able to close that loop a lot better than most, right? It's not going to be equal everywhere. So that's just one of the new realities that's going to be out there.

So want to say thank you again. Just to land on this, this is a big problem and it a big challenge for our industry. We need to really make that our clients are aware of it. It's hard to collect first party data. It's even harder to enrich that with PII to create identity, but the creation of an ID graph and the creation of device graph. I see any real way around that for us going forward as marketers in the future, and it's a challenge that that we all have to rise to. And I wanted to say, thank you to everybody.

Previous Video
Browser Privacy: The New Normal - Cory Underwood, Search Discovery
Browser Privacy: The New Normal - Cory Underwood, Search Discovery

Discover how the new browser privacy initiatives affect you and what you can do about them.

Next Video
Digital Transformation: The Billion $ Scam - Steve Carrod, DMPG
Digital Transformation: The Billion $ Scam - Steve Carrod, DMPG

Discover what a successful digital transformation programme looks like with the customer at its center.