Do you feel prepared for the death of cookies? We wanted to know how well prepared the industry is, so we scanned the top 100 sites in the US, UK and Australian markets to find out. Join James Wawne, Managing Director and Principal Consultant, at DMPG, Cameron Cowan, Senior Director, Product Strategy, and Dylan Sellers, Customer Solutions Engineer at ObservePoint for a discussion on all things cookies and discover the answers to questions like:
- How large is the third-party cookie dependency?
- What business functions are most reliant on third-party cookies?
- How can you understand your exposure and mitigate risk?
- What are some cookie alternatives you can start implementing now?
- And more!
View the full report HERE.
Cameron Cowan: (00:03)
I'm joined today by my colleague, Dylan Sellers. Dylan, why don't you say hi to everybody and introduce yourself?
Dylan Sellers: (00:09)
Hi, I'm Dylan Sellers. I'm a solutions engineer at ObservePoint site product experts. I work with anything as simple as just getting on the UI and as complicated as building custom integrations with our API. So I spent a lot of time in this space especially excited about privacy space and what it means for the new era of customer experience.
Cameron Cowan: (00:33)
Awesome. And we're happy to have you. We're actually gonna be pulling Dylan in throughout he helped lead some of the technical side of how we actually got and collected a lot of the data we're going to be reporting on too. Also with us today is James Wawen. James joins us from DMPG, our partner agency that we've been working with and really had a lot of the good ideas about how we should go about this whole process and why we should be learning the things that we're going to be presenting. So, James, why don't you introduce yourself to the audience.
James Wawne: (00:59)
Thanks Cameron. Hey everyone. I'm James Wawne. I'm the MD at DMPG in Australia. We are a digital experience enablement services consultancy with teams in London and Australia most recently. And we really specialize in helping consultants clients to leverage their data technology, people, and processes to extract value and deliver compelling customer experiences.
Cameron Cowan: (01:27)
Awesome. Thanks James. And James, how did you get involved in being a partner and working with ObservePoint?
James Wawne: (01:34)
The story actually goes back around three or four years when I was actually based out of London. I was working for large global energy company that had a pretty expansive digital landscape. And some of the many challenges that we were working with, around data quality assurance, ensuring and maintaining our tag landscape and ensuring that we were privacy compliant, and we had a good grip on things, necessitated really a solution that supported scalability. And that's where I became aware of the ObservePoint solution. And I actually worked with a couple of colleagues in that organization to procure the solution, bring it in, stand up those processes and really sort of automate a lot of that heavy lifting to ensure that we were in control of our data, that we had good data quality and had a clear understanding of our landscape and our exposure.
Cameron Cowan: (02:28)
Fantastic. Well, we are happy to be partnered with the DMPG and working with you. Now, obviously, this project specifically grew out of a little bit of necessity with the obvious changes that are happening around the globe in various markets around data privacy, around the way browsers are now collecting that data, certain restrictions they're putting in place. We wanted to cut past a lot of the hyperbole and the fluff in that the world is ending, and all cookies are going away, and really get some hard data we can share with our audience and our customers. Tell us a little bit about how the genesis of this specific project came about.
James Wawne: (03:00)
I think this sort of comes back to, there's been a growing sort of buzz around the industry as you touched on there. And I suppose that that most notably increased around January 2020 when Google Chrome announced that they would be phasing out support for third-party cookies. And I think the industry since then has, there's been sort of a growing focus around what that means. And not withstanding the sort of recent development where they've extended that deadline from 2022 into the back end of 2023. I think there's a lot of chatter in and around the industry. And again, as you touched on, I think it's really important from my point of view, when I'm coming at and evaluating problems or opportunities is to scale the problem. And there wasn't really much across markets in way of hard data. And so, you know, it was in conversation with some of the guys in our team, and with some of the guys on your side, that it really sort of came to me that we could undertake quite an interesting piece of primary research and put some bright lights over this, really get under the skin of the scale of the problem and dig into the details.
Cameron Cowan: (04:08)
I think you're right. It's getting hard data is the key to it. Not necessarily getting pulled along by all the clickbait and all the, you know, people kind of pulling their hair out, but understanding what is the true exposure, what's the true problem, and how do we, how do we address that now, Dylan, I want to bring, bring you in here for a second. James mentioned specifically third party cookies, and I think most people that are listening in today know the distinction between first and third party. But just so we're all level set, what is that distinction? And obviously Google has announced that they're going to be phasing out support, Apple and Mozilla have already started to block third party cookies by default, but it helps us just level set. What's the difference between a third party cookie and a first party cookie?
Dylan Sellers: (04:45)
Yeah, we'll start with the first party. I think those are the more innocent cookies, the cookies that we we take ownership for and responsible for implementing on page. They're set through our own network requests, our own network calls. And then the third party is essentially any, any cookie dropped or set by a third party vendor. So we look at the cookie domain attributes and we, you know, we certify that, all right, I guess we distinguish between those two and make it really clear in our reporting. But what's a threat, and what's going to be a concern here in 2023 is those third parties. And so the better we understand definitions and the better we understand the applications and really the consequences of phasing those out, the more prepared we will be. And the less shocked we will be here in 2023. And just in the years rolling up to 2023, when, when those changes have to happen, we have to be prepared.
Cameron Cowan: (05:46)
Okay. And I appreciate that because really the focus of this research that we did was very much on third-party cookie exposure. So we're gonna leave first party discussions for maybe another day, but as far as how we set up the scans, just gathering the data, then Dylan, obviously you were involved in that process. Talk to us a little bit about the methodology.
Dylan Sellers: (06:05)
So ObservePoint's, one of its core solutions is its web auditing technology. And what we did is in order to address and simulate user experience in these various markets and Australia, UK, and us that we studied, we started by configuring our audits to simulate that particular proxy or location. So we have visitors hitting the site from those locations. We then when needed or required, we provided some actions to occur or be executed on the page in order to simulate a user choosing their own adventure. Whether they're going to accept the cookies or decline. In certain geographical locations, we have an assumed response. Like in Europe, we expect people to be opted out by default, and in the US and Australia, we're still opting in by default.
We'll do a little bit of the breakdown later on in this webinar, but as far as the configuration goes, our audits will crawl once they've been configured to accept or decline cookies. And then we're just collecting the person's third party cookies, categorizing them for you and making it very clear the source and the origin of those cookies. And then, for this particular chat today, we're not gonna focus on the tags, those are also pulled in. And there's definitely a cookie tag relationship we shouldn't overlook. But definitely we zoomed in on cookies and their persistence in the sessions that we that we simulated when we crawled the sites.
Cameron Cowan: (07:44)
Great. And how did we pick which sites we were going to look, where do we get that list? How did we decide who we're going to look at?
Dylan Sellers: (07:50)
Yeah, so we had a, a list procured by similar web. They were kind enough to provide us a stack ranking of our top most visited, highest volume sites, basically. And we made some exceptions, some exclusions based on what websites with adult content, websites that were inaccessible. But we tried to curate that list, and then we wanted a really solid sample group. And we didn't want to contaminate with any more additional exclusions, that's pretty much as far as we excluded. And we made sure that that list was reflective of where people are going. And that's really what it's all about and making sure the study is most applicable to all of us, just to say, "Hey, where are the highest volume of people going and what is their response?" As far as cookies and privacy compliance goes.
Cameron Cowan: (08:46)
Awesome. And I know that in the data that I've seen we obviously picked a couple of very specific markets. James, I think you mentioned Australia, the UK, the US Tell us a little bit, James, as far as why we picked the specific markets that we did pick.
James Wawne: (09:00)
Yeah, thanks, Cameron. I think the reasoning here was to get quite a sort of interesting point of reference for comparison. So obviously Europe, and to some extent the US but certainly Europe at the forefront of pushing and championing, if you will consumer trust and privacy. Obviously on the back of the introduction of GDPR in spring time, 2018. And I think, as we can dig into the data, we'll see some notable and interesting differences though equally, quite a lot of consistency as well. But I think what we were looking to establish also in the selection of the data provided by Similarweb, is the fact that it's a reasonable working assumption as a starting point to assume that the top end of town that is the most heavily trafficked the most well-funded the most technologically advanced in, in theory websites will have a good handle on this.
And so the starting sort of point there was, if we use the quote "stop end of town," the most traffic sites as a proxy for what the current state looks like, which is the data that we really wanted to get a clear, clear view on. Then there would probably be learnings that could be sort of drawn from at the lower sort of mid and smaller size scale sites and the "smaller end of town." And I think, we had a crack at looking at that a little deeper by looking in the Australian market at the top 100, but then juxtaposing, that was the top sort of 101 through 200. And seeing if we saw a Delta there, and as we we'll dig into a little later, we didn't actually see that level of the strata as much differences perhaps one might have expected. And so therefore, you know, thus validating to some degree that working assumption that there is a fairly consistent level of exposure across the the sort of the digital web landscape. And I think that's something that therefore can be drawn across and is potentially applicable as a starting point, at least for everyone.
Cameron Cowan: (11:14)
You talked about the data, that's why people are joining today. So let's actually dig into some of the initial and high-level findings that we pulled from those scans that we ran. So give me a second to switch over, and here we go. So as we mentioned, there's a couple of different markets that we looked at, and as you said, we also looked at a couple of different breakdown segments within those, for example, the top 100 of Australia versus one-on-one through 200. And just talk to us a little bit, James, about what are some of the trends you either saw or the things that maybe you expected to see that you didn't see?
James Wawne: (11:50)
I think right out of the gate, most interesting thing was, was broadly speaking, and I'll come back to this, there's a notable difference, obviously between the UK, a 100 with the gray bar and the UK 100 opted in with the black bar, which I'll come back to. But at the highest level, and I think the answer to our first question we wanted to answer, which is what is the extent of the exposure here across the top end of town in terms of third party, cookie dependencies and broadly seeing what we can see is quite a consistent picture. So give or take around 82% of the sites we looked at had third-party cookie dependencies. And so therefore, you know, the key takeaway from that is look, this is something that is significant.
This is something that businesses need to pay attention to. You can't just ignore it. You can't just well, you can. Of course you can hope for the best, but that's not often a good strategy. But you know, what we saw was quite a consistent view there, as you can see with a couple of percent variance in terms of spread. I guess coming back to the UK point, this was quite interesting to see because we started off with, and touched on in methological approach. But basically arriving at the site from the corresponding sort of geo point of presence, running the crawl and pulling the cookies that we could see across those domains. And first off we saw quite a difference in the profile of the UK sites.
And of course, we dug into this and did cross validation, cross verification and fairly quickly determined that a lot of the advertising cookies were not being dropped in the first instance until such point, because obviously there was a much greater prevalence of consent management solutions. And so until you actively opt into those solutions, a good chunk of the advertising related cookies were not being dropped on the browser. So that was something that then Dylan and the team, I don't know if, Dylan, you want to perhaps briefly touch on that, how we looped in your script servicing correlates how we address that challenge.
Dylan Sellers: (13:53)
Each site had a unique, especially in the instance of the UK, right? In order to have a comparison between the UK opted-in scan results and the Australia and US markets, we had to basically configure our scans to visit each AH and accept cookies with each independent consent management solution. So with a variety of consent management solutions out there. It was a little bit of a fun task to attack. I'd say I wouldn't do it again, but we love automation at ObservePoint and so we always, we always look for the automated solution. But yeah, we went through and configured, for the UK especially, we had configured every single time we visited, we'd accept and reload the page and see the cookies that set in that after that experience.
James Wawne: (14:50)
Yeah. And I think where we see obviously quite an interesting visual on the impact, there is the shift of what we saw that prevalence around 69% in the UK of sites with third party cookies. You know, which raises other questions and things to look into, of course, for the organizations in the UK. But obviously when we did provide that explicit opt-in that, that brought the numbers much closer to the parity with the other markets. And so that then just, again, sort of reiterating the point of this is a consistent problem across the markets, you know, around 80%. It's quite interesting.
Cameron Cowan: (15:29)
And you call that exactly, what stuck out to me is that even in the default opted out state, 69% is still a lot higher than I was expecting, especially given all the new restrictions that exist in the EU. I would have expected that number to be south of 50. So that's, that's an interesting and core finding in and of itself. Now obviously the data we're looking at here, it's nice to see there is some consistency across markets with different regulations. It's interesting to see we're talking about the exposure that you have to third-party cookies, but really that's not saying how many third-party cookies exist at any of these sites. It's simply saying, are there any? So even within that subset of people that we scanned, you may have one, or you may have dozens or scores of different third-party cookies that are prevalent. Am I thinking about that the right way?
James Wawne: (16:18)
Absolutely. And I think when we look at the table of that data it's a little bit confronting, so I'll just walk you through what we're looking at here. And what was quite quite interesting to note was a very significant distribution and spread so range in terms of the number of third-party cookie domains. We counted across each of the sites we crawled. And this is sort of indicated, and this is partly why we included both median and mean here. Because obviously the challenges and potential floor of averages here, we've tried to mitigate to some degree with the use of those different approaches to arriving at an average view. And of course you can see that quite a significant variation between the mean and the median indicating that, of course, we've got some skewing going on.
And so actually also, if you look at the the maximum values there on any given domain that we crawled within the sort of three or 400 domains we've looked at in total, we found sites that had, and this sort of speaks to really the potential task at hand that companies may have ahead of themselves when they're looking at this. But of course everything from companies that didn't have third party cookie dependencies whatsoever, all the way through to companies that in the US market, we saw one, for example, one company had 110 different third-party cookie domains being set for the technologies and the capabilities that they had present on there, on the website. And again, on average I think it was around 20 cookie domains that we would expect to see in turn that related to I think 39 cookies unique named cookies as well.
And in terms of the actual unique domains we saw, we can get into some of the details of the distributions, perhaps. I think we, we provide a lot more detail in that report that will be made available at the end of the the webinar to those listening. But, we can see quite some quite interesting distributions there. Quite a lot of concentration, very much a Pareto distribution at the thicker end of the distribution tail. And anywhere from around 150, in the case of 147, in the case of the UK unique cookie domains which of course increased the 220 once we'd opted in. But obviously the US topping out there with on average 283 unique third-party cookie domains that we saw across the 100 sites that we crawled, which when we add those up and sum those, rather than counting them uniquely, translated to account just a under 2000 counts there, in terms of the total volume of third-party cookie domains we counted against the 100 sites that we evaluated.
Cameron Cowan: (19:11)
This is where I'm actually starting to see some of the trends I expected to see. While the UK, even in an opted-out state was a little bit higher than I expected to see there just overall distribution. I mean, the average is still only seven third-party cookies for the UK, whereas we see it much higher in the other markets. The maximum value is only at 49. So there's probably a couple sites that are skewed, but even that maximum isn't even coming close to what we're seeing in Australia, the United States. So some really interesting findings when we start to dig in a little bit deeper across markets, which I think is fantastic.
Now I do want to talk a little bit about like the types of cookies that we found. Obviously, not all, all cookies are coming from the same type of technologies are serving the same functions. James, talk to us a little bit about the different categories you observed as you dug into the data.
James Wawne: (19:56)
What we started off doing was getting an aggregate, but de-duped to market view across the landscape. And we stack ranked the counts that we found for those. And so almost starting from the bottom up of this table, any domain, of which we counted around 460, I think de-duped in total, across all markets, we had to basically go through this manual exercise of codifying and identifying for the purposes of informing this analysis, the actual nature of the cookies that we saw, or the cookie domains that were setting those cookies. And so where we, from a methodology standpoint, where we given that distribution, and the fact that it was accompanied with a Perino distribution, any domain that we saw appear only once we bucketed as unclassified. So effectively we deemed that this equated to a non prevalent domain for want of a better way of putting it, and as such from a classification perspective, we group those within that bucket. And then any that we couldn't get to the bottom of and these are still pretty low volumes in terms of what they equated to we bucketed any that we couldn't identify as unknown for the others, which represents the lion's share of the cookies. We encountered the ones of interest you might say that occurred two or more times. We bucketed them in a slight variation from the, if you like standard approach of necessary functional measurement and personalization. We split these out a little bit further to make the distinction between measurement as well.
And so effectively unnecessary cookie was anything that was basically required that was necessary to the site. So things relating to log-in, and things relating to navigational payments, perhaps, that was defined as a necessary cookie. And in the main, these were typically, from a best practice perspective, be instrumented from and set from a first party context to obviously mitigate potential risk of it being in a third-party context in getting blocked, and then potentially introducing issues of from a user experience perspective. The second group there, functional, we've basically used to bucket non-essential cookies, but cookies that potentially were important for a certain user settings. So language settings, localization, things of that nature and also things relating to functionality on the site that was potentially not essential, but again, not something that should be ignored. So it might perhaps relate to third-party solutions like live chat, social sharing, et cetera. And then the next real advertising measurement and personalization are fairly self-explanatory. Obviously the distinction between measurement typically being analytics and attribution solutions and potentially tag management solutions. I think we bucketed within that also. And then finally personalization solutions, your AB testing solutions and things of that nature.
Cameron Cowan: (23:16)
Awesome. And so let's now talk about that actual distribution and what we thought was I guess in a little way, not overly surprising to me advertising, taking the lion's share, but certainly a fair bit of distribution throughout the other types as well. Talk to us just really quickly about what you saw from that distribution. Yeah, absolutely.
James Wawne: (23:36)
And I think obviously as we might've expected from a starting point, you know, we see on average there, and you can see the bottom of the table and the advertising the majority of third-party cookies identified 84.9% did relate indeed to advertising technology. But that said, there were some notable dependencies, shall we say, of almost a second in rank relating to measurement? Obviously we could also call out there that the UK, that actually increased almost double the average to just under 12% of the cookies count, cookie domains that we identified, encountered pertaining to measurement. And then obviously where we might then look at grouping, if you like functional unnecessary cookies to just under 3% perhaps. Certainly if we include personalization as well, if we group those three together, we identify a potential exposure, should those technologies or solutions not be able to function properly. Once third party cookies are blocked and obviously one needs to look at that on a case by case basis, and dig into the details on a site by site basis. But there is a potential risk there for a degradation to customer experience that is worth noting. And obviously companies ahead of time should seek to investigate, understand, and get under the skin of, to make sure that there isn't that, if you'd like sort of critical shelf drop when third-party cookie support is phased out from Chrome and other notable sort of browsers, obviously to some extent already in play for Safari and Mozilla, as you touched on already, Cameron.
Cameron Cowan: (25:27)
Awesome. So that as I jumped out, then let's connect to our discussion board here. I just want to, obviously the purpose of today's discussion was to present those findings to really get to our audience. Some of these interesting things, as far as hard data, what is the exposure we're seeing and across markets and across categories? So I really appreciate the effort that both James, you and Dylan put into preparing this data. As you mentioned, James, after the webinar, everyone's going to get a copy of the full report, including some analysis and deeper dive, but just in our last few minutes here I want to take a moment and say, what's the, "So what?" I mean, obviously, as you said, Google has now pushed out their guidance almost a full two years. So a lot of people I think are breathing at least a little bit of a sigh of relief. But that's no reason to let up and pull your foot off the accelerator. What are you expecting your clients, and really just folks that are involved in the digital ecosystem worldwide to do with the insights that we've found here?
James Wawne: (26:24)
A good question. And I think it's a fairly simple one to answer in my mind. And I think first off, it's worth acknowledging that companies need to take note. So as you touched on already the Google delay, as it were, is a more, I would sort of position it as a reprieve, a bit of a stay of execution rather than a, "Oh good. We can not worry about this for another year or two. I think this is an opportunity for companies to go, "Right. Okay. Well, look, the data here clearly shows that there is a significant exposure." And on a company by company basis, my steer to clients is to make sure you get on the front foot, be proactive and start off with scaling the problem. Right? So half of the battle in any risk situation is to scale the risk. That's the starting point and the foundation from which companies can then sort of prioritize, raise the prominence within the organization, coordinate with team members across the organization to ensure that they are prepared and that they mitigate this ahead of time. So once they've scaled that risk, of course, then they should dig into as the second step the details. And then once they get into the details, of course, and they understand the ins-and-outs of the exposure and the risk of that presents themselves, they should build a roadmap to mitigate the risk.
Cameron Cowan: (27:55)
Yeah. And, and as far as quantifying, what is my exposure? What is the risk to my business? I mean, Dylan, obviously you've been here at ObservePoint for a while. We have technology that does this. What would your recommendation be to our active customers, as well as really anybody that wants to understand what is that exposure?
Dylan Sellers: (28:11)
Yeah, I'd say take advantage of some of the new reports we just kicked out. Cookie inventory categorizes your first and third party cookies easily for you. Run these scans and at least get a sample, a taste of how your website responds. I think I've worked with, in the past year I've worked with a number of clients and I have yet to find one to stump me and say, "Oh, look, I got an A+." All of them have some bad marks, and areas to improve, and the sooner they get on it, the sooner they're gonna be able to respond, the sooner they're going to mitigate that risk and not just scale the risk, but scale the solution too with ObservePoint.
Cameron Cowan: (28:54)
I think that's right. No business is a 100% perfect in their digital world. And as things change constantly, things break down and you want to keep an eye on that. But you also want to understand, not only from a high level, do I have exposure or not? I think we saw that in the data, 82% across the board, roughly have some level of third party, but also what is my range? Am I on that lower end where I only have two or three that I need to take care of, understand what they're doing now, and how I can and should be looking to replace them? Or am I on the far end? And I'm looking at a 100, 105, 110. I really got a big road to plow ahead of me. So I think there's all sorts of opportunity just to understand where we're at.
And then finally, James, you had mentioned a roadmap, and I think that's where I want to end off today. As far as building out a roadmap to address these things, even with the stay of execution, as you said from Google. What should people be looking do in the months and quarters ahead?
James Wawne: (29:44)
I think honestly beyond that, I think any roadmap worth its weight should also be underpinned by a coherent strategy. And so, in the first instance, and anyone that reads everything that I've written on this topic recently in and around the trust and privacy landscape, I think will recognize the point I'll make, which is around sort of renewing the emphasis and increasing the emphasis in broad terms also on the primacy and importance of first party data. So obviously where we're talking about third-party cookies to some degree, the risk exposure can be controlled, but obviously where we're talking about third-party data and third-party vendors to some degree, you're in receipt of on the receiving end, certainly when it comes to some of the top "end of town" solutions and technology platforms that you're working with. There is a dependency and control sits external to the business.
And so as far as businesses are able to, I think they need to sort of seize upon this focus at the moment to also position themselves with a future state strategy that very much puts first party data and first party instrumentation, and moving away from cookie dependencies entirely, where as far as possible, and that's something that's going to be receiving increasing focus and discussion, I think in the coming six to 36 months, really. And effectively I think companies to that end need to prepare and make ongoing investment in both technology capabilities, people processes, and around their first party data assets and how they curate and manage those solutions. Because I think, obviously the benefit that comes with first party is full ownership and control of everything around it. So everything through trust practices, everything relating to how you then use that data to personalize experiences across channels for users. So I think first party data would be first part.
The next point, I guess, briefly is going to focus around within the constraints of what their profile and risk is already, look to fix and shifting from a third party to first party context where possible, so work closely with your agency partners, work closely with your technologists internally and tech support on the solutions where you have those dependencies. But also, I think we're going to see fundamentally sort of a Renaissance within the ad part of that quadrant to contextual targeting. And there's going to be, if you like Back to the Future from obviously those skills and those approaches that we that we saw before they were supplanted by cookies, of course, going back some years now.
So I think the next couple of points I was just going to make on this one was looking to explore second party data, ship data partnerships. I think that's going to be a massive growth area looking to whether it's private data exchanges, whether it's a private data clean rooms, there's a whole bunch of buzzwords and emergent things that people need to be vigilant to and get their heads around and proactively look into. And then in final point, I suppose, in the context of where all of this originally started from working within the walled garden environment. So I'm talking Google, Facebook, Amazon, and others, notable big platform, ecosystem providers. You know, they're all jostling for position at the moment. You know, and we see this week to week, things seem to change deadlines, get pushed back initiatives to some degree get accepted or rejected by one or other party in frame. We've seen a lot of this going on in the last sort of 12 months really. And I think the only viable advice around those ecosystems is to remain vigilant and to plan, to be agile, and to plan to move quickly with agency partners as far as possible to leverage for the, the features and capabilities that they introduce or that they pivot towards. And yeah, keep a close eye on the news.
Cameron Cowan: (33:48)
Yeah, no, I think that's fantastic guidance. And what sticks out to me from what you've mentioned is being able to understand that shift from third to first party where possible. From a cookie perspective, obviously we want to consolidate, but even getting to more authenticated states. I think everybody's talking a ton about that right now, and that is that's vital to all of our businesses. So I think that's important. I think second party data is often overlooked building those strategic partnerships. Everybody has their own first party data. And up until now, everybody's been able to buy packaged third-party data, but really building the competitive advantage of having key strategic second party relationships I think is really huge. And then even with the third party advertising ecosystem, especially like you said, that Back to the Future of contextual targeting.
I remember when I started my career at Omniture back in what, 2005 now, that was what Yahoo, and Overture, and Bing in as, before Bing, Google, that's what all of them were doing. And so getting back to this idea of advertising two cohorts of people that show similar interests based on the sites they're on, it's not ideal, but it's still going to be possible. And so I think we have a lot of levers and a lot of options. We just need to make sure, as you said, build a roadmap and plan for it as we go forward. So fantastic analysis.
All right. Well once again, I want to thank you guys. This has been some great research. We're going to be sending it out to everybody that's registered today for the webinar. Dylan, appreciate your technical chops and everything you did to help us get the data and set this all up and deal with some of those nuances. James, can't tell you how much we appreciate both the partnership, but your in-depth analysis and the way you're looking at this. Really, as you said, and as a final thought, it's not just around technology, which is what we do at ObservePoint. It's not just around services which what you guys do at DMPG, but it's around all three technology processes and people, and being able to understand how do we get all those things to be working together so we can be in a better spot going forward.
James Wawne: (35:39)
Absolutely. Thanks, Cameron.
Cameron Cowan: (35:41)
All right, guys. Thank you everyone for joining us today. And until next time, we'll see ya.