Podcast What’s Advertising Got to
Do With a Free Internet? Season 1, Episode 1

Somer Simpson:

I think there’s a lot to be said about how we got to where we are because I’ve been in the industry for a very long time, pre-web, and I remember when the biggest struggle you had at the newspaper was running out of available space for the editorial, right? Because you plug the ads in, you left some space, you put the words in, and digital has completely changed everything. And it’s completely changed a lot of the aspects of this system. Walk me through, from your perspective, what are the most critical milestones in the past 20 years that had the biggest impact on where we’ve ended up?

Scott Messer:

Biggest things were… obviously, the change from ad network to SSP, I think is the biggest one. There’s some things that really facilitated that and went with it, and that was the invention of the cookie, seemingly harmless and yet grew in ways that no one could have imagined back then. Perhaps if they did, would they still have done it? I don’t know. Chances are: yes. And then how the SSPs grew over time, you know, in relationship to their counterpart, which was the DSP. If we think back to, say, 2012, which was really like peak ad network and just prior to programmatic, like the private programmatic side of things, it was a waterfall world, there were priority and quality actually mattered because advertisers wouldn’t use certain ad networks based on what their reputation was, and which publishers were inside that ad network. So there was differentiation and distinction between them, not just in the way that they function, but also in the quality of inventory that they were delivering. As the SSP grew up, and private programmatic became a thing, I remember match rates being a really big deal back then, because it wasn’t about what inventory you had, it was about what advertisers you could attract. And that put SSPs in an interesting position because they were really beholden to aligning themselves with DSPs that were ready to spend money on them–which sounds good, except it comes sometimes at the expense of protecting their actual customer, which is the publisher. And to the sense that AdX really invested in itself–it did, and it made an incredible product–the other SSPs tried to invest but weren’t close enough to the ad server to actually get over certain hurdles that they needed to become an integral part of the stack. And DSPs invested heavily in themselves. And they were throwing hundreds and hundreds of engineers and data scientists at these problems and built incredible machine learning algorithms that turn DSPs into something like the T-1000 from Terminator. And yet publishers were a bunch of like villagers with sticks trying to protect themselves. It’s not where–I remember in 2014, when we went fully programmatic, was what we said–this was not the world we envisioned. But I can definitely see how we got here.

Somer Simpson:

Yep. What impact do you think? Because I mean, as you sit there, you describe, you talk about the ad server, and the SSP, and the DSP, and all the pieces and parts that connect the two sides and make the transactions possible or the auction possible, and I’m thinking, “Google has one of those…oh, yeah, Google’s got one of those, too. Oh, yep. And they’ve got one of those as well. Right.” What is the impact that Google specifically has had? And what do you think about the way that they played this?

Scott Messer:

I think, firstly, you have to give a big tip of the hat to Google. Because what they did, they did fantastically well. You know, they took a relatively fragmented industry and consolidated a lot of it. They provide monetization flows for another huge percentage of the internet, almost everything runs some sort of Google product. And they scaled their sales teams; they educated buyers; they delivered tons and tons of money through their pipes to publishers. And they, for the most part, did it in an increasing fashion, where it got better and better. However, during that time of consolidation and growth, they cut off certain roads of access and avenues for competitors to build competing products. Not to say that it was bad or illegal in any way, but it’s fairly well known that their strength and consolidation kept other ad networks and SSPs out of contention because of how intertwined AdX was with the ad server. And that was a big thing. And so publishers invented header bidding to get around that–the header is essentially a second ad server that runs an auction that Google has to listen to. And that was one of the two big things that publishers did to thwart either Google or DSPs.

Somer Simpson:

So that was one of the handful of sticks that publishers had to make sure they can continue to exist in the world of the terminators. Yeah, I’m talking about Google, and we talk about the duopoly a lot. I think that Google had one kind of impact to the industry. Facebook had a very different kind of impact, which I think is maybe a conversation for a different day, because that’s more about, like: where do people get their news, and how do they consume it, and what do algorithms do to that? And not so much about the ad space.

Scott Messer:

But let’s talk about Facebook in the ad space for a second, though. What they did for self-serve advertising, and to create a universe large enough, that’s your conversation for late–the universe they created. But the ad stack they put into Facebook, and now they have like 2 million customers on, and 10s of billions of revenue flow through it, and an entire industry of direct-to-consumer products were fueled by cheap Facebook advertising. Publishers were arguably sustained by it; Facebook certainly was. But they built their own playground, and they monetized it very, very well. And then, you know, somewhere there’s a philosophical question of: do they owe the publishers anything? Or do they owe the users anything, besides access to the platform for that service, and that’s another big one. But they were commendable in what they did, and I think they really changed what advertising means to more people; they democratized access to advertising in a way that had never been done before.

Somer Simpson:

That’s true, that’s true. Within their own window, right, within the walled garden. I think a lot of the privacy changes that we’re seeing today are coming out of GDPR that went into effect in 2018; before that, e-Privacy Directive; CCPA and CCPR and dozens of others that are coming our way in the States. But it started a long time ago, right? The rise of of Google, the rise of Facebook, and data leakage happening. That was the big Facebook scandal, right? The government starting to finally pay attention to the consumer advocates who were telling them, hey, bad things are happening here. How has all of this new privacy regulation and an attempt for governments to manage the duopoly, if they can’t deal with it with a different approach, they’re gonna go after it and apply the same rules to everybody when it comes to respecting consumer privacy? How do you think that that has had an impact on where we’ve gotten to and the opportunity maybe at hand right now?

Scott Messer:

Well, to say that it’s had the desired effect would be an overstatement. Regulating markets is hard enough and regulating the murky world of adtech is even harder. Data and its broadest definition, which is really what needs to be addressed, is ubiquitous; it is everywhere. And when we were going down the GDPR path, and when you and I became great friends, I really understood where data shows up. And that becomes important because of how hard it is to regulate–and then we’ll talk about consolidation of power from that–but when you think about data, and a company that has to manage information that they collect from users on an anonymous basis–what did they read, something like an email address because they signed up for a newsletter, maybe you made a purchase through the website–seems alright. But then, like, employee records are also part of this, right? You know, a job application is data. Signing up for insurance. Everything that you do on the internet creates a data exhaust, and that exhaustive data is valuable to someone. And so when you try to regulate something that broad, you will almost certainly not get a great result. And inside of that, we saw the largest players be able to fold up their doors, and unroll the drawbridges, and entrench themselves within their own walls. And everybody else on the outside was left uncovered and still dealing with the regulations. And a very broad patchwork. I thought the IAB did a commendable job in their foray into consent management platforms and the TCF… 

Somer Simpson:

The transparency and consent framework? 

Scott Messer:

You should know that very well, Somer. Didn’t you build that?

Somer Simpson:

Maybe!

Scott Messer:

I thought there are some very commendable parts of that. But you could even see, when it rolled out, that not everybody was on it–and actually it represented publishers fairly well; some nuances down the road sprung up that probably could be addressed–but marketers weren’t on it. 

Somer Simpson:

And still aren’t, right?

Scott Messer:

And still aren’t! And we have both sides of the house. I was just on a call today, and we’re like, “Hey, you know, we use this on the publisher side, it’s covered by the consent framework, and somebody has a similar technology.” And these are on the marketer side, but they’ve never heard of the TCF and so, therefore, our consent management platform doesn’t have a place for them. And now all of a sudden, you’ve got to build a custom thing. And the most expensive words you can hear in adtech are ‘custom thing.’ So adoption is really hard. And that makes it tough for publishers and especially small ones, and usually what they’re gonna have to do is: sub-scale publishers, you’re gonna have to quit the things that were probably profitable for you or advantageous, and you’re gonna have to go with some watered-down version of it. That’s really what you’re gonna have to do. Whereas the walled gardens get to replace you in that function and use their data as a tremendous value asset for either use within their own garden or, you know, across your lawn.

Somer Simpson:

Yeah, it’s creating an interesting divide, even between them, even within the publisher realm. The biggest publishers, it’s still a struggle, but they have the ability to sell direct only. There’s a couple that I know of who have pulled the programmatic pipes out entirely and are just selling direct. But others, like you said, who aren’t big enough–they don’t have the resources, they don’t have the technology, they don’t have the technical resources–just go deeper and deeper into the Google stack, right? And then they are at the whim of how Google decides to define privacy and data and data segmentation, and then all that kind of stuff. So it’s really starting to split the publisher world as well, which is ultimately going to have an impact on consumer access to content.

Scott Messer:

One thing I do appreciate is Google has taken–more so in the last few years, I would say–a relatively pro-publisher stance, especially when it comes to news and journalism, as well as just private media, overall. And I genuinely think that Google wants publishers to succeed, and they want media to exist, and they want journalism to be great. And we’re seeing more investment on our side to help in that. But it’s not the same as…I don’t know if it’s going to ever be enough to fight advertisers leaving the media space, because their media is untargetable–or worse, unmeasurable.

Somer Simpson:

Yeah. And that’s the critical piece right there–the measurement piece. And I think that’s something that people don’t think about when they talk about, you know, “Oh, cookies are going away. You know, tell us about your cookieless solution, and how it’s going to keep allowing us to do exactly what we do today, just without cookies.” And measurement is the big thing, and attribution. And if you can’t measure something, you can’t target with precision. And that’s the thing that people have, marketers have come to rely on so heavily is the precise nature of the technology that allows them to target today.

Scott Messer:

And that was, I think, when you go back to 2014, the year of mobile, one of the years of mobile, when Tapad and Drawbridge brought out their multi-touch attribution models and could really show cross-device how frequency and reach impacted spend and purchase behavior on desktop. It allowed marketers to feel more comfortable spending in mobile environments, even though purchase behavior was still happening on desktop. And that really helped publishers out. Had that not happened, we marketers would never have realized that their mobile campaigns actually do impact desktop purchase. And that was all based on cookies and IDs. And this infrastructure that was used for targeting was then flipped over and used for measurement. And there’s so many examples of great companies that measure it in different ways. But those, I think, were two really big ones. And then you look at the DMPs becoming marketing clouds, right? That was all about measurement and attribution. Because that’s what marketers needed. And so if we can’t solve for measurement in the ‘cookiepocalypse,’ I think marketers may just pull up their dollars and go back to television and other places where the reach is a lot cheaper and they didn’t have measurement in the first place.

Somer Simpson:

Right. That’s actually an interesting thing, right? It’s the measurement piece that drove the improvements in precision that have led us down the path we’re on to kind of where we are now and the advances that we’ve made. 

Scott Messer:

And those pieces of precision have been terrible for publishers. The cherry picking of inventory seems harmless and innocuous in some places, but it creates real disparity in your inventory. And so you know, you’re taking largely similar goods to the market. But they have highly differentiated prices, depending on the buyer. And a publisher doesn’t really know the difference between one or the other. Two users can have incredibly different value, even in the same highly similar ecosystems. It’s about what their abandoned shopping cart history looks like. 

Somer Simpson:

And what about contextual, too, right? 

Scott Messer:

Contextual is a big piece of it. I mean, two users in the same contextual environment should have the same value. But with cookies, you got more bid density. That’s what match rates are really about right now. It’s not the match between your DSP and your SSP. It’s about how much stuff does this user have that matches with how many buyers and that equals how many bids. And then over time, you can push up the CPM for that user, because the DSPs learn that this user is valuable, and they should bid a penny more than they did last time. DSPs are very good at finding floor prices; they are driven, like the T-1000 to find Edward Furlong, to drive the price down. And they will do it systematically and incredibly efficiently. They will wait for impressions, they will look for the floors, and then they will bid at them. Because there’s nothing that a publisher can do to really defend it. So the first thing they did was they introduced header bidding, which was to bring more demand into the stack; that had the secondary effect of creating bid duplication, which actually drove outsized returns for publishers–same inventory, but because of bid duplication, advertisers bid against themselves and drove the price up on themselves. So that was one very useful stick for publishers. And until DSPs can figure out supply path optimization, they will not be able to undo bid duplication, and publishers will continue to do it. I don’t think that they will be able to figure out supply path optimization on their own. But publishers can be persuaded to stop duplicating bids. The secret to better supply and cleaner paths lies with the publisher, not the advertiser. It’s not the DSP. The DSP has the stick. And the publishers have the carrots. So marketers should think long and hard about how do you work with publishers to reduce their duplication? The loss of cookies will make this difficult, but you only really need one SSP, two SSPs, at most. Seven is a lot. Yeah, there’s a little bit of unique demand. I can see the need for things like Amazon’s got unique demand; Google has some unique demand; a few sales teams here or there. And then you see The Trade Desk create OpenPath. That’s a very interesting move for them. They want the advantages of the SSP, without doing any of the SSP work–which sounds like a direct pipe into publisher inventory from which inventory can be extracted with incredible precision and lack of price.

Somer Simpson:

And no regulation.

Scott Messer:

No regulation, and no one to protect them. I see where things like OpenPath could be really good. And we could see more trade desks just have direct pipes and want to compete on price. But if there’s no tools where publishers can control deals and yield and build sort of their storefront inside of it, it’s really just bad news for publishers. But if you can do that, then you see things like what RTK.io was doing, and Rubicon acquired them, where they were effectively building In private SSPs, which are–we call them $0 floor deals in our world–for our top agencies and partners, at least, we create what we call our $0 floor deals and you essentially get an open auction pipe. But it’s inside of a private deal. It’s all our network, all the inventory; it’s very clean; it’s very cheap, because we use our best rev share partners; and it’s just us–there’s no fraud inside of it.

Somer Simpson:

It sounds like Christmas.

Scott Messer:

It does sound like Christmas. Here’s our phone number; you can sign up for more of it, whenever you like!

Somer Simpson:

Nice. It’s interesting to me, like… when we talk about the marketplace, it is like two sides of a coin, or it should be in an ideal state; in a perfect world, everything would just work, right? Marketers’ money would have good ROI, and publishers would get paid what they deserve for their inventory. But there’s all this silliness going on in the middle and thumbs on scales and various things. What do you think that we should be doing as participants in this ecosystem to do a better job of meeting in the middle? Because continuing to invest more and more and more and creating more Terminators, does it help to bring balance?

Scott Messer:

No, but I think that the loss of cookies will make marketers think twice about not working with publishers. It’s agencies in the middle who really control the spin; I would love to see more agency advocacy of building bridges with publishers and working closer with all of the tests that we’re doing. I mean, maybe it’s just my publisher chair that I sit in, but I don’t see a lot of news about testing from the agency side. They’re the ones who really stand the most to lose if the ‘cookiepocalypse’ takes a bad turn. But I want to see more collaboration; I want to see more pledges to buy inventory directly from publishers; I want to see more collaboration around contextual and a standard for it; seller-defined audiences, or something that’s really interesting, that I think if the buy-side gets around, the publishers can make a really good product for them. But whatever the outcome is, it’s not going to be like what yesterday was. So whatever you were doing yesterday, is not the solution for tomorrow. 

Somer Simpson:

You’re hanging out with your fellow publisher colleagues from other groups. What kind of advice would you be giving publishers today to be able to participate more effectively in this ecosystem, to be able to stay on the ground, find more tools, more sticks, so to speak?

Scott Messer:

Advice to fellow publishers–in what to do and finding sticks–is to take inventory of what you have now, and what you have in the future, and to engage in this conversation with us, and to participate in tests. Start to form your own opinions. And really think about what is best for all of publishing, and not just yourself. The slippery slope of DSPs was caused by weaknesses in the publishing ecosystem. The need for revenue made publishers take away their floor prices, and not watch it, and just open the floodgates. But we all know that it’s much easier to lower the price than it is to raise the price. So something about when you’re looking at first-party identifiers–your email addresses converted into hashed IDs or phone numbers or whatever it may be, take real good care of those. It is one of the last things that publishers have that marketers value. Put those in private auctions; do not open them and let them out into the open web, because they will buy your $5 user for $1.10 in Safari; they will then buy it for $1.09 tomorrow. Oh, and your chrome CPMs will decrease because they’re finding quality inventory in Safari, and you will just be lowering the boat. But if you can take that ID, and you can stick it into a market for $5, you’re getting fair value for it, because it’s probably worth much more than that to an advertiser, and you’re keeping your chrome CPMs on par. So really ask not just ‘What are the vendors doing, who are creating these IDs; where else are they going?’ but put pressure on your SSPs to create a better product that protects you, their client. If your SSP is not doing you any favors, you need to get a different one. And don’t take the easy money because in the long run, it costs you a lot.

Somer Simpson:

That brings to mind the scene in Braveheart, where they drew the line, and they’re standing at the line, and they’re yelling, ‘Hold!” I feel like, after almost 30 years in publishing or publishing-related industries, that is my biggest wish for publishers, that they do that: they draw that line; they hold it; and don’t let anybody pass. Because you’re right, this feels like: those users, that’s the last thing.

Scott Messer:

And, to that sense, I think there’s some really good publisher groups out there that are protecting thousands of publishers in mass that previously would have had to navigate these waters on their own, as they did in the 2009, ‘10, ‘11, like the ad network days. But there’s some really smart people representing them. And I want all the publishers to collectively hold hands and draw that line. And we will get what we need. And ultimately, that’s a fair price for our users; it’s high quality media on the internet; it’s a better user experience. People will actually be excited to go visit websites, and not afraid. If you look at ad blockers, publishers took a decent stance, not just against ad blockers, but against poor user experience. There are a lot of standards that were created and adhered to. And you don’t hear as much about ad blockers these days. Because most websites are okay to visit. And so when we do this, and that’s a good sign, we can accomplish what we put our minds to.

Somer Simpson:

100%. If there was one piece of tech, one Terminator that someone could build for you tomorrow for the supply side for publishers, what would it be? What problem would it solve for?

Scott Messer:

I’d be remiss if I didn’t say that it was the big CPM button. That’s the one we’ve always wanted: you get to push the button with the big CPMs. But that’s not realistic. What a great question. I would like to create a system where publishers can understand why an individual impression was purchased by a DSP.

Somer Simpson:

That’s cool: transparency all the way through, just as you’ve got transparency, in some cases, from the demand side, all the way through.

Scott Messer:

I know what signals we’re putting out in the bid request. I don’t know what signals are appended by the DSP, and I don’t know why a DSP chose that impression. Was it context? Was it audience? Was it ID? Was it random? That stuff would help me out. And we could better package and find your inventory for you, create more of what you wanted, less of what you didn’t. And that would be super helpful. But I’m not optimistic we’re gonna get one of those.

Somer Simpson:

Well, let’s see. Let’s see what I can do. Let me go talk to the engineers over here. Perfect. All right, so this has been a great conversation. Again, super appreciate you doing this, Scott. I’m sure I’m going to ask you again. Or just, you know, come out to visit and or you come out to San Francisco. But again, thanks for being on. And we’ll talk to you soon.

Scott Messer:

Thank you so much, Somer. Appreciate everyone at Quantcast and the mission that you’re on on behalf of publishers in a wonderful ecosystem. Thank you.

Somer Simpson:

Awesome, thank you. This podcast is brought to you by Quantcast. Our mission is to radically simplify advertising on the open internet. We are the creators of a new and innovative, intelligent audience platform that empowers brands, agencies, and publishers to know and grow their audiences online. The Quantcast Platform, powered by Ara®, our patented AI and machine learning engine, delivers automated insights, marketing performance, and results at scale to drive business growth responsibly. Our solutions are leveling the playing field for our customers when it comes to effectively reaching audiences online and helping them power a thriving free and open internet for everyone. Connect with us today at quantcast.com.