1. Podcasts
  2. The Digiday Podcast
  3. Can web users trust ad tech’s server-side shift?
The Digiday Podcast

10 Nov 2026 · 38m · 13 chapters

Episode artwork

Can web users trust ad tech’s server-side shift?

In short

The episode explains the “Trusted Server” specification and the broader “server-side shift” in ad tech, focusing on whether web users can trust server-side ad delivery and identity signals after Apple’s identity-resolution crackdowns.

Guest

Rowena Lam, Senior Director of Product at IEB Tech Lab; a leading spokesperson for Trusted Server.

Key claims

Trusted Server moves ad/identity-related calls from the browser to the publisher’s server, improving publisher control, potentially page speed/viewability, and respecting user privacy choices more “surgically” than client-side setups where browsers can block everything. It includes audit logging, attestation, and relies on an “edge cookie” set server-side. Publishers must decide how consent signals (e.g., GPC/ CMP) are applied and what gets shared downstream.

Notable examples

A food-hall analogy (browser calls to third-party “stations” vs server delivering a complete “meal”); mention of publisher adoption (e.g., Penske Media) and testing across Safari/Chrome/Brave/Firefox.

Written by AI. May contain mistakes.

Chapters

0:45

Understanding Trusted Server

Discover the significance of Trusted Server and its implications for ad tech.

“limit opportunities for ad tech companies, advertisers, publishers.”
2:36

Analogies to Explain Server-Side Shifts

Explore analogies comparing browsing experiences to food halls.

“food hall, but the kind of food hall where, and this I'm talking about like right now, when I go to a website, what happens and what that's like.”
4:42

Benefits of Moving to Server-Side Operations

Learn why publishers should adopt server-side techniques for better outcomes.

“time for publishers to start thinking about not, you know, continuing to keep everything within the browser itself.”
7:00

Concerns Over User Trust in Server-Side

Examine the user trust issues associated with server-side technology.

“And as I guess as you say, it's become more popular over the last couple of weeks.”
9:53

Audit and Transparency in Trusted Server

Understand how audit logging and transparency work in Trusted Server.

“Like there's nothing for you to do at that point.”
12:18

User Controls in a Server-Side Environment

Discuss the level of control users retain in a Trusted Server setup.

“And so I, yeah, I guess if you like don't actively decide to do that type of vlogging, it can be hairy.”
14:00

User Transparency and Control in Ad Tech

Learn about the level of transparency and control users have with Trusted Server in ad tech.

“or even just, I'm in California, may have global privacy control enabled, but it may not.”
17:20

The Role of Trusted Server in User Trust

Discover how Trusted Server impacts the trust relationship between users and publishers.

“And I would say, like, probably in terms of the feature set, thought about and considered.”
19:48

Understanding Edge Cookies and Privacy

Explore how edge cookies function and their implications for user privacy.

“There's a 64 character string that's set that's computed based on the device's IP address and a secret key that's specific to the publisher or the domain that I visit.”
22:54

Challenges of Server-Side Shifts for Users

Examine the potential gains and losses for users in the server-side shift of ad tech.

“Like it's kind of an issue to a certain extent, right?”
Show all 13 chapters
28:00

Understanding the Server-Side Shift in Ad Tech

Learn about the benefits and challenges of the server-side shift in ad tech, especially for users and publishers.

“And the transparency piece, like, again, I acknowledge that this is not trusted server.”
31:37

Exploring Trusted Server and Its Capabilities

Discover the specific functions and future developments of the Trusted Server in ad technology.

“So I definitely do think that that's like the big piece.”
32:44

Technical Insights and Cost-Effective Solutions

Gain insights into the implementation and cost considerations of trying out Trusted Server technology.

“And we're looking for market feedback on it.”
A man walking through an airport, asking his phone a question

Ask about this episode

Get the gist, explore an idea or ask about something said.

Claude

On The Digiday Podcast, what are the key points of this episode?

Searching your listening

Transcript

Automatic transcript. May contain errors.

0:09Hello and welcome to the Digiday podcast, a show for anyone wondering why is the hottest tech in ad tech right now the server? I'm Tim Peterson, Executive Editor of Video and Audio at DigitAid Media. And today I am joined by who I'm going to call the most popular person in ad tech at the moment, probably, Rowena Lam, who is the Senior Director of Product at IEB Tech Lab, and also the face of Trusted Server, which is the specification that the industry seems to really be rallying around at the moment after Apple started cracking down on identity resolution, and, you know, spurring a lot of concerns about like the power of the browser and how that can limit opportunities for ad tech companies, advertisers, publishers.

0:54Trusted server predates this crackdown by Apple, but seems to be central to what's becoming the server side shift that the industry appears to be looking to make. So Rowena, thanks so much for joining me. Thank you so much for having me. And it's so funny that you would say I'm the most popular person in tech right now, that's rarely the case because they usually focus on privacy related issues. And usually nobody wants to talk to me. So happy to be here. I am very happy to talk to you, especially because like, so I'm trying to better understand trusted server because again, it feels like right now the most consequential specification and at the moment, but you know, as folks adopt it for the near future.

1:39And I also just always love any sort of infrastructure stuff um i like to nerd out on these things and and i'm also as i've been researching trusted server more i even like put together a mock test with some websites and like back-end vendors implementing the trusted server spec to better understand how exactly it works testing it across safari and chrome and brave and firefox and with all you know gdpr tags, everything to try to understand, okay, what is this going to mean for me as just a person in the internet in the near future? But I think before we get to all that, we should probably break down like how exactly trusted server works or what's different about it.

2:23And I was, I think in analogies, have you ever been to a food hall? Yeah, of course. I'm a total foodie. So you're already kind of speaking my language already. Awesome. So I was thinking different kind of food hall, but the kind of food hall where, and this I'm talking about like right now, when I go to a website, what happens and what that's like. So I'm thinking of, it's like a food hall where you go to the host. When you first arrive, you place your order there and then they give you tickets and you have to go to the different stations to pick up your food. That feels to me similar to my browser goes to a website, that website gives me the HTML.

3:05But a lot of times, my browser then has to make calls to the different third party vendors who may need to like load things on a site. Those are the food stations that in the food hall I would have to go to. Does that check out? I know my analogies aren't always perfect, but is that captured enough? Yeah, yeah. I would say that's definitely an accurate depiction of that. Yep. Okay. So then it feels like the trusted server version of this would be, I go to the host, I place my order, they give me a number, I go sit down at my table and the human server brings all my food to me. Yep. They go out and they do the collecting of whatever it is at each individual stall on your behalf.

3:47And then they bring it all the full meal, the complete meal to you all at once. Which is pretty convenient for me. It's like I'm not having to go to the different food stations. That said, I'm having to trust the server that they're bringing me the food that I actually want and that I like the food that's coming. Why is trusted server necessary or something that companies should be moving to even before the Apple stuff that we'll get to? Yeah. So and as you mentioned, we started on this journey and started working on Trust a Server, which is open source software, not just kind of specifications, but actually open source software that publishers can can take and use.

4:31because with all of the signal loss just generally and signal loss conversations generally that was happening at that time when this initiative was first started, it was pretty clear that it was time for publishers to start thinking about not, you know, continuing to keep everything within the browser itself. So I would say that that's really where it, that initiated. There's, I mean, I'm sure we'll get into it. There's a lot of other features of trusted server and benefits for publishers and actually even to consumers to a certain extent that I would say is not necessarily a new concept. This concept of moving things over to the server, like things happen server side today, right?

5:22Yeah, connected TV server side is very popular. Yeah, that's a great example. So there is a lot that already happens server side. But I don't think that there has been a real push for when we think about open web to go that route. And so to me, it's like, okay, should we start actually thinking about it and doing things in a slightly more modern way? And there's a whole bunch of benefits to doing it beyond just, you know, like people try to think about this, like, oh, we're just mitigating perhaps like Apple's, you know, decisions and rest. And it's like, yeah, sure. That's one of the things. It's not all of it.

6:10That's a nice benefit now. Yeah, exactly. It's like, that's the point, right? Like, are we saying that now's the time, which I think it is. right yeah because i was at um ib's annual leadership meeting this past year and there was a session on trusted server where it was anthony katzer from ib tech lab and then i believe it was an exec from penske media talking about how like penske had adopted it was either penske or arena um but it was a publisher talking about how they had adopted trusted server and a lot of the benefits they were talking about was like improved page load speeds and then there are like ad blocking benefits or benefits for a publisher to not be as susceptible as ad blocking.

6:53Those seem to be the two primary benefits they called out at that time. Right. Yeah. And so definitely, those are what we talk about quite a bit. And as I guess as you say, it's become more popular over the last couple of weeks. And we've been having conversations with additional publishers who are also just trying to understand trusted server and what it does is some of them are interested in sort of the speed component of it because like that's what impacts viewability right improve viewability ultimately leads back to more money and so I don't know I've been saying trusted server is all about giving the publisher kind of control again of their stack.

7:40And I say like why they should care is because it will, the whole idea is that it will ultimately help publishers make more money on sort of their existing audiences. Like that's the reality here. We're not saying trusted server obviously solves for all problems that publishers are facing. There's a whole lot of them. And so it doesn't, you know, it's not going out and finding you additional users as a publisher. But the hope is that it will help you better monetize those existing users. Right. Yeah. Because one of the challenges in the client side version of all of this is a web page. I can go to a web page and that web page can give me its content, whatever article I may be there to read.

8:24But then the ads that it wants to show me, my browser would be making calls to different ad servers to fetch those. but because the browser is the one making those calls, the browser can also decide not to make those calls, which is how ad blocking can work. Whereas with the server side, the webpage is the one delivering me all of that. And so conceivably the browser could block, but then it's blocking everything, including the article that I'm there to read in the first place, right? Exactly. And this is exactly what you just described is also what makes going server side, I would consider just more privacy preserving in general and lets publishers actually better respect the consumer's privacy choices ultimately because when it's not the browser and you're not saying the publisher is not telling the browser like, hey, okay, now go and call this particular partner and the publisher themselves are controlling it, they can decide and be pretty surgical ultimately about, you know, who they would want to call which partner and sort of what data that partner can get.

9:41And so if a consumer, you know, makes certain privacy elections, you can be much more surgical about it versus what happens today as a publisher, the second that you, your instructions say, hey, web browser, go out and call this, like, you've lost control. Like there's nothing for you to do at that point. And there's a whole lot that you can do about what gets shared. When you are talking, it's the publisher's owned server that's making those calls. Yeah. So I want to spend a lot of time on this part because this is the part that I want to better understand because I get the point of now, if it's a publisher I trust, I maybe can trust them to be looking out for me, that they're only going to be bringing things back to me, to my browser, to my device that are reputable, that I would trust.

10:38But especially with changes Google's been making with Google Search and AI overviews and AI mode, I find that I'm clicking more links to domains that I'm not as familiar with. And so one of the things I'm wondering with trusted server is how much trust do I then as a user have to have in the website owner under trusted server? Whereas in the client side version of this, I can have tracking protections installed, whether they're browser based or browser extension based, but kind of do things on my side to try to protect myself. also see all the different calls that are being made from my browser when I'm going to a website, all of that is getting taken away from me.

11:25And so my concern becomes, what if I don't have a trust in the website that is kind of running the trusted server? Yeah, that's an interesting question. And I think that more broadly, sometimes there is a fear of just broadly like server side, because then it's not automatically observable to the consumer's browser and you can't automatically inspect it from that point. But what we've done for Trusted Server in particular is we've made sure that all of this is considered. So like what you can do server side is pretty robust audit logging. And so part of what's included in Trusted Server is, I would say like pretty robust logging and auditing that's made available.

12:15So anyone who is, you know, looking to deploy trusted server kind of gets that out of the box. So it's there and it's available. And so I, yeah, I guess if you like don't actively decide to do that type of vlogging, it can be hairy. But we've purposefully built that into Trusted Server. And I don't know how much I can talk about this at this point, because there's obviously this is a living, breathing software product. And so there's pieces to this, including attestation at multiple levels that we are thinking about. And the whole point of attestation is to address the trust piece on multiple levels, by the way.

13:02We're talking, you know, potential like trust between partners who would be operating in the supply chain, trust that because it is open source, you know, software, that this is kind of unmodified directly, you know, at the GitHub level, unmodified version of the open source project. So you kind of can understand that some of these guardrails that are naturally built into trusted server because we've considered it are still there. So it's things like that that are being built and worked on and iterated upon today. Okay. And so like those logs, is that purely at the domain side of things or how much of this gets exposed back to me as a user?

13:49Because granted, most people online aren't pulling up dev tools and looking at like the network requests that are coming, but they may have Privacy Badger installed or uBlock Origin Lite or things, or even just, I'm in California, may have global privacy control enabled, but it may not. Right now, I have these tools available to be able to see what are the calls that are being made, what data may be collected about me and how that data may be shared. And I may have some controls over to what extent those calls can or cannot be made. With Trusted Server, what, if any, transparency as well as controls do I have as a user?

14:38Yeah, so from a transparency standpoint, it is up to the publisher domain to make the specific details available to the consumer ultimately. but you know some of the extensions for example that you've mentioned or like gpc or whatever whatever other signaling and controls that the consumer has that is all i would say accounted for so meaning there we have things like cmp integrations in trusted server and as part of that, as you can imagine, that also includes ability to understand if, you know, GPC is configured. And ultimately, that does get respected. Does that only get respected in states where like there are laws in place that warrant GPC as a privacy opt-out?

15:35So we don't make that determination. And at Tech Lab, we typically don't, right? Like we don't make that determination at all. It's up to the publisher to decide. And also, I guess, like, have conversations with their partners about where or when they would decide to apply GPC. And I say that because, you know, some may decide that they actually want to be more restrictive, even if the law doesn't explicitly say that you have to, right, like, comply with GPC. And we hear about that. So we don't there, I guess I would say it's like, there's no defaults per se and trusted server on like the full details.

16:16Like we're not trying to make legal determinations on behalf of publishers. What we do require publishers to do is make decisions on what exactly and how they exactly they want to operate as it relates to respecting consumer privacy. And And that is the piece that's built in natively to trusted servers. So it's not a matter of like, yep, you're using trusted server and it's a free for all. It's like, no, you as the publisher, when you deploy, there are certain decisions and they're not necessarily net new ones. I'm just assuming at this point, right? If you're a publisher who has dealt with privacy regulations for a long time, you already know what your specific position is and how you handle it today.

17:03So I would see it as not different from a publisher decision standpoint today. Right. And I guess this is where it keeps coming back to that trusted part of trusted server feels like the most consequent or the most important element of it. And even, you know, when Trusted Server was first announced, it was just like, OK, how important is like the trusted part or like the trusted part felt more important when it came to what's happening on the back end among those companies. now it feels like the trusted part is just as important if not more important for in terms of the trust between a user and the website that for me as a person on the web to be able to trust that a website is using trusted server in an above board fashion for me you know what's funny is that i actually think of it as the like the opposite is like since the beginning yeah since the beginning of trusted server, respecting consumers' choices was always at the forefront.

18:11And I would say, like, probably in terms of the feature set, thought about and considered. Whereas the trust between partners was a thing that was acknowledged. And like, we need to have a solution for it because we need to ensure that the partners in the supply chain, like, feel comfortable and will trust this. And we do have a feature set that addresses that now. But I actually, I would say that like the way that it was considered as trusted server development has been happening has been in the opposite direction. And it's part of like we we worked on CMP integration. It was like one of the first things.

18:52And we considered that as part of like the like MVP of trusted server. It was included even as part of the POC. Like, because we knew we're like, without that, no, this is a no-go.

19:09With Trusted Server, there's the edge cookie, which correct me if I'm wrong, but is a first party cookie that the publisher, the website domain drops. And that seems to be like the linchpin for so much of this in terms, at least in terms of like how identity resolution can work through trusted server given what Apple's been doing. Is that right? Yep, that's right. It is set, an edge cookie is set on the server. So not, it's not a cookie that's like set via JavaScript. It's set server side. And it is the, I would say like the linchpin. Yeah. And so the edge cookie, there are kind of two primary components to it.

19:52There's a 64 character string that's set that's computed based on the device's IP address and a secret key that's specific to the publisher or the domain that I visit. And then there's like another string that gets added to that to make it a unique tag. Is that right? Yes, you got that absolutely right. Yeah. So I was messing around with Trusted Server, you know, pulled it, created like a mock publisher site and then a mock vendor site. And one thing that I noticed was when visited the website, ran through trusted server, all that, then I cleared cookies in my browser. This was in Safari. And then I went back to that website.

20:39A new edge cookie was created, but because the 64 characters of that edge cookie were created based on the, my IP address, which hadn't changed and the same publisher domain that I visited, that was able to, I was able to like reassociate the two edge cookies because they had the same ingredients there. Is that okay? Is that, are you asking me for, I have to say, I guess I'm not at all. Yeah, because like I cleared the cookies and then went back to the website, But then because of how the edge cookie is configured, it gives, whether it's a website or a vendor, whoever gets those two different edge cookies would be able to reconcile them despite me having cleared the cookies.

21:35cookies and so I get that like that's not the same as me opting out of tracking but it feels like the kind of thing where if I'm clearing the cookies in my browser a reason I would be doing that is because I don't want to be tracked or I don't want to like be associated with my previous browsing history yeah I guess what I would say is that so there's the edge cookie portion of this for trusted server for the actual identity, that doesn't mean that that edge cookie itself automatically makes it to any of the downstream vendors that a publisher engages with. And actually, and this does kind of go back to my point of like, this is where a publisher does have more control.

22:22It's like you would have to like proactively decide that this edge cookie gets sent to anyone. And so nobody downstream would automatically receive an edge cookie. It's a proactive choice. Right. And so that comes back to like where the trust is. And it feels like so much of what we're getting at here is it feels like it's going to become even more important for websites to establish that trust with users off the bat and to like reinforce that trust over time of if the website's going to be doing all of these things server side on the back end that's effectively like a black box to me as a user i need to be able to trust that the website is doing right by me if i'm passing a gpc opt-out that it's honoring that and it's not passing along the id tag for that um is that generally the idea and is there anything you would recommend for publishers or that's being somewhat you know incorporated into trusted server for that trust component to be as much of a priority as like any of the technical adoption that would be needed yeah i i guess like the way i think about it is that But this doesn't, like the trust issue is the same in some instances in my head, like meaning the consumer trust of the publisher's site is sort of the same.

24:01Like it's kind of an issue to a certain extent, right? Like if you don't, I guess like if you're a publisher who doesn't have a set, like you don't have trust regardless of whether something's happening server side or on the client side. To your point, if I think about an average consumer, like most don't go into developer tools, like don't do all these things, probably largely because they don't understand it. Right. So I don't think like this conversation of consumer trust is necessarily different when we are talking about going server side. And I would say that, you know, we're in a world today also now that there are privacy laws, regulation globally that actually mandate that publishers respect consumers choices about how their data is used.

24:58And so the obligation of the publisher to do that and respect the consumer's choices has been there for quite some time. And so to me, it's more of like you have to continue to do that and do that correctly. I feel like when we're talking client side, there's a lot more oopsies that can happen. Whereas server side, like, you know, as with anything, like it's if you configure it incorrectly, like there could be an issue. But there's at least still a little bit more control. So I still do see it as, yes, consumer trust. But that's not different from today. Like you have to establish that consumer trust today.

25:41And if you do an OOPC. I guess today it's the idea that like if I have the browser, the browser can be my defense as a user or at least like my representative in all of this. Whereas it feels like in a server side world, I would be losing that. And I'm kind of on my own in the wild west of the World Wide Web.

26:07So it's so funny that you say that because I think that's like a weird thing to expect. Or the consumer has a relationship with the publisher, presumably, if you're going there. That's the thing, though. Like with, you know, Google's made such a big thing of, oh, with AI overviews and AI mode, we're actually surfacing all these new websites. All these long tail websites are getting new referral traffic. And I'm even saying in my own browsing that like I'm getting directed to websites where I don't have that relationship. And that you don't have an existing relationship. I see what you're saying there.

26:46And I can understand that point. But I guess I just question and I don't know if it's like totally the right answer to say, like, as a consumer, then the browser is going to be my shield against. I think it's more just like I don't know. I don't know that I have any other shield beyond that. And some of this is like Apple's done a really good job with marketing around privacy and kind of positioning Safari as that shield. Brave and Firefox obviously do a lot of that. and it doesn't feel like there's been enough marketing from the industry of like oh yeah no we're looking out for you as well and not just like looking out for you but actually giving me as a person who would want these things options or just like the means of transparency and control and so the reason why like i'm going on about this is it feels like trusted server is such a foundational technology and like will be such a fundamental shift if the industry does fully make this server side shift and trusted server is at the center of that and so it feels like there's an opportunity right now for trusted server to enable a better version of the web not just for website owners and ad tech vendors but for users as well as you're yeah because as you're saying like the client side version is far from perfect but i haven't been as clear on like what is it that i'm gaining as a user and as well as like what is it that i'm losing as a user in this server side shift yeah i mean i ultimately i think gaining is what you are gaining is just that's hard so this is like what publishers are getting is just more control but it is like an extension of the user as well, meaning that whatever your choices that you're making, just publishers are going to be more capable of actually respecting them.

28:50And the transparency piece, like, again, I acknowledge that this is not trusted server. This is just generally when we're talking about server side, period, right? That transparency has to be something that you include and think about at the out, right? or from the beginning, I should say, from the beginning. And I think we've done that with Trusted Server. That sort of is the point. We're not trying to, its purpose is not for it to be a black box. And then I just go back to regulation again, because, you know, there's, we are at a point today where there's lots of regulation to protect consumers.

29:30And like one of the cornerstone pieces of regulations, again, I'm not a lawyer, I'm not going to interpret it. But like, what I understand is that transparency to the consumer is kind of step one, actually. Right. And so that's that's a requirement. And we acknowledge that. And so when we think about trusted server, of course, like there has to be the capability for publishers to then surface that level of transparency to the consumer. And on that note, what are specific capabilities or mechanisms in trusted server that you'd want to call out that publishers can be using to surface that visibility to the end user?

30:12I guess there's not specific consumer-facing capabilities to surface it to the consumer. So, like, the publisher would still need to figure out, like, right, like, the exact mechanism to present it to the end user. And they do that today, right? Like whether it's privacy policies or, I don't know, some other mechanism to let the user know how they're collecting, using their data, et cetera. But what trusted server does is, I guess, give you the, I would say, like the primitives to do that. So all of the like audit, logging, et cetera, it's sort of the primitives so that a publisher can then take that and present that back to the user.

31:00It also kind of does the reverse, which is like reverse meaning like if a publisher shows a consumer and gives them certain rights, trusted server can then effectuate those rights downstream. Right. Yeah. And then again, the onus is on the publisher to be passing along whatever the consent signal or whatever else it is to the other. So it all comes back to who's operating the trusted server and how are they earning that trust. right this has been a really helpful conversation again as someone who is looking at trusted server as the biggest thing in edtech at the moment the most consequential thing anything that folks should be keeping an eye on in terms of the development of trusted server what's next on the roadmap here so i mentioned attestation um and and just like furthering the trust and trusted server to your point for on multiple levels.

32:06So I definitely do think that that's like the big piece. We've got a pretty robust roadmap that hopefully we'll be talking more publicly about in the coming weeks. But yeah, it's ready for testing. You've clearly played with it. And I would say that's exactly what folks should do. Hopefully you found that the documentation was fairly simple to follow for you to kind of get something up and running on your machine to just see how it works. And that's what I would encourage everyone to do is go ahead. It's open source. It's open source code. So like go ahead and try it out. And we're looking for market feedback on it.

32:49There's a lot of Claude and Claude Code Fable 5.1 sessions implementing. Okay, explain to me how this works. weight. Basically, it felt like technical reporting of, okay, let's implement this now. Let me kind of pick at why does this work? Why was this the result? How did this, is this actually a signal? And then also just using all of that to then inform this conversation so that I could come here and be able to have a hopefully somewhat informed conversation with you. I love that. The team, the engineering team who works on Trusted Server likes to describe it as If you are like just mildly technical, like you don't have to be an engineer, mildly technical, which is what I consider myself as a product person and not an engineer.

33:37You can you you actually can get it up and running and just see what it does and test it. Yeah. Yeah. It was like, I think,$20,$10 a piece for two Cloudflare domains and then 14 day free trial of parallels to run a virtual machine on my desktop. and then, I mean, all the browsers are free and then it's, you know, a cloud subscription on top of that. So yeah, pretty cost-effective way to check out this technology. Yeah. Awesome. Well, Rowena, I really enjoyed this conversation. I appreciate you so much for coming on and helping me to better understand Trusted Server. Thank you so much. Happy to be here.

34:24Thanks for listening to this episode of the Digiday podcast. If you enjoyed it, please leave us a rating and a review on Apple Podcasts, Spotify, or wherever you're listening. Get more from Digiday with our daily newsletter sent out each weekday morning. Visit digiday.com slash newsletters to sign up.

From the publisher

“Trust” is one of those words that gets thrown around a lot in ad tech circles. But it’s going to become even more important if ad tech completes its server-side shift.In response to Apple’s identity resolution crackdown, the online ad industry has rallied around server-side ways to sidestep browser-based controls. IAB Tech Lab’s Trusted Server is the biggest example of this. Not only does the open-source software promise publishers faster page-load speeds, it enables publishers to better control their ad tech stacks and minimize their exposure to browser-based blockades.Trusted Server originated at a time when third-party cookies were under siege and signal loss was a major concern.“When this initiative first started, it was pretty clear that it was time for publishers to start thinking about not continuing to keep everything within the browser itself,” said Rowena Lam, senior director of product, privacy and data at IAB Tech Lab, on the latest Digiday Podcast episode.Taking ad calls out of the browser, however, has the effect of taking them out of users’ sights. A browser can see – and block – the calls a web page asks it to make. Once the calls run on a publisher’s server, though, users have to trust the publisher to a greater degree.For example, Trusted Server introduces the edge cookie, which is a server-side identifier that websites use for users. The edge cookie is created, in part, by hashing a user’s IP address via a secret key specific to the domain the user is visiting.The downside of this from a user privacy control perspective is the edge cookie can be used as a persistent identifier. A person can clear their browsers’ cookies, but if they visit that website again while connected to the same IP address and the website uses the same secret key, a new edge cookie would be created that can be matched to the previous edge cookie, effectively undoing the user’s cookie-clearing. And if the website passes on that edge cookie via server-side calls to ad tech firms and others, those companies would be similarly positioned to use the edge cookie effectively as a persistent identifier connected to whatever other identifiers they have in place, all beyond the user’s visibility.And this is where the trust in Trusted Server really matters.“That doesn’t mean that the edge cookie itself automatically makes it to any of the downstream vendors the publisher engages with. And actually this is where the publisher does have more control. [The publisher] would have to proactively decide that this edge cookie gets sent to anyone. Nobody downstream would automatically receive an edge cookie,” Lam said.Lam also noted that Trusted Server was designed from the start to support integration with consent management platforms. That would give websites a way to share a user’s consent status across the programmatic supply chain. But of course, that requires the user blindly trusting the site to do so.