Latest / Signal & Noise: Bringing Clarity to Media, Data, Tech and AI / The Activation Gap: Rob McLaughlin on Why First-Party Data Still Isn’t Moving the Needle
Transcript
- 0:03This is signal and noise, bringing clarity to data, tech
- 0:07and AI. Welcome back to Signal and Noise
- 0:15everybody. I'm Brett House.
- 0:16I've got my Co host Rio Longacre.
- 0:19And today, we're talking about one of the most overused, trite
- 0:24phrases in modern marketing and digital advertising.
- 0:27There are many of them. First party data, maybe even
- 0:30zero party data if we want to go down that path, but really from
- 0:34a very different angle. We've all gotten to the point,
- 0:36you know, we've all understood the point over the last 1520
- 0:39years in digital advertising for those of us that that came from
- 0:41this space, that first party data has been sort of, it's
- 0:46considered to be critically important to brand success in,
- 0:49in the advertising ecosystem. And and the real question I
- 0:52think is if it matters so much, why is so little that actually
- 0:56making its way into media activation?
- 0:58Shockingly little, yeah. Shockingly little, according to
- 1:01the thesis of Rob McLaughlin. Is it I Did I pronounce it
- 1:04right? Or is it?
- 1:04McLaughlin, OK McLaughlin with an F.
- 1:08Yeah, which is one of the thesis of your new companies or I guess
- 1:13not new you you founded it in 2022, right Audiences, which for
- 1:17those that don't know is a cloud native first party data
- 1:19activation application. And and your big thesis is that
- 1:23the problem isn't that brands don't have first party data,
- 1:25it's that they're really not there's some something blocking
- 1:28their ability to actually activate that.
- 1:30And we'll talk about that dissect this in this episode.
- 1:33But your background, Rob, great to meet you.
- 1:37You know, welcome to the pod. I know you've spent years
- 1:40working at the intersection of data marketing activation at Sky
- 1:44Barclays. And then you had a, an exit as
- 1:48well with a loop. Is that the, is that the name of
- 1:51the company loop in 2023? So that was a consulting and
- 1:53tech company and you exited that.
- 1:55And then, but it seemed like there was a simultaneous play to
- 1:58found audiences right, right around 2022.
- 2:02And, and tell us what you're doing there and tell the
- 2:03audience a little bit about yourself.
- 2:05Oh, thanks, Brett and Rio. Brett, great to be here with
- 2:08you. So thank you.
- 2:09Yeah. I mean, most formative bit of my
- 2:12background was as head of data at Sky.
- 2:16So during the Comcast acquisition.
- 2:19So we were the biggest media company in Europe of its type at
- 2:23that time. And then, yeah, becoming a
- 2:25global company with, with this sort of being gobbled up by the
- 2:28North American giant. And it was an amazing time.
- 2:32I I was head of data across the buy and the sell side of media,
- 2:36which meant we had this incredible.
- 2:38Yeah, incredible resource which could be could be purposed for
- 2:42many things. That's common for a lot of
- 2:45organizations, but kind of like the timing was pretty special.
- 2:49So that was around 20/16/2017 and Sky was doing something
- 2:54crazy then, but something which is very normal now, which was
- 2:59migrating all of its data onto cloud.
- 3:02So for them, it was taking everything from its own premise
- 3:06data centers into Google Cloud and as I say, very normal now.
- 3:12Very, very, very weird. And you're right, back then,
- 3:16that's where everyone was just starting.
- 3:17Like the people don't realize like the whole cloud, the
- 3:20transition of all these like these applications, data sets
- 3:23that didn't really start until like 2012 to 2014 very earliest.
- 3:27And I don't think it even accelerated until a couple years
- 3:29after that. You're absolutely right.
- 3:31And I don't make a secret of the fact that when, you know, when I
- 3:38took the, the job at Sky, no one told me that this was the plan.
- 3:42And I was very happy to be living in a world of, you know,
- 3:47enterprise data warehouse on premise, like EDW type language,
- 3:53leaning into the the likes of the like classical sass, like
- 3:57Adobe and that kind of stuff. And and then this happened.
- 4:01I was like, whoa, I don't even know what that is.
- 4:04Got to Google what that is. And and and it ended up very
- 4:10quickly becoming the most influential sort of force on my
- 4:14career. So, so we yeah, the the question
- 4:17was forced you to get. Technical and to learn this new
- 4:20domain, right? Everyone was learning, right?
- 4:22It's not like there was a big bench of with experts with tape
- 4:26expertise at that point. It was a new thing.
- 4:28No, no. And, and even more than I would
- 4:30say because they're both worlds, the three worlds existed on
- 4:35premise, data management, cloud data management.
- 4:39And then like this grey area in between of people who built up
- 4:43martech ad tech stacks from third party tech.
- 4:47And yeah, I, I probably sat in that in that middle of like a
- 4:50very happy buyer of lots of tech, probably also with a yeah,
- 4:53I came from the old school web analytics, digital analytics
- 4:56background. So, you know, probably, yeah,
- 4:58some chops, but but yeah. We had Adam Greco on the show a
- 5:04little while back. Long time, long time, long time
- 5:07Ally, long time fan. Love what they've been over
- 5:10there as well. I mean, yeah, very relevant.
- 5:13You know, we got, well, the team and I got a front row seat to
- 5:17how to use cloud compute and storage inside of a gigantic
- 5:22enterprise to support the buy and sell side of the media.
- 5:26And yeah, we're very lucky. So we, we built all this stuff
- 5:31on GCP and accelerated the business.
- 5:34So you know, what we built there was yeah, in in the UK and
- 5:38Europe was immediately adopted in, in North America.
- 5:41And we found we were in our thinking and execution 10 or 15
- 5:46years ahead of a lot of companies because, because they
- 5:49all had started their cloud transformation, just as we said,
- 5:52you know, it became well, the obvious and Deroga thing to do.
- 5:56And it's funny you mentioned I, I was like, OK, cool, well, we
- 6:00should go and help other companies make use of AWS, Azure
- 6:03and GCP in the emerging snowflake data bricks, big query
- 6:07side of the world as well to, to feed advertising on the buy and
- 6:10sell side. That feels like something which
- 6:12people want to do and need to do quickly.
- 6:14So, so we left and started Loop Horizon, which was that
- 6:18consulting business very relevant to our audience's story
- 6:22because like, you know, 10 or 15 projects in making good money,
- 6:28you know, people would pay us to tell them how to use AWS and
- 6:32Snowflake to pump data into, you know, Google ads, Meta, the
- 6:37trade desk, etcetera. I was like, now you make it new
- 6:40every time and get paid a few 100,000 lbs dollars to do it.
- 6:44But like, maybe because we're recommending the same framework,
- 6:49maybe this could be great training wheels for a software.
- 6:54And yeah, I'm guilty. I can only do one thing at a
- 6:56time. So once, once that I once
- 6:59that's. The ADHD like I have right?
- 7:01You got a hyper, you got a hyper focus.
- 7:03It's something like that, something like that, something
- 7:06like that. Metalize hyper focus and and and
- 7:08pivot, right? I mean, that's.
- 7:10I'm diagnosed. There you go.
- 7:11Do it. But yeah, once the idea got in
- 7:14my head, it was, it was. I knew that's what I wanted to
- 7:18do. I knew that I wanted to start a
- 7:21pureplay software company and and that's what audiences is.
- 7:25Awesome. Well, well, that's exactly why
- 7:27Rob, we wanted to have you on here to have this conversation
- 7:29because as we all, this is an industry that for the last few
- 7:31years we've been talking about first party data, first party
- 7:34data. I mean, I've probably been on 15
- 7:37panels. We've talked about the value of
- 7:38first party data and 1st party signals and blah, blah, blah,
- 7:41right. I mean, it's been a big thing,
- 7:43but I mean everyone talks about first party data like it is the
- 7:46strategy headline. It's it's not, but it's
- 7:49incredibly valuable and it can actually change your outcomes if
- 7:52you use it correctly. I posted this recent case study
- 7:55in which this is first CPG company, a seed audience that we
- 7:58use the first party that outperformed the platform native
- 8:00targeting by wide margin in terms of new to brand customers,
- 8:04in terms of average order volume.
- 8:06The things the brand really wanted to focus on in order to
- 8:10grow their grow their business as opposed to business as usual.
- 8:14So to me that's the whole point. It's a reminder that when you're
- 8:17using it correctly, first party data signals can be very
- 8:21valuable more so than platform signals.
- 8:25And but if this is true, like why aren't more brands using it?
- 8:31I just been a big push, but adoption still pretty low.
- 8:33I remember even as recently as a couple years ago, looking at, I
- 8:36think as one of the media cleaning rooms, only a single
- 8:39digit percentage of brand advertisers were actually using
- 8:43first party data to for targeting or measurement, right,
- 8:47which are the most bait or even basic use cases like
- 8:49suppression. So it's it's, it's kind of
- 8:52shocking. So like it's been cool to see
- 8:54what you're building. We'd love to hear about that and
- 8:56like to understand. OK, Like, like audiences, your
- 8:59thesis seems to be the model that sells how I have to change,
- 9:01right? Keep that in your brand's cloud,
- 9:04move signals instead of data sets and reduce the privacy and
- 9:09operational friction. And many brands are experiencing
- 9:11that has really slowed this whole thing down.
- 9:13So we'd love to unpack what you're doing, why it's
- 9:15important, what's working, what's not, and like where do
- 9:18you see things going? So with that, Brett, maybe first
- 9:20question is over to you. Yeah.
- 9:22You know, you've worked at what we talked about it, you, you,
- 9:23you worked in a bunch of major brands and pretty complex data
- 9:26environments. You learned early technically
- 9:28how to do this sort of stuff. So what what convinced you that
- 9:31like now was the time to develop this software and and what is is
- 9:35getting in the way of brands activating their first party
- 9:40data? It seems like everybody you
- 9:41know, it's almost common knowledge that everybody's.
- 9:44Right at this point. Yeah, to some degree.
- 9:47But you're claiming that that's actually not the case.
- 9:49I mean, what? What's, what's breaking in this
- 9:52process? Well, you guys then and really
- 9:55you did a great job there representing, you know, it's, it
- 9:58is, it is, it is common understanding now that using
- 10:03first party data in advertising is powerful.
- 10:06And you know, at audiences we we don't feel the need to tell
- 10:09people that very often. Yeah, usually, usually they want
- 10:12to do it. And yeah, yeah, yeah, you want
- 10:16to do it whether it's for the buy side or the sell side.
- 10:18But you don't. You don't need to convince
- 10:19brands it's important anymore. I think that are we past that or
- 10:22are they still arguing with you? Why should we do that?
- 10:25No, we're, we're past that, we're past that.
- 10:27And so, so, so there's a strong desire whether an advertiser,
- 10:32you know, to be able to use it for acquisition growth, you
- 10:36know, cross sell, upsell exclusion, etcetera, or feeding
- 10:39algorithms. And on the sell side massively,
- 10:43right. If you're a publisher or a data
- 10:45Rainer, being able to get that your, your, your signals to
- 10:49supply super, super, super powerful.
- 10:52Originally, yeah. Originally our hypothesis was
- 10:58that it was a deadly cocktail of data privacy, data security, and
- 11:04the difficulty of doing anything regarding data and technology.
- 11:08Yeah, well. Especially with GDPR in the in
- 11:11in the UK and Europe in. Germany, yeah.
- 11:14GDPR, DCPA, all similar. Alphabet soup of privacy laws
- 11:18and wherever every country has them at this point.
- 11:20So yeah. It really is.
- 11:21And then, you know, in big organizations, once you might,
- 11:25you might be successful with talking to a DPA, DPO, sorry,
- 11:31then you're gonna have to talk to infosec.
- 11:33So we spend a lot of time talking about privacy in ad tech
- 11:37and advertising. But if you're interfacing with a
- 11:41large enterprise, you have to do that.
- 11:43But then there's information security, OK, that where it's.
- 11:48It's it's and maybe legal too gets involved, right?
- 11:51I've had that. In the end, these guys often sit
- 11:54within that part of the organization.
- 11:56So, so yeah, absolutely. And you know, there and, you
- 11:59know, there are at audiences that the the guys in privacy and
- 12:04information security are often our allies.
- 12:06And that's great. But they've had decades of
- 12:09training in saying no. So they they say no every day to
- 12:13lie. That's.
- 12:14What they say by default, right? They they say no and then OK,
- 12:17but here's what you need to do to the boxes you need to check,
- 12:20right? Yeah.
- 12:20And is there a lot of education about the technical elements of
- 12:23this for? Us for us there is there,
- 12:26there's loads. They become our allies.
- 12:28I'll explain why. But yeah, what we we thought is
- 12:32against privacy and data security, OK, there needs to be
- 12:36a solution which which doesn't require the enterprise to copy
- 12:39their data and give it to someone else.
- 12:41And what we saw in the background, and we're lucky
- 12:44enough to have that front row C to the transformation, was
- 12:50enterprises had quietly built up this incredible capability using
- 12:57cloud infrastructure like AWS, Azure and B using data
- 13:01warehousing like Snowflake data bricks and Big Query.
- 13:04They had built in the background the capability to to store,
- 13:10manage and process this data themselves.
- 13:13Fundamentally, we looked at SAS and you know, thinking of the
- 13:16live ramps, thinking of the whole CDP space, and we said,
- 13:21yeah, they've got some queue apps, but most of that is
- 13:24reselling storage and compute. So what about if we just did the
- 13:28queue app and that's what audiences is by the way.
- 13:31And in that journey, as we found out, you know, as we.
- 13:34Doing the what? Did you see that?
- 13:35Again, yeah, the application, creating the application and
- 13:38then we give the application to run in the enterprises cloud
- 13:43owned data. Yeah, native environment in
- 13:46their AWS, Azure, GCP or more recently really exciting in
- 13:51their Snowflake, their data bricks and.
- 13:56Does this eliminate the need for data onboarding like with live
- 13:59ramp that was 1? Of our yeah, yeah.
- 14:01That was one of our thesis at New Star.
- 14:03We were. We were.
- 14:03It would. Be a pretty big cost too, right?
- 14:05I mean, it's not not inexpensive.
- 14:07Yeah, no, exactly. It's a huge cost.
- 14:09It's a huge task. And in our whole thesis at New
- 14:11Star, we were sort of competing on at least the identity
- 14:13resolution and onboarding side of things in certain places with
- 14:17with live ramp was if you build this native within the
- 14:20environments like the snowflakes was, was our kind of our first
- 14:23prototype. There is no need for data
- 14:25onboarding because you're doing the data matching and
- 14:28synchronization, unification, all that stuff within that that
- 14:31environment which is protected by the brand.
- 14:34But it's funny, I've probably had a dozen companies over the
- 14:37years tell me, oh, we have a live ramp takeout strategy.
- 14:39We're going to eliminate onboarding with some native
- 14:41identity solution. I've had CDP say it, I've had
- 14:44different. So, so I have heard that quite a
- 14:46bit. But I guess, Rob, the question I
- 14:47really had here though, is I know, I remember around
- 14:512018-2019 I the concept of data clean rooms.
- 14:56I've, I, you know, I got really excited about it for very
- 14:58similar reasons about branch should be using first party data
- 15:01and the ones I knew and I found this was wild.
- 15:03I, I would talk to Bryce. Do you hear the same things you
- 15:04just described? Oh, well, legal says you can't
- 15:07do it for whatever reasons, But then finding out like some like
- 15:11the same brands who were saying they protect their first party
- 15:13data, they adhere to privacy laws, are uploading audiences
- 15:17directly to DSPS, right? Just bypassing this, just
- 15:20pushing it up. Oh, we're just sending to DV
- 15:22360. I remember it being shocked and
- 15:24realizing they're doing this because no one in the
- 15:26organization really knew, right. And, and then the DSP will say,
- 15:30oh, you know, we only do the match.
- 15:32We don't, we, we don't, we, we, we delete it afterwards.
- 15:34We're only keeping a match data and it's encrypted, but I don't
- 15:38know, I mean, I wouldn't give, I wouldn't be given Google or the
- 15:41trade desk or any DSP is my first party directly.
- 15:43If I was a brain, I would do it zero copy.
- 15:45But I mean, is that what you've experienced?
- 15:47And so I think that's where things were a few years ago.
- 15:49Then the data cleaning was kind of struggle to take off.
- 15:51So I'd like love to maybe hear from you.
- 15:53Why do you think that was? And like like what, what forced,
- 15:56like what, what caused that? I think they're taking off a
- 15:58little bit now, but like, where are things now?
- 16:00They're taking off a little bit now.
- 16:01They've all been purchased, right?
- 16:03Or they or they're. I mean, I mean, they've all been
- 16:06purchased before they hit critical mass and adoption.
- 16:11Habu. You know, and bless them, you
- 16:13know, You know, bless them. Those businesses didn't deliver
- 16:17on the adoption which they promised.
- 16:19Like, these are not big businesses, OK?
- 16:21They're full of great people. I get why you might want to buy
- 16:24them and push them all into WPP. So you've got some people who
- 16:27actually know something about data, and that's useful.
- 16:30Yeah. But that technology has
- 16:33unfortunately failed to be adopted, failed to take off.
- 16:37Yeah, our view and it's funny you said new start and then live
- 16:40ramp, but all of this very, very.
- 16:42So we had a fork in the road early doors where there was
- 16:45people saying, hey, one of the biggest problems is on boarding.
- 16:48Hey, could you help us on board data into live ramp into, you
- 16:52know, the DNP of choice, whatever it was into ACDP,
- 16:55whatever. And we're like really that the
- 16:58value add if all we were was a data on border into somebody
- 17:02else's platform. That's that's not our mission.
- 17:04Our mission is to get data and then signal into advertising,
- 17:09we'd be stopping short. So what we do on board into
- 17:12those platforms. But what we what we thought was
- 17:14what we found very quickly was yes, privacy, yes, security,
- 17:18yes, like the difficulty of it. But cost of doing this is was
- 17:27escalating more and more and people were doing very, very
- 17:31reductive things. So they were choosing what small
- 17:35selection of data they might onboard because the cost of
- 17:37onboarding data was so expensive.
- 17:39It's like wow OK so I all choosing.
- 17:41So they have this incredible data set, but they're saying,
- 17:43oh, because it's so expensive, we're only going to take a small
- 17:47amount instead of saying we'll put put a lot of it just to
- 17:50unlock maybe things we don't know the note, no, the unknown
- 17:52unknowns or whatever. Right.
- 17:54And in this era of of models and algorithms, this is that humans
- 18:00shouldn't be shouldn't be curating and trimming the data
- 18:03which is then going to be used by the models.
- 18:05We all know that that's that was clearly that can't be the
- 18:08future. And and and these escalating
- 18:11costs of like I want to push more data that that's how and
- 18:15yeah, a sorry reality in advertising.
- 18:18It's and wait for a moment for me to finish this point.
- 18:21Advertising is one of the most unrewarding places to push data
- 18:27to. If you've got data and you want
- 18:29to get value from it, you, you you use that data in direct
- 18:34channels like e-mail and SMS to begin with.
- 18:36Believe me, you get response. Advertising, you don't get
- 18:40response. And the models of the data on
- 18:44borders, the DNPS, the live ramps, the CDPS, they charge
- 18:48very, very similar rates for onboarding into, into
- 18:52advertising as they do into those direct channels.
- 18:55And we just said there's got to be a way to change this.
- 18:58And what we found is in the solution where we run our app
- 19:02natively inside of the infrastructure of the client.
- 19:05But you know, it's, it overcomes the privacy and the security.
- 19:09It also just leapfrogged the cost because all of the storage
- 19:13and compute is already happening inside of the client's
- 19:17infrastructure. There's no cost to pass on.
- 19:20Yep. So we've got B to B clients that
- 19:22push 10s of 1000 records a day because that works for them.
- 19:26And we've got retailers who push 90 to 100 or more million
- 19:30records a day into the buy and sell side of media.
- 19:34They're not paying data and data out costs, which sometimes can
- 19:37duplicate, right? You send your data out, gets on
- 19:39board, you bring it back in, and where are the big costs?
- 19:42Like where brand might get hit like if they're let's say they
- 19:45are uploading just like how are these cloud providers like
- 19:48dinging them like and why does it add up so much?
- 19:51Yeah. I mean, it's, it's, it's and
- 19:53it's the it's the CDP's. It's the on borders like live
- 19:56brand and similar. They get charged in two ways,
- 20:00you know, and they should talk for their model, but it's really
- 20:03it's storage and it's compute. So it's number of records is the
- 20:06traditional way. And there's a lot of that still,
- 20:09understandably, because the more records you store, the more it
- 20:11costs. And yeah, it's hard work and
- 20:14then the processing. So then you want to you want to
- 20:16push it to Google? OK, well, I want to push 100
- 20:19million to Google. That's going to cost this.
- 20:20I want to push 100 million to Meta.
- 20:22Well, that's all the work again. Just for that next channel.
- 20:25And yeah, then I want to push it.
- 20:27On average, our clients what push the what, 12 to 15
- 20:31different channels and they want to push everything every day.
- 20:34So yeah, that's a gigantic amount of compute and and yeah,
- 20:39that that means large bills to those SAS platforms.
- 20:43What we realised is because our app runs inside the
- 20:47infrastructure, the cloud infrastructure, which the CTO
- 20:50has, the office of the CTO has already sanctioned.
- 20:53Yeah, all of those costs are already accounted for and
- 20:56actually they're very small. Actually they're very, very
- 20:58small. It's another secret of our
- 21:00industry or the ecosystem. The cloud storage and compute
- 21:04that ACPG or a retailer pays is on a much lower rate than the
- 21:09cloud storage and compute that a software company pays.
- 21:12OK, you go and they can tell that the Subs because if they
- 21:17know that your business is running a software company,
- 21:20they're going to juice you for it.
- 21:21If they think your job is selling fruit and vegetables,
- 21:25well, yeah, you they get you guys, you're getting a much a
- 21:30much better rate so yeah. And yeah, just a spoiler, but
- 21:34yeah, that's why we go with a flat rate that we, you know, you
- 21:38cannot on board billions of records into the ecosystem for,
- 21:43for the same amount of money because, you know, I don't pay
- 21:46an AWS bill. Our our organization doesn't
- 21:50have a Google cloud license, which we, which we run all of
- 21:54this on. So we don't have to pass on
- 21:55those costs. And the movement and the
- 21:57mitigated movement process, I mean, not only reduces costs,
- 22:00but I'm assuming it reduces a lot of the legal compliance,
- 22:03data governance issues that you might run into when you start
- 22:06moving between platforms. Some of our biggest advocates
- 22:09are data previous services and infosec and that's a long, long
- 22:14standing kind of joke. Or bet at audiences that I won't
- 22:19walk on stage. No, next year at, you know,
- 22:23Adweek New York, I'll walk on stage at, you know, info
- 22:27security world in, I don't know, Basel or somewhere like that.
- 22:32That'll actually, because that's a.
- 22:33Different crowd. It's a different, it's a really
- 22:35different crowd, but you can make that crowd laugh.
- 22:39Well, yeah, they're the biggest fans.
- 22:41And this is a direct quote, a yeah, a DPO data privacy officer
- 22:47at a major retailer in the UK said, you know, I'm used to just
- 22:50receiving the next madness which marketing has dreamt up and
- 22:54having to say no. And for the first time, I can
- 22:57say yes. And and these guys, and you have
- 22:59met them, these guys love to represent the business.
- 23:02And you know, they, they they love to echo the corporate goals
- 23:08all around growth, but also around responsibility.
- 23:11So that putting them in a situation where they can say yes
- 23:15is is is really cool. So, Rob, I, I'm really glad you
- 23:18brought up the cost element or two of this.
- 23:21And I think there's maybe even another dimension love to
- 23:23explore this with you. So I've run into this when
- 23:26looking at, let's say, solutions for onboarding first party data,
- 23:29right? For media, not for, you know,
- 23:32other things, right for for, let's say, for me, for
- 23:36advertising use cases. I think a common issue is, you
- 23:40know, the way that most brands buy media, they give their media
- 23:44agencies budget and then the media agency says, OK, we're
- 23:46going to take the budget and yeah, obviously there's going to
- 23:48be tech taxes and other brand safety ad verification taxes
- 23:54that come out of working media, but we're going to control all
- 23:57that. So I think the issue becomes,
- 23:59OK, that if a, if a, if a head of media then finds, oh, there's
- 24:04going to be additional cost for data onboarding, additional cost
- 24:07for this, additional cost for this, it becomes, I don't even
- 24:10have a budget to pay for that. You know, the agency is not
- 24:12going to pay for it, right, Because they don't want it to
- 24:13come out of their body. It would literally come out of
- 24:15the bottom line, right? Where does the money come from?
- 24:18And like do is that an issue run into like who's gonna like, I'm
- 24:21not, it's not the amount necessarily, but like just what
- 24:24budget does it come out of? What P&L does come out of?
- 24:26Have you run into that? Yeah.
- 24:27Well, we had to have a very clear line of sight or of who
- 24:31pays for it. And for our business model, it's
- 24:35the data owner who pays for it, OK.
- 24:38They're the ones who have to get comfortable from a privacy and
- 24:40security perspective. They're the ones who are going
- 24:43to receive the ultimate value from using their data in the buy
- 24:47or the sell side of media. So our teams interface directly
- 24:53with the advertiser or the day or the publisher or the data
- 24:57provider, OK, whoever's data it is, they're our client.
- 25:01Media agencies are our best friends and they're great
- 25:04advocates for what we do. But we we never have and
- 25:08probably within my tenure, never will be paid via an IO.
- 25:13We never come. We never come out of the cost of
- 25:15the media. It's not just an additional kind
- 25:19of toll or tax that comes out of the out of working media.
- 25:22It's separate typically, right? Hey, I mean you, you, that's,
- 25:25that is a that was a great phrase coined regarding tax.
- 25:29And we use that all the time. Yeah, we, we do not believe
- 25:33there should be a tax on activating your own data in
- 25:37media. Yeah, it should be a competitive
- 25:38advantage. This, this, this, this phase
- 25:41we've been through where you have to pay a toll in order to
- 25:46to to use to the data which you've already won the trust of
- 25:49your customers. You've already done the legal
- 25:50work you've already done. You've already built the
- 25:52infrastructure to to house it and process it.
- 25:56This next era, you do not require a third party.
- 26:00And actually third parties are problematic in all of those
- 26:03areas, you know, to hear them. Third parties are problematic
- 26:06for privacy, security and cost when it comes to.
- 26:09And and a media efficiency which, which to your point about
- 26:12working media and how much is actually going to middle men
- 26:14like during that whole process, that's where you lose value
- 26:17because your, your dollars aren't going 100%.
- 26:20And you know, for publishers and data owners, it's even more
- 26:23acute, right? They're actually selling the
- 26:26data. So they're not using their data
- 26:29to to sell more fruit and vegetables or, or more sofas or,
- 26:33or whatever it is. They're just selling the data.
- 26:37You know, this is, this is the data is the commodity.
- 26:40So they're hyper margin sensitive and and that's what
- 26:44that's been the most that's been actually the fastest growing
- 26:46part of our business in the last 12 months has been with
- 26:48publishers and data providers. That's interesting Atlantic.
- 26:52So so. You made an interesting point
- 26:53there. So I think like the data owner.
- 26:55So I think that if if someone in media finds OK, we don't have
- 26:58budget to handle this. But you're right, it's
- 27:00critically important. So I actually think it elevates
- 27:02it to not just in the media silo, but this is a general
- 27:05marketing issue, right. It should, this should go up to
- 27:08the CMO and you know, if they don't have the authority to do
- 27:11it, I mean, who's going to right.
- 27:12So I almost think that it is that critical.
- 27:15But then maybe even going back to your like your initial point
- 27:20earlier you had mentioned there was a 3% of brands or three like
- 27:24the 3% number. Maybe you can collaborate on
- 27:26that because I fight, I mean, I didn't find that shockingly low
- 27:29having worked in the space a little bit love for you to maybe
- 27:31explore that a little bit. And again, like, how like are
- 27:34you seeing needle move at all now?
- 27:36Yeah. I mean, look, that's the state
- 27:37of the union and we we can see a lot of good work both, you know,
- 27:42with our teams, our clients and and the rest of the industry
- 27:44which it's progress is being made at last.
- 27:47But yeah, that that. I don't know where you got that
- 27:51set from. Yeah, yeah.
- 27:53So we got it and it's. Just for the audience, 3% of
- 27:56media spend is informed by first party data on the brand side,
- 28:00which is a minuscule percentage if you think about it in the
- 28:04larger scheme of things. So it's absolutely the case and
- 28:07we heard it direct from from the brands, but more importantly,
- 28:10actually, we heard it direct from the people doing the
- 28:13buying. So the guys at WPP, at Densu in
- 28:17Publicis, yeah, who account for, you know, that huge percentage
- 28:22of all, all media. I heard, I heard two years ago,
- 28:26I heard from a it was about 18 months ago, I heard from one of
- 28:28the retail media networks that only 1% of advertisers were
- 28:33using first party data in the in, in their in their clean
- 28:37room. This and this is not long ago
- 28:39it's I it's going up from that, but it's for retail meeting the
- 28:41most obvious place that you would do it right.
- 28:43So the 3% number of overall spend didn't surprise me.
- 28:47Keep going. Yeah, and I look I it, it's
- 28:52difficult because the truth is, and it's hard to say sometimes,
- 28:57the truth is the media is going to get bought anyway.
- 29:03Like the data is super powerful, but the, the game in media and
- 29:08marketing is, is to get the reach, meet those objectives.
- 29:13No campaign is going to be paused whilst they wait for the
- 29:16data to make its way to ADSP. Like the campaign goes out the
- 29:21door every. I'd say almost almost to the
- 29:23contrary, right? I think that it's like no one
- 29:26wants things to get paused, right?
- 29:28So not only will the campaign not be paused, but anything that
- 29:31creates or adds friction will be potentially not even considered
- 29:36or or maybe de prioritized. Is that true?
- 29:39Friction and cost you, you which you guys represented earlier.
- 29:42And, and that's where, yeah, the office of the CMO is super
- 29:47important. You know, of course they're the,
- 29:49they're the beneficiary of this. But the benefactor of of these
- 29:55new models, especially where audience plays, is actually the
- 29:59office of the CTO, OK, which is great news and that's a good
- 30:03point. It's it's the technical
- 30:04application of the data and they're the sort of masses of.
- 30:07That domain, hey, it doesn't mean that suddenly CTOS care
- 30:10about advertising that it's and it's OK, they don't.
- 30:14But what they care about is they made gigantic enterprise level
- 30:17strategy decisions to invest in cloud computing and to to.
- 30:23To date, the office of the CMO has barely been a stakeholder of
- 30:28theirs. Every time a marketing division
- 30:33buys another third party software where the data needs to
- 30:36be copied and put somewhere else, the CTO has to make an
- 30:40exception to their enterprise infrastructure and that that
- 30:44kills them and. That what's happening and that
- 30:46can be in the dozens of platforms.
- 30:48I think the average CMO gets has about 50 to 60 different point
- 30:51solutions that they leverage, which it's not all copying data
- 30:54necessarily, but still it's. But there's a lot of discussion
- 30:59and research around the cost of all of those solutions.
- 31:02I don't think that that's obviously important, but
- 31:05breaking the strategy, breaking the enterprise strategy.
- 31:08As marketers or people in marketing, we might, we, we have
- 31:12a tendency to think that marketing is extremely important
- 31:15in large organisations market. The CMO is often not a real C
- 31:20isn't actually. Sad it's it's really AAC suite
- 31:24position that's and many orgs have even eliminated it.
- 31:28They have like a VP marketing app but even once.
- 31:30In customer office or. Even something like that, it's
- 31:32become less important or less powerful.
- 31:35Yeah. And I tell you what, they're not
- 31:36in the suite, they're not in the meetings.
- 31:40Contrary to that, the CTO always is, OK?
- 31:44The CTO always is. And that's because technology,
- 31:47just real technology decisions, infrastructure decisions impact
- 31:51risk and growth. And actually, I'd have to drop
- 31:54this point. The game changer is when we
- 31:59found is when you act as infrastructure.
- 32:02Very unsexy word, but when you act as infrastructure for data
- 32:08activation, you get treated seriously.
- 32:11Yeah. Yeah.
- 32:12Because you're, you're coming in there and the CTO is coming in
- 32:14there and saying, hey, we're going to make this our, our use
- 32:17of our data is going to be exponentially more efficient,
- 32:20right? We're going to be able to use
- 32:21this to create second party audiences, to expand our reach
- 32:25beyond our own first party data set.
- 32:27And we're going to be able to, yeah, hold on a whole bunch of
- 32:30our our vendors, to your point around sort of the vendor bloat,
- 32:33to the point solution bloat, right?
- 32:36There's an efficiency play. And then I think there's an
- 32:38effectiveness play here that that seems to be like you said,
- 32:41in the hands of the CTO in terms.
- 32:43Of maybe. Maybe I'm a gigantic
- 32:45multinational publisher. Maybe I'm a massive big box
- 32:49retailer. I want to get that you're so
- 32:55they're so data rich, gigantic amounts of data.
- 32:59And if they they have limited to no barrier to getting all that
- 33:03data, say to the supply side like an index exchange or a
- 33:07trade desk, they can unlock those retail media strategies,
- 33:11etcetera, etcetera, which they've been after.
- 33:14If it's working off like that infrastructure, which they've
- 33:17already Phil, it's it it Nothing's a no brainer in large
- 33:23organizations, but what it is, it's tick, tick, tick value.
- 33:27Yes, but but how far away are we when you work with large
- 33:30organizations like that? Because I've seen a ton of that
- 33:32on on in my career. How close are they to getting to
- 33:36that, that promised land of kind of unified data across very
- 33:39large siloed organizations with multiple data sources?
- 33:43I mean, are, are you finding that that's we're getting closer
- 33:47and closer to to big brands getting more sophisticated in
- 33:49that, in that way? We are, we are and, and and I
- 33:55talk about it from data for advertising, right.
- 33:58So there are other use cases that can be more complicated.
- 34:01Yeah. Honestly, I believe the last 10
- 34:03years there's been a lot of distraction.
- 34:07You know, a lot of talk about, I'd say one of my favorite ones
- 34:10is identity resolution. OK, So look, I grew up in
- 34:14marketing and media. Media loves scale because it
- 34:17loves reach, OK? And it loves addressability.
- 34:22Identity resolution in my view is and femur to that OK,
- 34:29identity resolution. Look, and this is my point of
- 34:32view, Identity resolution is about picking the one ID that
- 34:37you're going to use out of many IDs.
- 34:40The most persistent of the IDs. Yeah, I'm not interested in
- 34:43that. I'm born out of media and
- 34:44advertising. As long as there's permission or
- 34:47you're comfortable to use with the permission for those, I DS
- 34:50we should use all of them and then the demand and supply side
- 34:54will dictate what's useful or not.
- 34:56Don't worry about that. Don't waste any time doing that.
- 34:59Remove as many barriers between your data and the trade as
- 35:04possible and you'll be rewarded. You'll be rewarded in that.
- 35:08And yeah, so, so actually, and Matt, my Co founder, like often
- 35:15talks about this. There's there's been a like a
- 35:17data inferiority complex in many organizations where they're
- 35:21like, Oh no, but we've got multiple silos of data or we've
- 35:24got yeah, we still need to clean up that data.
- 35:27Well, is it permissioned? Yes.
- 35:30OK, let's get to work. And you can bring multiple
- 35:33sources of that. In the end, you can push, you
- 35:35know, many of our clients, you know, they have four or six or
- 35:39eight different divisions and they've been built through
- 35:42acquisitions. So they've all sat with multiple
- 35:45warehouse tables for each division.
- 35:47Yes, story as old as time and they've had this some confusion
- 35:51over the years of hey, we should create a single customer view,
- 35:54bring it all together. But financing in these
- 35:56organisations will never finance a, a cross organizational
- 36:00project like that. So you know, years go by, it
- 36:02still hasn't happened. If you remove the the the
- 36:08engineering and the cost of pushing let's say 8 different
- 36:11data sets into the same DSPC for example, and then the DSP does
- 36:17all of the resolution, yeah, you don't have to do it.
- 36:21And also. It's interesting you're saying
- 36:24though, because I think looking at marketing versus advertising,
- 36:28right, I mean, and especially I also think it depends on the
- 36:30industry too, right? When you have additional
- 36:32compliance, but like advertising, you don't need to
- 36:34be as precise, right? Like if you're marketing to your
- 36:38own customers, you, you better have, you better know who
- 36:40everyone is, right? Better not.
- 36:42But if it's advertising, you're just trying to drive demand.
- 36:45You know your fidelity doesn't have to be quite as precise,
- 36:48right? Because you're.
- 36:49As you're just in, in, in, in, in, in, in, in the campaigns
- 36:53that going out at any given point in time, the advertising,
- 36:57in the advertising campaigns, any point in time, it's not
- 37:01there are going to be far fewer customer targeted campaigns than
- 37:05there are acquisition campaigns. OK.
- 37:07I'm sure someone will find an exception to that.
- 37:09But like if I walk into, you know, a telco and look at their
- 37:14campaign, lay down 980 to 90% of it is acquisition and then yeah.
- 37:2020 Should we just think from CRM, right?
- 37:21Where, where, where or even or even service emails, right.
- 37:26I mean those you better be get right.
- 37:27But you're right for advertising, yeah, that's not
- 37:29our. Job, that's not our job.
- 37:31And also for those people who are doing their job.
- 37:33And another great thing that gets a lot of airtime at the
- 37:36moment, we talk about CRM managers being like the most one
- 37:39of the most undervalued roles in organizations because they can
- 37:43bring everything they've been doing in direct for such a long
- 37:45time. Like yeah, but but isn't isn't.
- 37:51Customer growth as well as customer acquisition, isn't that
- 37:55all in a sense fall under the, the, the category of
- 37:58advertising, right. So if I'm a retailer and I'm
- 38:01saying upsell, cross sell opportunity, you might also like
- 38:04suggestions that are being sent to me by e-mail or whatever
- 38:07direct channel they've got with me.
- 38:09Isn't that advertising? That is, that is and you can do
- 38:14a lot of good work there. But what I what we find over and
- 38:17over again is advertisers, the majority of their focus is
- 38:23campaigns which are acquisition focused and the role of data by
- 38:27the way is gigantic in those. So first, exclude all your
- 38:32deterministically exclude all of your existing customers.
- 38:35Everyone feels smart and clean and clever after they've done
- 38:38that. We suppressed our customer looks
- 38:40great. OK.
- 38:42That's something that you ship like people, you know, decades
- 38:44ago. They have, But not in
- 38:46advertising, man. Not in advertising, not in
- 38:49advertising you. I don't know.
- 38:50Whatever. No, you're right.
- 38:51You're right, like direct mail industries like that has been
- 38:54doing it, but. You're getting ads for them on,
- 38:56you know, in the open Internet, in the Water Gardens and on CTV
- 39:00all day. But.
- 39:01You're right. Like, no, you're totally right.
- 39:03Resolution. Problem.
- 39:05Yeah. Well, well, no, Yeah, Yeah.
- 39:06Nice. It's it.
- 39:08They're not suppressing and they're wasting money talking to
- 39:10existing customers. And there's a massive customer
- 39:12experience problem in that as well, because you're showing
- 39:14offers which you're not eligible.
- 39:15For well, the worst is retargeting when you've already
- 39:17bought something when like the same pair of sneakers follows
- 39:19you for weeks and you're like, I bought this is crazy, right?
- 39:22What AI mean? Retargeting is maybe this, you
- 39:24know, it's the scale is not as big, but it's a lot of money
- 39:26that's being wasted. It's a lot.
- 39:28So look, we we just find in the first month or so of working
- 39:31with a client suppressing existing customers for
- 39:34acquisition, that's huge, powerful, super powerful it's.
- 39:37So easy. It's so easy.
- 39:38The fact. That not doing it's crazy then
- 39:40obviously seeds for look alikes. Whatever your version of a good
- 39:43looking customer is, get that into the channels and keep and
- 39:46keep refreshing those as much as possible.
- 39:49And you know. What else is good I've seen to
- 39:51Rob is that like not only can you, but then like, but that's
- 39:54only just the most basic stuff with first party down.
- 39:56Then when you pull in lifetime value, average order, volume,
- 39:59things that you know because like their customers.
- 40:02And then you can begin to do really amazing things.
- 40:04What are our best customers really look like?
- 40:07What customers should we be marketing to, right?
- 40:09And then like, how do we take these and then extrapolate these
- 40:12from seed audiences to bigger audiences that are out there?
- 40:14I mean, I think that's something very few people are doing.
- 40:16And I'm sorry I cut you off, but I got excited.
- 40:18Very people, not enough people are doing it.
- 40:21That's where the huge value is and I would link it back to like
- 40:25the relationship between data signal and advertising is very
- 40:30different between the relationship between data and
- 40:32direct channels. OK, and that is why it's data is
- 40:36so rewarding when you use it for suppression and sees for look
- 40:39alikes and advertising and when you use it for targeting to
- 40:42upsell cross sell your existing customers, advertising is very
- 40:46disappointing. OK, I still think there's lots
- 40:48of good work to be done and it should be, but the role of other
- 40:51channels is far more successful there.
- 40:54Like you walk and also you walk into the media agencies, they
- 41:00are very adept at building acquisition campaigns.
- 41:03They're less adept at customer growth.
- 41:05OK, so we we pick as few fights as possible and OK well.
- 41:11We're up on the agency front and I'd love to know, like how have
- 41:14you been working with agencies? I mean, I mean, agencies are
- 41:16changing. I think you are becoming like
- 41:18building their own. They're trying to get better
- 41:19technology. I mean, I think some are doing
- 41:21it better than others. They're starting to have their
- 41:22own platforms. The big focus on bringing their
- 41:24own identifiers so they can have some IP.
- 41:27How are you working with them and like what challenges have
- 41:29you had and what successes? Love for you to talk about that
- 41:31a little bit. Yeah, we, we love the fact that
- 41:35all of the major networked agencies and many of the Indies
- 41:39that we work with already, they're big supporters of our
- 41:42approach and the way we do it. Fundamentally we they get to do
- 41:48their job. When we have a client in common,
- 41:50they get to do their job. But the ad account running are
- 41:54enriched with the data signals which we're providing.
- 41:59So like, literally, they don't have to lift a finger, they
- 42:01don't have to pay for anything. It's just their clients are
- 42:06giving them tools which are, yeah, super powerful in order
- 42:09for them to do their job. So we love that they're super
- 42:13smart and get it very, very quickly so far.
- 42:19And this surprises some people so far like those acquisitions
- 42:22that publicists have made that WPP have made in the clean room
- 42:26and and and similar space. It doesn't really doesn't really
- 42:30impact us because I think there are multiple reasons and I
- 42:34don't, I don't really if that was kind of part of your
- 42:37question. But fundamentally the data
- 42:40owners are the advertisers or in the third party case the
- 42:44publishers or the data or the data providers agency.
- 42:49In my view, I worry for agencies when they talk as if they are
- 42:54the data owner, OK that they're in often the data owner.
- 42:59OK they're they're they're they're they occasionally are a
- 43:02trader of data in between a data owner and the advertiser.
- 43:08Publicis is maybe the best exception with Lotomy where they
- 43:13really have acquired some some serious data sets.
- 43:17I think it's an incredible business.
- 43:18I will say I'm not allowed in. Epsilon publicist with epsilon
- 43:21as. Well, right.
- 43:22And then yes. And then.
- 43:24Omnicom and Axiom, right? So you're getting, Yeah.
- 43:28But we wouldn't be alone in saying, yeah, are those third
- 43:34party owned data sets, You know, is that really the future?
- 43:39I'm not saying it's not, but I don't think it's the whole of
- 43:41the future. I'm pretty confident that the
- 43:44future contains the data owner, the advertiser being the data
- 43:49data owner, and them being able to use their data alongside.
- 43:52Which I think is inherently more powerful.
- 43:55And in Rio when we talked to George Musey, who, who will he
- 43:58will be publishing shortly, he, he talked a lot about that like
- 44:01like all of these database kind of identity resolution
- 44:04providers, the Epsilons, the axioms, the new stars, they all
- 44:07came from the same kind of core database marketing fabric from
- 44:1125 years ago, right? So there's not a ton of
- 44:15differentiation argument arguably within their data
- 44:18themselves. And I and I think that sort of
- 44:20supports your thesis that really the data owner has the power in
- 44:24terms of the quality, the depth, the consistency, the real time
- 44:28nature. Yeah, they know their customers,
- 44:30their data, first party data is going to be.
- 44:32Yeah, it's going to. Be much better.
- 44:34Let me ask you if this is. Relevant to the brand.
- 44:35Oh, totally, totally. And and you mentioned clean
- 44:38rooms, Rio and and Rob, I'd like to hear your perspective on
- 44:40this. Is that to me, the number one
- 44:42use case of of sort of clean rooms or at least native data
- 44:45architecture where you can is second party use cases, right?
- 44:49Where you're, you know, you're a retailer partnering with the CPG
- 44:53company. You're a, a, a, an airline
- 44:56partnering with a hotel company to say, Hey, we're going to
- 44:58cross target, cross pollinate. Because we know if your
- 45:03customers arriving in Miami and staying at the, at the, you
- 45:06know, on JetBlue that the Marriott might be interested in
- 45:10targeting them with the hotel offer, right?
- 45:12Have you seen a lot of those use cases?
- 45:14Because I found that that they're honestly a lot of brands
- 45:16aren't particularly sophisticated, which surprises
- 45:18me. Despite all the clean room hype
- 45:21that we've heard in the last 20 years or 10 years.
- 45:24Have you seen a lot of those use cases being sort of leveraged
- 45:27and thought about? So we've seen a lot of work in
- 45:30that space and a lot of use of clean rooms.
- 45:33You know, I'd say only slightly tongue in cheek, a lot of use of
- 45:37clean rooms to create very expensive Venn diagrams.
- 45:42OK. Of how these how this data could
- 45:45could meld and be used together. You might imagine I've got two
- 45:49issues with that. One, that that data that you
- 45:56brought together on a clean room, that's not the match data
- 46:00for the DSP, that's the two brands data.
- 46:04You haven't done the match of the DSP, so it's not your real
- 46:08audience in media yet, OK. And and two, you've had to
- 46:13compromise privacy and security in order to just create this
- 46:17Venn diagram, which isn't even actionable in media yet, because
- 46:21in order to make it actionable in media, you're going to have
- 46:23to onboard it into a DSP. Yeah.
- 46:27And that that might be a 50% or? 30% if you're using Amazon
- 46:31Marketing Cloud or ads or ads data Hub ADH, in theory, you you
- 46:35are pulling in platform signals, you know, for the match.
- 46:39So I think those would be the exceptions, but I've been
- 46:41further just for generic, for generic cleaner and like no, you
- 46:44would definitely would not be. Well, see that you see, yeah,
- 46:48that worries me though, because what you're a brand strategy, an
- 46:53advertiser strategy is going to be Amazon centric.
- 46:57Yeah, like that, that might work for then your Amazon DSP
- 47:02campaigns. But I'm also want to, I want to
- 47:04push out to Disney Plus and I want to push out to the trade
- 47:08depth. Or for one channel 100% that
- 47:10would not. That would not work for others.
- 47:12It's very important, you know, I, I, I, I throw shade on, on
- 47:16something, but I want to say the reason I do that is because the
- 47:19answer I believe is we believe is just staring us all in the
- 47:22face, which is brand A and brand B push their customer files
- 47:28encrypted of course, to a DSP or to multiple DSPS and make that
- 47:33overlap happen at that DSP at the point of media.
- 47:36OK, then you're doing it with your, you're finding that
- 47:39overlap with the real available audience and you can action it
- 47:43straight away. I probably don't need to say
- 47:46this, but there's no associated cost for doing that as well.
- 47:51And the data was going to get be of that DSP anyway.
- 47:54So all of the Dpas and stuff like that that you'd need to
- 47:57sign in order to work with that DSP, we're always going to be
- 48:00signed anyway. You're interesting, Robbie bring
- 48:02up like in the future, if we really think like some kind of
- 48:05agentic trading will start taking place between let's say
- 48:08buy and sell side, right. I mean in theory, like you know,
- 48:13we Brett, you remember the conversation with Doctor Fu the
- 48:16other day? He was basically saying he
- 48:17thinks the future is direct buys on the the you know, the
- 48:22advertising, you know, the buy, the buying portals for the
- 48:25different publishers from theory, a solution like yours,
- 48:27Rob, it could just be sit in the middle, right?
- 48:30And you're allowing like a encrypted push from, you know,
- 48:33from brand over to over to the advertising portals for these
- 48:38for these publishers. You're doing the match directly
- 48:40there. In theory that could happen,
- 48:41right? And I think that it it really
- 48:43can. And yeah, the commercial, the
- 48:48unlock that we see is we've removed the cost of data
- 48:54transportation. Yeah.
- 48:57Because you make it easy. So instead of having to build
- 48:59integrations with 15 different like, you know, 1520 different
- 49:02like let's say places you could buy CTV inventory, like your
- 49:05platform can make that very easy.
- 49:06I think that is kind of interesting.
- 49:07I. Mean we make it easy, but it's
- 49:10also you do more of it, it doesn't cost you more like like
- 49:14we have a retailer who on boards probably into 30 different DSPS
- 49:20every day. Yeah, there's some of it's
- 49:22delta, of course. So, but, but for the purpose of
- 49:25my point, I'll act as if it's the whole file.
- 49:28They're pushing like 12 million records to, yeah, 15 different
- 49:33DSPS every day Like that's and they could double that.
- 49:40They could it, it, it it it's. Cost won't cost won't scale as
- 49:45they as they as they increase their volume, you're saying?
- 49:49That's the amazing there is. When you remove the incremental
- 49:52cost of signal delivery, then you can put that signal
- 49:57everywhere. And that's where an agenda,
- 49:59especially an agentic approach can really pay off because
- 50:04agents can go and seek value of using that data everywhere.
- 50:10You're not saying hey? Yeah, without having to like
- 50:13somebody internally having to manage that complexity.
- 50:15Yeah. And making a judgement call,
- 50:16right? We can afford this quarter.
- 50:19We can afford to push it to the Trade Desk and to Meta, but
- 50:22that's it. And then you put this out on the
- 50:25hold of the rest of the ecosystem.
- 50:26Yeah. Well, and Doctor Fuke kind of
- 50:28made a good point that that there's so much consolidation
- 50:30that there really are a handful, maybe 10.
- 50:33You can count them on two hands, places that you really need to
- 50:36push your data. Yeah, I think you said 1015,
- 50:38Max. Yeah, because you're already
- 50:40with those platforms in mind, the walled gardens, whatever you
- 50:43want to call them, the retail media networks, you're hitting
- 50:4495% of of your of your addressable audience anyways.
- 50:48So what do you need all of these extension plays and the further
- 50:51you extend out the the kind of less trustworthy the inventory
- 50:57gifts. Right.
- 50:58I got really, I got a worried the other like a couple of weeks
- 51:01ago talking to one of the one of the companies that runs one of
- 51:05the biggest, well, a lot of the big retail media networks here
- 51:09in the UK at least. And they're talking to me about,
- 51:12you know, they've got 5 or 6 big retailers, supermarkets on their
- 51:17books and they're saying, yeah. So once we've, you know, we're
- 51:20planning to bring this data together and then we want to
- 51:22extend it. And I was like, in the UK, if
- 51:25you've got five of the biggest supermarkets and you've got that
- 51:29like that data set is going to include everyone in the UK are.
- 51:33You going to? Extend that, right, Yeah, that
- 51:36doesn't, that doesn't that doesn't like.
- 51:40And the answer in that is because of the associated cost
- 51:45or perceived associated cost to bringing all that data.
- 51:47They weren't planning to bring all of the data together.
- 51:49Yeah. So they were trimming it and
- 51:51trimming it and trimming it. That's like, well, this is this
- 51:54valuable asset and that's the power.
- 51:56When you, when you remove the cost, the incremental cost per
- 52:00line of data, then the industry can get to work, the ecosystem
- 52:04can get to work. And yeah, it's only going to,
- 52:06it's only going to be accelerated.
- 52:08And it's in arguably it's real reach versus this sort of, you
- 52:11know, expansion play where you're reaching people that
- 52:14don't may not even exist. It may be inventory that's just
- 52:17completely fraudulent. That's kind of Doctor Fu's
- 52:19point, right? Like first part he did ensures
- 52:21that you're this is real reach and you're and if you go
- 52:24publisher direct or platform direct, the likelihood of
- 52:28hitting your audience increases, right?
- 52:30I, I, I think we've, we find that the people who are obsessed
- 52:35with audience extension, reach extension, synthetic audience
- 52:40data, they're the people without the data.
- 52:46And I get that. If I yeah, I get that.
- 52:48I get that. But we, we work in service
- 52:52anyway. We work in service of the data
- 52:54owner. And very often it's been first
- 52:58party. But it's very important for me
- 52:59to say we've seen an explosion in data, data providers and
- 53:05publishers. People like Experian and others.
- 53:08You know, if you're a publisher, your core, your core business is
- 53:13publishing. You're not a data business as
- 53:15much as they might have tried to say that you're not, You're a
- 53:18publisher, OK? And that's hard work to do
- 53:21publishing and to create content, activate getting value
- 53:26from their data sets. That's they're not looking to
- 53:29add massive costs. And actually it stopped them
- 53:32from being able to do data providers.
- 53:34They've stopped and you'll see it.
- 53:35They've stopped describing themselves as data and
- 53:38technology companies. They don't say that anymore.
- 53:40They used to 10 years ago. You told.
- 53:43You're in our TransUnion. They got hey, yeah, we're a data
- 53:46technology company. They don't anymore because
- 53:48they've they've not they found that they they they are
- 53:51brilliant. They are brilliant data
- 53:53companies and they rely they must rely on other people.
- 53:56And yeah, sadly they are tactical relationships often
- 54:00with the likes of Live Ramp and others in order to get snippets
- 54:03of their data on advertiser by advertiser basis.
- 54:06Intermediate. Well, we are very, very happy to
- 54:11be doing for them on both sides of the Atlantic is being able to
- 54:14say, hey, we can, you can on board your gigantic valuable
- 54:19data sets, intermedia, Yeah, in their entirety and it doesn't
- 54:25have to destroy the margin of your data business.
- 54:29That's a good. Point so so.
- 54:30Rob yeah, we unfortunately don't have too much time.
- 54:33We're about, we're about a little under an hour, but we'd
- 54:36love to maybe to hear from you like how if anyone wants to get
- 54:39a hold of you, wants to find out about your business, how would
- 54:42they do that? I mean, what's the best way?
- 54:44Thank you. Well, yeah, we're very, we're
- 54:50very excited to be servicing clients both in North America
- 54:53and in Europe. Our headquarters are here in
- 54:57London. So yeah, always happy to have
- 55:01people over. But yeah, you should get in
- 55:03touch with myself or the rest of our team here.
- 55:07Go to weareaudiences.com and, and, and see for yourselves.
- 55:12We've got a, you know, we've got, we've got some great
- 55:15vertical leads who've really like affected understanding how,
- 55:19you know, if you're a publisher, if you're an advertiser, A
- 55:23retailer or, or a data provider that how can you with as little
- 55:29disruption as possible be able to bring that, enable signal
- 55:34delivery and unlock those revenue streams which, yeah,
- 55:37which, which as an industry we've been waiting for for so
- 55:41long. So maybe before we go, we can do
- 55:44some quick hits. I think we had a couple of
- 55:45minutes left. What do you say?
- 55:46Sure. Well, there's Rob getting that
- 55:50right. Yeah.
- 55:51So, so how about the first one? Biggest myth about first party
- 55:54data? Yeah, that people don't have it
- 55:58or they don't have enough of it. That that advertisers don't have
- 56:01enough of it. Yeah, that's a myth.
- 56:03That's it. It's wrong.
- 56:05Yeah. It's more about utilization,
- 56:06effective utilization of that versus not having access,
- 56:09access, access, yeah. What's the most over hyped part
- 56:13of the data stack or tool in the data stack?
- 56:16CDP, it's irrelevant, completely irrelevant.
- 56:19And I include composable in that.
- 56:21Completely irrelevant. In an area of cloud and data
- 56:24warehousing. Yeah, I worship.
- 56:26I worship. We worship at the feet of
- 56:29snowflake data bricks and Big Query.
- 56:31That's where the action is. That's where the data is.
- 56:34Nothing else is relevant. Nothing is really.
- 56:36You put anything in between it, composable or otherwise, you're
- 56:39bringing cost, you're bringing complexity.
- 56:41Don't do it. OK, that's it.
- 56:43Very well said. So suppression reach targeting
- 56:46the retail media, which is the most undervalued use case
- 56:50suppression. Yeah, suppression.
- 56:52We talked about it, right? It's massive, yeah.
- 56:56What kills activation faster? Legal, legacy processes, tech
- 57:01stack, etcetera. Cost.
- 57:10Yeah, Yeah. Cost, cost.
- 57:12So, so on that topic, I mean, I think we could probably end with
- 57:15this is what what is one thing and we've talked about it within
- 57:18this this podcast is that brands spend too much money on right
- 57:21now. Yeah, third party technology,
- 57:24they need to, they need to go. They need, they need to turn
- 57:27around. If you're a marketer, you need
- 57:29to turn around and say, hey, what cloud infrastructure?
- 57:35I know it's boring and it doesn't sound very marketing,
- 57:37but what cloud infrastructure? What data warehousing
- 57:40infrastructure? OK, that's where all the data
- 57:43is. Your dreams will come true when
- 57:44you lean into that. Yep.
- 57:48All right. This is great.
- 57:48I think we we obviously could have kept talking.
- 57:50That's the topic we're all three passionate about.
- 57:53I love your enthusiasm here. Yeah, but maybe we could do make
- 57:55it do a Part 2. I mean, I think it's, it's a
- 57:58cool topic area. So thanks for joining.
- 58:00And thanks everybody for listening and thank you very
- 58:03much. Yeah, thanks for having.
- 58:04Thanks for joining, Rob, and thanks everybody for listening
- 58:06that's made it this far through the podcast.
- 58:08You can reach us at Deb Deb, Deb dot signal and Noise dot AI and
- 58:12subscribe and share signal and noise on Apple Podcasts, Spotify
- 58:17and YouTube. And thanks and see you next
- 58:19time.