WEBVTT
00:00:00.160 --> 00:00:02.720
Welcome to Executive Conversations by the Papers.
00:00:02.799 --> 00:00:04.000
I'm Dwayne Jeffrey.
00:00:04.160 --> 00:00:09.759
In this episode, I'm joined by Darmuk Thoma, Head of Fraud and Data Strategy at Ad Data.
00:00:09.839 --> 00:00:17.280
We're exploring how email and data intelligence is used to detect fraud earlier and the role data strategy plays in staying ahead of it.
00:00:17.920 --> 00:00:22.559
For someone meeting Ad Data for the first time, what does your company actually do?
00:00:23.199 --> 00:00:27.519
We're heavily focused, in fact, almost entirely focused on email, right?
00:00:27.679 --> 00:00:32.079
So what I primarily look after is the fraud prevention side of it.
00:00:34.079 --> 00:00:36.159
A lot of people don't even know the brand very well.
00:00:36.240 --> 00:00:40.799
And it's but it's actually quite a mature, it's a 26-year-old company.
00:00:41.520 --> 00:00:43.920
It's the largest email network in the world.
00:00:44.159 --> 00:00:53.439
Um so what that kind of translates to is we process just about two billion email interactions a month globally in about 190 countries.
00:00:53.759 --> 00:01:08.079
Um and that gives us insight into you know kind of what the email activity is happening, and that allows us to do things like fraud prevention, say how old the email, how good it is, what it's doing, um, and allows us to do a lot of stuff in terms of fraud prevention.
00:01:08.239 --> 00:01:10.000
So that's that's kind of it.
00:01:10.159 --> 00:01:16.799
We try intentionally try and keep very simple, keep it, you know, easy to use, easy to understand, and uh easy to deploy, of course.
00:01:17.040 --> 00:01:17.519
Absolutely.
00:01:17.599 --> 00:01:24.879
And then when you're talking about financial companies, fintech companies, payments companies, what are they utilizing uh ad data for?
00:01:25.200 --> 00:01:29.519
Well, it's primarily kind of type KYC type things, right?
00:01:29.680 --> 00:01:33.519
So uh in in finance, there's a lot of different checks that have to happen.
00:01:33.599 --> 00:01:36.879
You know, there's credit checks, KYC checks, uh AML checks, and so on.
00:01:37.040 --> 00:01:38.959
Quite a lot of them are quite expensive, right?
00:01:39.120 --> 00:01:42.640
So we're not, you know, we're quite easy to deploy, quite cost efficient.
00:01:42.719 --> 00:01:47.599
So a lot of the stuff is used very early in the funnel uh for fintechs and and finance companies.
00:01:47.760 --> 00:02:00.799
So we're you know, literally on the point of, you know, as soon as they're creating an account with the with the company, we're able to tell them, yeah, that in the the the name associated with the email that you got, that's that's correct.
00:02:01.040 --> 00:02:14.159
The email is you know 20 years old and it's quite active, it's operating lots of different locations, and uh the IP is quite close to the address that you're so we have a lot of insight very early in the funnel, and that gives them confidence of the customer.
00:02:14.240 --> 00:02:27.280
Um, so you'd be surprised, like I've seen like certain like 20% of what the fintech could get could be just rubbish, just you know, nuisance, either intentionally fraud or just just really um, you know, it could be bought uh created accounts and stuff like that.
00:02:27.360 --> 00:02:32.080
So they're able to get rid of a lot of that very early in the funnel for relatively minimal cost.
00:02:32.400 --> 00:02:32.960
Absolutely.
00:02:33.039 --> 00:02:40.719
So there's of course a lot of different signals out there, but most of us treat our email address as you know, just something more mundane.
00:02:40.879 --> 00:02:42.639
We even barely even think about it.
00:02:42.800 --> 00:02:49.120
But your company tends to treat it as it's one of the most revealing signals of somebody who's really online.
00:02:49.919 --> 00:02:52.960
So, what does an email address actually tell you?
00:02:53.280 --> 00:02:58.879
Well, I I mean why data is important is effectively why is because email is so important, right?
00:02:59.120 --> 00:03:05.840
Because uh I've had I've had lots of different emails, but my primary email I've had since I was about 16 and I've kept it with me.
00:03:05.919 --> 00:03:13.759
But I've moved I've changed devices, of course, every couple of years, moved addresses, I've you know, uh uh people even change names, but they'll get their email.
00:03:13.919 --> 00:03:14.560
Yeah, right.
00:03:14.719 --> 00:03:21.280
So in a digital world, it it really is the crux of your your digital identity.
00:03:21.439 --> 00:03:27.280
In fact, most of it's you're surprised a lot of our uh clients refer to it as the email ID, not the email address.
00:03:27.680 --> 00:03:31.120
And that's a bit telling, I think, you know, because there's like it's it's so important.
00:03:31.280 --> 00:03:35.599
A lot of the time we'll see on account opening, you'll see uh email.
00:03:35.680 --> 00:03:35.840
Yeah.
00:03:36.000 --> 00:03:41.120
Nothing else, not your name, not your just enter your email and whatever you know, other data point they need.
00:03:41.199 --> 00:03:45.039
But that's it, just to keep the it it to a minimum and then they can follow up from there.
00:03:45.280 --> 00:03:56.639
So it's so important and it's so you know, has a such a long tenure uh within that that evolution of a profile and that it it makes it valuable from a you know customer point of view, and well, evaluation standpoint.
00:03:56.960 --> 00:03:57.520
Absolutely.
00:03:57.759 --> 00:03:58.080
Okay.
00:03:58.560 --> 00:04:02.080
So let's bring that into sort of like the real world, right?
00:04:02.159 --> 00:04:03.599
Like I want to imagine that.
00:04:03.759 --> 00:04:12.159
So imagine I'm an online lender, I'm approving loans, um, and a real share of that application, of course, comes from real people.
00:04:12.319 --> 00:04:21.839
Um, but without getting too technical, how do you know something about an email address that helps me tell who real customers are from the rest of them?
00:04:22.319 --> 00:04:25.360
So I suppose there's two scenarios, like the the good, the good path, right?
00:04:25.519 --> 00:04:31.360
The happy path, which is we hope most of your your client base, 80% hopefully, is is is good and or plus.
00:04:31.600 --> 00:04:35.839
Uh and what what when they come in and we get the email, it's all real time.
00:04:36.000 --> 00:04:39.199
Like so our environment, it's like uh it's almost like the Visa Credit Guard network.
00:04:39.279 --> 00:04:41.120
It's emails constantly interacting with us.
00:04:41.279 --> 00:04:44.639
We'll see, you know, we see billions obviously coming back, but we see all the interactions.
00:04:44.959 --> 00:04:49.279
So when a when a client queries us, we're querying it in right that second, right?
00:04:49.519 --> 00:04:51.040
What's happening with that email right now?
00:04:51.199 --> 00:04:52.720
So we give that context back.
00:04:52.800 --> 00:04:56.240
So it's all really fresh, really high high context data.
00:04:56.399 --> 00:05:01.120
But the what we're gonna say in um in the good flow is yeah, Dwayne.
00:05:01.199 --> 00:05:04.399
He we saw him first time for uh 16 years ago.
00:05:04.560 --> 00:05:10.240
He's uh the address that's the one we normally see, the name, that's the one we normally see with this email.
00:05:10.480 --> 00:05:15.360
And you know, IP looks good, he's not uh on a tour proxy or anything like that.
00:05:15.439 --> 00:05:20.319
Uh and the distance between his address that he gave and that IP is five miles or whatever, you know.
00:05:20.720 --> 00:05:21.839
So happy bath.
00:05:21.920 --> 00:05:22.480
Everything looks good.
00:05:22.560 --> 00:05:29.199
And then there's probably other checks that the the bank or finance company is going to do, and they hopefully align with that, right?
00:05:30.000 --> 00:05:35.519
Now, on the fraud side or you know, malicious side, it's a very different look, right?
00:05:35.600 --> 00:05:39.519
It's like the email, quite often actually for lending email is created that day.
00:05:39.600 --> 00:05:40.079
Yeah, right.
00:05:40.240 --> 00:05:41.120
So the email is new.
00:05:41.199 --> 00:05:48.480
And if you think about it just practically, if you're applying for a loan as a genuine person, you're not going to create a new email and then immediately go apply for a loan.
00:05:48.639 --> 00:05:50.879
It's a it's a period of a telling point, right?
00:05:50.959 --> 00:06:13.920
But it doesn't end there because you know, some of our B2B lenders, for example, we'd see the um them replicating, we've seen replications of Citibank, uh, McDonald's, uh, all these other you know, companies that they make, they make a domain, they use disposable domains, but they make a new domain to look like the company and then they apply for commercial loans based and pretending to be that company.
00:06:14.480 --> 00:06:19.439
Uh and that's just the start of it, and then they try and do, but we can catch it so early.
00:06:19.680 --> 00:06:24.240
So we look at that, how the domain is constructed, when it was constructed, how active is that domain.
00:06:24.399 --> 00:06:35.279
We even have the ability because we have such visibility to all that activity, we start to look at if a domain comes in, because there is new domain, so it's not just because it's new and doesn't mean it's bad.
00:06:35.360 --> 00:06:36.399
Uh it might be new business.
00:06:36.720 --> 00:06:40.160
So we say, All right, what are the usernames uh that have we seen?
00:06:40.240 --> 00:06:43.519
Let's say we've 10,000 in the last uh month.
00:06:43.759 --> 00:06:46.079
How what was construction pattern in that username?
00:06:46.160 --> 00:06:48.160
So was it Dwayne 101, Dane Two?
00:06:48.319 --> 00:06:49.519
We see that a lot, you know.
00:06:49.600 --> 00:06:51.600
So the sequential side of it, and we can detect that too.
00:06:51.680 --> 00:06:54.879
And we start looking back and pattern analysis and stuff like that.
00:06:55.120 --> 00:07:05.519
So from a very one single point or you know, a couple of single points, we can determine quite a lot about what's happening and what's you know what's going to happen with that with that client.
00:07:05.839 --> 00:07:06.399
Absolutely.
00:07:06.560 --> 00:07:21.920
Well, so when when let's just imagine when one of those fraudsters then sets out to pass as a real customer, um what about that long-standing email address is really genuine and is probably the hardest to fake?
00:07:22.319 --> 00:07:26.800
Yeah, I mean, it's it's actually really rare nowadays because you have different solutions.
00:07:26.959 --> 00:07:36.240
We're not necessarily the only company that you know offers email data, but we are the only company to do it the way we do it and and how we and you know, with the tenure we have.
00:07:36.800 --> 00:07:41.920
So what you know from a from a fraudsters, you know, they've of course they'll try and replicate it.
00:07:42.000 --> 00:07:49.279
They might maybe create an email uh six months ago and then leave it kind of you know settle or even use it a little bit to make it look look genuine.
00:07:49.519 --> 00:07:50.480
But it's very hard.
00:07:50.560 --> 00:07:59.759
Like we we s we have such high traffic volumes that we have uh uh you know, you break break the email out from um a single data point to a timeline.
00:07:59.920 --> 00:08:03.600
So it because it is that's what all effectively profiles are, right?
00:08:03.680 --> 00:08:10.560
They're a big long timeline and they they should evolve gradually, some changes, maybe an address, maybe a location, an IP, that kind of thing.
00:08:10.879 --> 00:08:15.360
They shouldn't change suddenly and they shouldn't just appear out of nowhere and then start transacting.
00:08:15.439 --> 00:08:19.759
So there's a few few telling points like that, and that's what's so hard to replicate.
00:08:20.000 --> 00:08:23.839
You know, people look at like we get asked frequently, oh, do we use you know social data?
00:08:23.920 --> 00:08:33.039
And we don't, because social data, I myself I used to work in the very early days on Facebook, and I can tell you that um there is a lot of fake accounts, right?
00:08:33.360 --> 00:08:42.240
And it's especially in this world, you know, with AI uh and and script and you know things like that, it's not that hard to create those social profiles.
00:08:42.399 --> 00:09:02.480
So just because they have a social profile with an email, it's not always now more and more, maybe it gets a bit better because I was you know, uh, but we have a effectively a walled garden in terms of our our environment in turn in terms of our data, and we're very protective over that to make sure that it has that keeps that integrity, you know, because it can't be replicated.
00:09:02.559 --> 00:09:05.279
They can't insert, and that's effective what you're doing with social.
00:09:05.600 --> 00:09:11.360
You're inserting a new data point into a database to make it look legit, but they can't insert it into ours like that.
00:09:11.600 --> 00:09:17.600
Um, so it does, it m maintains its integrity, and we have ability to kind of filter out anything else.
00:09:17.840 --> 00:09:18.399
Absolutely.
00:09:18.639 --> 00:09:30.879
So I think especially now, you know, when you hear data points, when you hear email addresses, when you social data, um, everybody, of course, thinks of AI and sort of like the impact of what AI will have.
00:09:31.039 --> 00:09:36.480
Um, we're already seeing AI being able to spin up a convincing fake person, right?
00:09:36.559 --> 00:09:46.559
From the deep fake, fake documents, fake background stories, um, and just being able to create that and actually do damage by causing a lot of fraud with that, of course.
00:09:46.799 --> 00:09:51.440
So, what is the one thing that frauds that can still not really easily fake?
00:09:51.759 --> 00:09:52.559
History.
00:09:53.200 --> 00:09:55.919
History is the hardest, right?
00:09:56.000 --> 00:09:59.519
So you can backdate documents, so you can fake those, right?
00:09:59.600 --> 00:10:03.519
But then if you exist, it's it's it's hard if you're a new company to do that.
00:10:03.679 --> 00:10:08.639
But if you're in this withstanding company, you know, who's been there, we know they weren't around.
00:10:08.720 --> 00:10:14.159
This particular profile or this particular email in our case wasn't there two years ago, three years ago, whatever.
00:10:14.320 --> 00:10:24.960
So even though now, yes, all of this has appeared as a perfect looking profile with all the verifications, it's possible they they moved and you know, they now move to a country like we see that with you know immigration.
00:10:25.039 --> 00:10:26.799
We'll see like a whole new profile coming in.
00:10:26.879 --> 00:10:33.120
But again, we treat them initially as correctly, in my view, you know, as suspicious initially until they prove, you know, tenure.
00:10:33.360 --> 00:10:37.519
Um so we're still kind of insulated from that.
00:10:37.600 --> 00:10:48.559
Uh and and and in fact, maybe even become more important uh in that because uh the verification steps required in that in that become, you know, yeah, like the quality, as you say, is is amazing, uh some of the stuff.
00:10:48.720 --> 00:10:52.240
But we managed to kind of again keep that, keep our integrity, I guess, again.
00:10:52.879 --> 00:11:06.000
So I think, yeah, history sort of like this two-dimension thing where you you have probably static data, you have an email address, you have a name, you have other data points that you're collecting, but then history makes it another dimension of correct.
00:11:06.480 --> 00:11:06.639
Correct.
00:11:06.720 --> 00:11:06.879
Yeah.
00:11:07.039 --> 00:11:18.559
Do you think that's an advantage that will actually last, or is it just a matter of time before we'll have AI learning about that and creating history for those profiles, or just using AI for those?
00:11:18.879 --> 00:11:19.840
It's an interesting one.
00:11:20.000 --> 00:11:27.519
I mean, uh it everything there's a lot, like uh I've described it before as the this is the industrial revolution for the internet, right?
00:11:27.679 --> 00:11:34.320
AI, the agentic side of it, and it's going to fundamentally change quite a lot of things, but not everything, right?
00:11:34.399 --> 00:11:37.519
There's some fundamentals that just can't be changed, right?
00:11:37.600 --> 00:11:49.279
Which is one, for example, when you're going on to buy some sneakers, well, yeah, now your AI agent that's gonna do it, but the difference is you're still getting those shipped to your house.
00:11:49.440 --> 00:12:00.799
Okay, so now your shipping address is still important, and you're probably going to have like uh, you know, want to get the receipt for your purchase, even though the AI agent, and that's probably gonna go to your email.
00:12:00.960 --> 00:12:09.200
And you know, so there's a lot of that, you know, um, I was just talking Visa earlier, and they and talk about the agentic flow, and then our email was there and all this.
00:12:09.600 --> 00:12:10.960
I said, Well, does that go back to the merchant?
00:12:11.120 --> 00:12:12.240
Oh, it does, yeah, of course.
00:12:12.399 --> 00:12:20.000
So a lot of this, those core fundamentals will will kind of transcend through post AI and still will be there.
00:12:20.159 --> 00:12:29.039
So it's what it's changing is, you know, and I've tried to explain to people is that the difference with agentic commerce is not you know, you're not directly dealing with the merchant anymore.
00:12:29.200 --> 00:12:32.080
There's an in there's a there's a go-between bit and that wasn't there before.
00:12:32.159 --> 00:12:35.519
And that's fundamentally, you know, if you try and simplify it, that's the difference.
00:12:35.600 --> 00:12:38.879
Yeah, you have a you know a butler that's doing it for you now, you know.
00:12:39.120 --> 00:12:48.320
And so, and the likes of um, you know, what Experian are doing into that is is just verifying that connection, you know, and and stuff like that.
00:12:48.399 --> 00:12:51.039
So that's that's primarily, but I do, you know, I do to answer your question.
00:12:51.120 --> 00:12:55.440
I think we're gonna, you know, be as important or even more so going into the future.
00:12:55.519 --> 00:12:56.960
Um, because we're still expanding.
00:12:57.120 --> 00:12:59.039
We've more, we've more and more customers every year.
00:12:59.120 --> 00:13:03.039
And we've been going 26 years and we're processing billions, but it's getting bigger every year.
00:13:03.519 --> 00:13:06.080
So at the moment, there's no indication it's gonna change.
00:13:06.320 --> 00:13:06.639
Yeah.
00:13:06.879 --> 00:13:21.360
And of course, you know, you have plenty of companies who are already in that fraud space, like even though you're in the email space, but then applicing applying that to payments companies for fraud prevention, of course, and identity verification.
00:13:21.440 --> 00:13:24.639
Um, and we see, of course, a lot of acquisitions also in that space.
00:13:24.799 --> 00:13:29.360
You have MasterCard and LexisNexis acquiring different companies for that.
00:13:29.600 --> 00:13:36.480
So to really set you apart, like what does add data do that others don't or can't?
00:13:37.120 --> 00:13:41.840
Uh well, I what one of the things in fraud, like we're not a fraud platformer, right?
00:13:41.919 --> 00:13:53.120
We didn't we and and I intentionally didn't want to add like device capabilities and all the other, you know, stuff that does work is just that I wanted this product, it's it's very effective.
00:13:53.279 --> 00:13:59.360
And I actually used to be a customer because I used to when I was in fraud product for transunion, I used to use the data and I knew how valuable it was.
00:13:59.600 --> 00:14:02.399
So it was one of the big reasons why I wanted to come here.
00:14:02.799 --> 00:14:07.759
But uh it was very easy to use, it was very understandable, it was very easy to deploy.
00:14:07.919 --> 00:14:11.519
And it's like that's it's that's its strength, one of its strengths.
00:14:11.759 --> 00:14:17.360
The other thing as well is, you know, when we when we look at, you know, who do we compete with?
00:14:17.519 --> 00:14:18.480
Well, not many.
00:14:18.559 --> 00:14:27.440
I mean, there's others that have the data, but there's nobody that has our tenure, that has our, you know, um our quality, and um, again, just the easy deployment.
00:14:27.759 --> 00:14:34.240
So it's it's an in, you know, again, it's changing so much right now, and there's a lot of companies out there.
00:14:34.480 --> 00:14:42.080
Um, but we don't, you know, we we plan on expanding our our our reach um in in terms of our different countries and the data.
00:14:42.240 --> 00:14:48.000
We've been we have we've generally launched about two new features to the product every year since we've developed it.
00:14:48.159 --> 00:14:52.799
Um so we just plan on kind of enhancing it, make it more, let's say, AI suitable.
00:14:52.879 --> 00:14:58.240
And we've really already released three that have that direct uh connection uh features within the product.
00:14:58.320 --> 00:15:01.519
So and that's it, really, is just to make it more you know, more available.
00:15:01.840 --> 00:15:02.799
More available to others.
00:15:02.960 --> 00:15:03.919
Yeah, absolutely.
00:15:04.240 --> 00:15:10.000
So Experian uh is a name that most people probably associate with credit scores.
00:15:10.399 --> 00:15:16.320
Um but they acquired At Data, a 26-year-old email data company.
00:15:16.559 --> 00:15:18.799
Um what do they get out of the deal?
00:15:18.879 --> 00:15:22.240
Like how does that expand their reach or the other way around?
00:15:22.320 --> 00:15:28.639
How does that work for At Data and how can you leverage Experience um assets and what I mean you provide?
00:15:29.039 --> 00:15:32.559
We work we were quite excited about this deal for multiple reasons, right?
00:15:32.799 --> 00:15:36.639
Because one, they were actually a customer of ours for 18 years.
00:15:37.039 --> 00:15:41.519
So we we knew them so well already, and they knew the you know the value.
00:15:41.840 --> 00:15:51.360
And they knew what we had was unique because in the fraud industry, what you'll find when you talk to different companies is most of them are buying from each other.
00:15:51.759 --> 00:15:52.000
Right.
00:15:52.080 --> 00:15:56.240
And what they they their uniqueness is they their combination, whatever combination they end up with.
00:15:56.799 --> 00:15:58.159
But we were buying from no one.
00:15:58.320 --> 00:16:03.279
We owned entirely owned at the end of the entire data set, and that was one of the things that made us unique.
00:16:03.360 --> 00:16:11.120
And of course, from an experience point of view, because they are effectively one of the largest identity companies in the world, um, but they didn't have that.
00:16:11.360 --> 00:16:11.679
Interesting.
00:16:11.840 --> 00:16:12.000
Right.
00:16:12.080 --> 00:16:17.840
So they didn't have the email, they didn't have the history that we have, they didn't have the the kind of deployment ability we have.
00:16:18.080 --> 00:16:22.480
So for them it was it was a pretty, you know, yeah, pretty obvious fit.
00:16:22.639 --> 00:16:25.039
Uh for us, what do we get out of it?
00:16:25.200 --> 00:16:28.879
Um you know, simply their reach, right?
00:16:29.039 --> 00:16:32.879
They have, you know, I've I'm I'm still only kind of getting to grips.
00:16:32.960 --> 00:16:38.879
It's four months in, and now every finance, anything to do with finance, oh yeah, we're an experienced customer.
00:16:39.039 --> 00:16:41.039
Every single one of them, or you know, without fail.
00:16:41.200 --> 00:16:43.200
So now we go, well, do you want to use it?
00:16:43.279 --> 00:16:45.440
And they go, Yeah, well, we have a contract already, so it's easy.
00:16:45.600 --> 00:16:53.440
You know, so the the market just opened up to some of the biggest companies in the world, uh, to us, you know, so it is definitely mutually beneficial.
00:16:53.600 --> 00:16:59.600
And and what they like as well is that it's not a regional product, it's not something just for North America, it's not Southern Europe.
00:16:59.679 --> 00:17:03.360
It's we can deploy anywhere, like I said, and we have 190 countries running now.
00:17:03.600 --> 00:17:06.079
So um that's also a bit unique for them.
00:17:06.319 --> 00:17:06.799
Yeah, absolutely.
00:17:06.960 --> 00:17:10.480
It also looks to me like they have different data sets.
00:17:10.559 --> 00:17:19.519
Like I think you already mentioned, like because you have that Vault data set actually that you've been building with a lot of that historical data that's unique, that's your moat.
00:17:19.680 --> 00:17:21.279
And then you have all these other things.
00:17:21.440 --> 00:17:26.720
You have uh credit scores, you have insights into businesses information, all of these other things as well.
00:17:26.880 --> 00:17:38.079
Like, what are what is like a use case that you can share that that that a company has been benefiting from ad data, but also the experience data sets that actually allows them to do more than just the standalone service.
00:17:38.400 --> 00:17:49.839
Well, I suppose one of the examples, what we're going to do is like we're gonna be taking some of the experience, like their one of the things I love about them is that the data quality they have is exceptional, right?
00:17:49.920 --> 00:17:54.799
Like their procurement of their data, so their identity data, I mean, from my perspective, is what I'm looking at.
00:17:54.960 --> 00:17:59.359
So is so we're gonna take some of that and enrich the fraud product that we have, right?
00:17:59.519 --> 00:18:01.200
Some of the some of their global data sets.
00:18:01.359 --> 00:18:05.759
So from an identity perspective, we're going from good to amazing, is the goal, right?
00:18:06.240 --> 00:18:27.839
Um, so what that means um is we'll be able to tie, you know, like with stuff like credit header data, we'd tie that to the email, tie that to so it becomes incredibly accurate because all that data is really is high quality, but not only high quality, it's it's refreshed very regularly, updated, and quite accurate because it's the credit reporting system in the US, it's incredibly important.
00:18:28.000 --> 00:18:29.359
Most most countries it is.
00:18:29.759 --> 00:18:39.200
So we get the access and benefit of that, and that just makes um because our accuracy is generally pretty high, but now we're moving into a different league with um different data, you know, data points.
00:18:39.440 --> 00:18:49.039
And we are now moving into there, because um you may have seen in the press um Experiende launching a suite of uh agenc products.
00:18:49.119 --> 00:18:49.279
Yeah.
00:18:49.359 --> 00:18:54.400
So now the email now is being is will be integrated as part of those agenc products as well.
00:18:54.640 --> 00:18:57.680
So there's a lot of different things happening right now.
00:18:57.759 --> 00:19:07.279
Um come back to you in a year to say, you know what exactly how it's all worked out, but um, but there's a lot of opportunity in you know different regions and different, you know, for different companies.
00:19:07.599 --> 00:19:18.240
Um and we we have new opportunities like so um, you know, companies like Booking.com, which wouldn't traditionally be um an experienced customer.
00:19:18.319 --> 00:19:19.839
Yeah, you know, that's not that's out of side.
00:19:19.920 --> 00:19:21.200
So they're excited about that, right?
00:19:21.279 --> 00:19:24.640
Because we get to go in and have a big impact very fast and you know, and all that.
00:19:24.720 --> 00:19:40.000
Like, and we've like great companies like that, like DLocal and a few others that that we work with uh that would be traditionally outside of their wheelhouse, but we we've made a big impact with them uh in terms of their KYC and fraud prevention stuff in in you know, again, many different markets.
00:19:40.240 --> 00:19:43.599
So yeah, I think that's it's it it very much is mutual.
00:19:43.759 --> 00:19:49.039
I mean, we we're now we were a company 55 and they're a company of nearly 22,000 people.
00:19:49.200 --> 00:19:55.039
So we're not gonna change their world now necessarily, but we'll we'll probably punch above our weight in terms of what we can do for them, all right?
00:19:55.359 --> 00:19:55.920
Absolutely.
00:19:56.079 --> 00:20:04.400
So um I think more and more people, of course, are expecting AI assistants to do their shopping or their booking for them as well.
00:20:04.559 --> 00:20:12.480
Um, and when it's a bob placing that exact order, how does a business know if that's real or trusted person behind that?
00:20:12.640 --> 00:20:16.640
Is that does that link back to the email addresses again or is there a different way?
00:20:16.880 --> 00:20:18.000
Yeah, well, like that.
00:20:18.079 --> 00:20:21.759
I mean, on the experience side, that's why they've launched the agent trust stuff.
00:20:22.400 --> 00:20:26.960
Basically, that is um it it is I wish I was more part of it.
00:20:27.039 --> 00:20:38.799
I'm I'm getting into it now, but uh so that is where they are um bringing in uh the link between the uh agent to the basically the person who's ordering, right?
00:20:38.960 --> 00:20:53.920
And that means that they can uh assign an ID and that ID gets passed through um as you know, when through the flow, and then basically from the you know on the on the on the user customer side they can see that it has a an approved uh experience ID, right?
00:20:54.000 --> 00:20:55.599
So it's a trusted customer.
00:20:55.839 --> 00:20:58.880
So it's a really cool system, very efficient, very low touch.
00:20:59.039 --> 00:21:09.200
Um and but that happens because we're like we were saying earlier, is that there's there's a go between you know that wasn't there before and how do you link all three, because there's three parties, right?
00:21:09.279 --> 00:21:12.000
You link them all, how do you link them all together and trust that?
00:21:12.400 --> 00:21:20.240
The issue I do see at the moment there's very little traffic, to be honest, because most of the the um You mean agentic traffic.
00:21:20.799 --> 00:21:25.599
Yeah, exactly, true agentic traffic because most of it's been blocked by the website's malware and stuff like that.
00:21:25.680 --> 00:21:27.680
Like because traditionally speaking, bots are bad.
00:21:27.759 --> 00:21:29.680
Yeah, you know, stop, period, nothing, you know.
00:21:29.759 --> 00:21:32.240
So we didn't want any bot activity, but these are all bot activity.
00:21:32.319 --> 00:21:39.680
So customers they have to, or companies will have to change their view of that because it'll nearly 90% will be good after in a short while.
00:21:39.920 --> 00:21:42.480
And the challenge will be who's good and who's bad, you know.
00:21:42.559 --> 00:21:47.440
Um and it's that that is going to be an issue, you know, with some of the the new AI models that are being launched now.
00:21:47.759 --> 00:21:48.000
Okay.
00:21:48.480 --> 00:21:54.160
So I think you already touched upon it, but how much of that, how much of that traffic, right?
00:21:54.319 --> 00:21:57.440
How much of the transactions is already actively happening?
00:21:57.680 --> 00:22:05.119
Are we still looking at very small single Single digit or a couple of thousand transactions per day or per month.
00:22:05.680 --> 00:22:07.119
So it's is it growing?
00:22:07.440 --> 00:22:16.400
It is growing, uh, but like most of it's still being blocked because of the that that, but like so there's a difference between agentic commerce and then AI being used.
00:22:16.480 --> 00:22:18.400
And so AI being used is very different, right?
00:22:18.720 --> 00:22:19.440
It's prolific.
00:22:19.599 --> 00:22:21.920
We see, like I said, we've launched features because of this.
00:22:22.000 --> 00:22:27.279
So we've launched um uh gibberish detection, sequential email detection, and tumbling.
00:22:27.440 --> 00:22:33.440
And those three features are basically a direct result of uh massive script creation of emails, right?
00:22:33.519 --> 00:22:41.359
So we see, like I said, Dwayne 101, Dwayne 106, 106, and they goes up like thousands of them gets created, and then they get spread across different websites.
00:22:41.519 --> 00:22:45.839
So the AI side of it and the AI malicious use of it is not new.
00:22:46.079 --> 00:22:48.160
Now, this is this is commerce, so it's different.
00:22:48.240 --> 00:22:50.799
There's an agent and it's it's designed to the purchase of it.
00:22:50.960 --> 00:22:58.400
Uh yeah, we we haven't seen it, but like it, I would say it's months uh before that's gonna massively shift and shift quickly.
00:22:58.480 --> 00:23:12.559
So most merchants have to create a start start working now, basically, to be prepared, make sure that they have the right solutions in place, make sure that they have solutions that have the durability to withstand, you know, because a lot of them you take, let's say, device ID.
00:23:12.799 --> 00:23:22.480
Well, it has no relevance in an in in AI because the device ID wasn't that device, it wasn't your device, it was it was the with the exception of things like the comet browser.
00:23:22.559 --> 00:23:26.480
I don't know if you're from Perplexity, that that actually uses your computer to do it.
00:23:26.640 --> 00:23:28.480
So maybe that becomes popular, I'm not sure.
00:23:28.559 --> 00:23:29.839
We'll have to see how that evolves.
00:23:30.000 --> 00:23:48.559
But um so yeah, but I I think all of that infrastructure is being created right now, uh, in both both from the agentico side and actually the provider side, you're starting to see like no then you know conferences and stuff that's all these uh you know agentic-based kind of solutions are starting to merge.
00:23:48.640 --> 00:23:49.599
They weren't there last year.
00:23:49.839 --> 00:23:59.200
So that all that's kind of starting to come together, and as soon as they all come together, that's when you'll see an explosion in terms of the traffic because it is quite ready, I think, um, and people are ready for it.
00:23:59.519 --> 00:24:01.440
So just have to see how it evolves.
00:24:01.680 --> 00:24:03.039
See how that evolves.
00:24:03.359 --> 00:24:17.200
Um, I think another thing that I continue to see, but I think probably a lot of people are also noticing, is that you have some of the bigger tech companies who are providing a lot of email services, like an Apple with their iCloud services, Gmail, of course.
00:24:17.359 --> 00:24:27.920
Um, more and more are letting people giving people the option to hide their email addresses, you know, and privacy rules, of course, due to regulations, GDPR and other things that keep tightening, right?
00:24:28.240 --> 00:24:30.720
Much more uh is being done on that way.
00:24:30.960 --> 00:24:36.240
As more people are trying to amass their email, does that actually hurt your business?
00:24:36.720 --> 00:24:42.240
I mean, the so there's only one major one, really, and it's Apple who've done it.
00:24:42.319 --> 00:24:50.720
Gmail allow change, but actually it's quite a cumbersome thing in that they they only allow one change a year, and it's uh it means you'd have to change all your accounts, right?
00:24:50.880 --> 00:24:53.440
Your access accounts, and people aren't inclined to do that.
00:24:53.680 --> 00:24:58.880
Uh but Apple have done it differently where they they have um Hide My Email service, right?
00:24:59.039 --> 00:25:09.680
So predominantly more on apps than it is on the website, but uh for us it represents about 2% of our total traffic, so it's still relatively low, and growth rate is limited.
00:25:09.920 --> 00:25:18.240
Um so with that one, it's interesting because we actually see because it's it goes back to a website and it keeps using the same email.
00:25:18.319 --> 00:25:21.119
So we actually do have we see returning, even though they've highten it.
00:25:21.279 --> 00:25:22.319
We actually see the return rate.
00:25:22.480 --> 00:25:26.799
So instead of one email, a person has 40 because they're going to 40 different websites.
00:25:26.880 --> 00:25:30.000
So we see all of those 40 and potentially could merge them back, you know.
00:25:30.079 --> 00:25:36.400
But um, so no, um, you know, for now it hasn't had any impact from a compliance perspective.
00:25:36.640 --> 00:25:51.920
GDPR largely from a you know, GDPR has provisions for fraud prevention, you know, so it allows a lot more, you know, and rightly so, you know, to to for companies to do to have more access to data to perform fraud prevention checks.
00:25:52.160 --> 00:25:57.599
So we haven't had um like we do our GDPR compliance and doing one at the moment with Experian as well.
00:25:57.839 --> 00:26:01.599
So uh yeah, it hasn't been a big issue on that front either.
00:26:01.920 --> 00:26:02.079
Okay.
00:26:02.880 --> 00:26:17.039
So as more shopping is being done by those AI agents and more people are hiding those details, does email still stay at the center of proving who somebody is, or does it become just another signals among many?
00:26:17.200 --> 00:26:18.960
Uh so what do you think?
00:26:19.359 --> 00:26:22.960
I suppose it's uh it's always a signal among many, I guess.
00:26:23.039 --> 00:26:27.440
Uh it's a case of um in data science, like we have like a ranking, right?
00:26:27.599 --> 00:26:30.880
So so all your signals, what how how does it rank?
00:26:31.039 --> 00:26:37.519
And it is it is top, like it's starting top three, and it very, very rarely deviates from that.
00:26:37.680 --> 00:26:41.599
Um so do I see it having that into the future?
00:26:41.839 --> 00:26:42.799
Almost certainly.
00:26:43.039 --> 00:26:50.960
Um there is no, like I said earlier, there's no indication right now that it's going away tomorrow because it does there's nothing that replaces it necessarily.
00:26:51.119 --> 00:26:57.279
The shift with email actually has been that it's not used so much for communication anymore.
00:26:57.440 --> 00:26:57.759
Yeah.
00:26:57.920 --> 00:26:59.200
You know, that's been the shift.
00:26:59.279 --> 00:27:02.160
So it's it's it's use, but that doesn't necessarily matter to us.
00:27:02.319 --> 00:27:08.400
But the use of it as a as a literally a back and forth communication tool, and I mean I'm private, still used to certain degree in work.
00:27:08.480 --> 00:27:15.359
Um, but IMs and WhatsApp and stuff is is more predominant now, you know, where we would have sent an email and I used to pay someone and someone like that.
00:27:15.599 --> 00:27:20.000
But but it's use in terms of you know checkout and as a kind of the your address.
00:27:20.319 --> 00:27:28.880
Uh because again, it is that it is it is an ID slash address because if you're buying digital goods, your physical address doesn't matter anymore.
00:27:29.119 --> 00:27:34.799
So but you still need an address and it's something for those confirmations to go to, and that's email.
00:27:34.960 --> 00:27:39.119
And I now I don't see where that it would go other than that, you know.
00:27:39.920 --> 00:27:40.720
Absolutely.
00:27:40.960 --> 00:27:45.279
So I tend to be a very positive future-focused person.
00:27:45.440 --> 00:27:52.559
Um, but however, sometimes when you're building something, you're building a new company, there's always roadblocks uh ahead of it.
00:27:52.720 --> 00:27:56.559
Uh, what is something that you believe that needs to drastically change, right?
00:27:56.640 --> 00:28:12.880
It could be regulations, it could be technology itself, it could be just the way that companies are operating or maybe utilizing emails that you believe like if we can change that, we can actually get a lot more innovation, we can, you know, grow substantially, not just as a company, but overall as businesses.
00:28:13.119 --> 00:28:14.000
Yeah, yeah.
00:28:14.240 --> 00:28:18.799
Um I mean, like the rate, the rate of progress is phenomenal.
00:28:18.880 --> 00:28:26.880
And it's, you know, and right now my my issue has been is actually around with AI, is that it's that transparency.
00:28:26.960 --> 00:28:30.319
There's a lot of fear-mongering with it, uh, you know, it's particularly that.
00:28:30.480 --> 00:28:34.000
And a lot of people they're they're afraid of what they don't know.
00:28:34.319 --> 00:28:42.559
So it's AI, AI, AI, and it's gonna either, depending on your viewpoint, it's gonna destroy everything or create everything, you know, it depends on what you look at.
00:28:42.720 --> 00:28:52.559
So that transparency and that insight is is a big issue right now for for consumers, businesses, everything, because they're they're petrified, most of them, uh, as to what happens.
00:28:52.880 --> 00:29:04.079
So I don't have an answer for it, unfortunately, but like the uh is that transparency into like you know, if we were able to do it nicely and have a nice clean okay, over the next five years, this is what's going to happen.
00:29:04.240 --> 00:29:07.680
Everybody can kind of settle down and and and work towards that.
00:29:07.920 --> 00:29:12.240
Um, but the other issue I think is the confusion.
00:29:12.319 --> 00:29:27.599
For I like I'm on the I you know, okay, I build these products and stuff like that, but I do sympathize with the buyers, the the companies out there, because you go to conferences and there's thousands of companies, hundreds and thousands or whatever of companies, and for the most part, they all look the same.
00:29:27.680 --> 00:29:27.759
Yeah.
00:29:27.920 --> 00:29:30.559
And you're like, how if I was a how do I pick one of these?
00:29:30.640 --> 00:29:32.160
And how do I know I got the best one?
00:29:32.240 --> 00:29:40.160
How do I it's a massive challenge for for for companies to find a correct vendor for them.
00:29:40.319 --> 00:29:45.119
Um, and a lot of the time I think they're just sick of people, you know, pounding their door and stuff like that.
00:29:45.359 --> 00:29:53.680
Um so I think they're the two for me, if you could, if they were those two issues, you know, it would make things a life a lot easier.
00:29:53.839 --> 00:29:58.880
Um, you know, the transparency and and it I suppose they're they're both effectively about transparency.
00:29:59.039 --> 00:30:08.079
Um so yeah, um being able to filter out some of the vendors that are suitable for your industry and that noise, and then you know, transparency into the AI.
00:30:08.160 --> 00:30:11.440
Now, hopefully the transparency into AI is kind of self-fulfilling, right?
00:30:11.599 --> 00:30:14.240
That over time we'll start to get used to it, like we did with the internet.
00:30:14.559 --> 00:30:18.319
I mean, you could argue it was the same going back about 30 years now with the internet.
00:30:18.480 --> 00:30:22.799
We're like, oh my god, what now, you know, all the bricks and mortars, this the high street will never change.
00:30:22.880 --> 00:30:24.960
And yeah, it has, it has changed dramatically.
00:30:25.119 --> 00:30:26.240
So we're probably seeing it.
00:30:26.319 --> 00:30:32.960
I just hope that it doesn't happen too fast that people, you know, catch them unaware that they have time to prepare and and and and kind of get ready for it.
00:30:33.279 --> 00:30:33.839
Absolutely.
00:30:34.000 --> 00:30:36.880
Well, thank you very much for joining Executive Conversity Race.
00:30:37.039 --> 00:30:39.599
It was uh great having you on as again, and uh till next time.
00:30:39.759 --> 00:30:40.400
Yeah, thank you.
00:30:41.279 --> 00:30:45.440
If you like this episode, hit like, follow, and subscribe for more.
00:00:00.160 --> 00:00:02.720
Welcome to Executive Conversations by the Papers.
00:00:02.799 --> 00:00:04.000
I'm Dwayne Jeffrey.
00:00:04.160 --> 00:00:09.759
In this episode, I'm joined by Darmuk Thoma, Head of Fraud and Data Strategy at Ad Data.
00:00:09.839 --> 00:00:17.280
We're exploring how email and data intelligence is used to detect fraud earlier and the role data strategy plays in staying ahead of it.
00:00:17.920 --> 00:00:22.559
For someone meeting Ad Data for the first time, what does your company actually do?
00:00:23.199 --> 00:00:27.519
We're heavily focused, in fact, almost entirely focused on email, right?
00:00:27.679 --> 00:00:32.079
So what I primarily look after is the fraud prevention side of it.
00:00:34.079 --> 00:00:36.159
A lot of people don't even know the brand very well.
00:00:36.240 --> 00:00:40.799
And it's but it's actually quite a mature, it's a 26-year-old company.
00:00:41.520 --> 00:00:43.920
It's the largest email network in the world.
00:00:44.159 --> 00:00:53.439
Um so what that kind of translates to is we process just about two billion email interactions a month globally in about 190 countries.
00:00:53.759 --> 00:01:08.079
Um and that gives us insight into you know kind of what the email activity is happening, and that allows us to do things like fraud prevention, say how old the email, how good it is, what it's doing, um, and allows us to do a lot of stuff in terms of fraud prevention.
00:01:08.239 --> 00:01:10.000
So that's that's kind of it.
00:01:10.159 --> 00:01:16.799
We try intentionally try and keep very simple, keep it, you know, easy to use, easy to understand, and uh easy to deploy, of course.
00:01:17.040 --> 00:01:17.519
Absolutely.
00:01:17.599 --> 00:01:24.879
And then when you're talking about financial companies, fintech companies, payments companies, what are they utilizing uh ad data for?
00:01:25.200 --> 00:01:29.519
Well, it's primarily kind of type KYC type things, right?
00:01:29.680 --> 00:01:33.519
So uh in in finance, there's a lot of different checks that have to happen.
00:01:33.599 --> 00:01:36.879
You know, there's credit checks, KYC checks, uh AML checks, and so on.
00:01:37.040 --> 00:01:38.959
Quite a lot of them are quite expensive, right?
00:01:39.120 --> 00:01:42.640
So we're not, you know, we're quite easy to deploy, quite cost efficient.
00:01:42.719 --> 00:01:47.599
So a lot of the stuff is used very early in the funnel uh for fintechs and and finance companies.
00:01:47.760 --> 00:02:00.799
So we're you know, literally on the point of, you know, as soon as they're creating an account with the with the company, we're able to tell them, yeah, that in the the the name associated with the email that you got, that's that's correct.
00:02:01.040 --> 00:02:14.159
The email is you know 20 years old and it's quite active, it's operating lots of different locations, and uh the IP is quite close to the address that you're so we have a lot of insight very early in the funnel, and that gives them confidence of the customer.
00:02:14.240 --> 00:02:27.280
Um, so you'd be surprised, like I've seen like certain like 20% of what the fintech could get could be just rubbish, just you know, nuisance, either intentionally fraud or just just really um, you know, it could be bought uh created accounts and stuff like that.
00:02:27.360 --> 00:02:32.080
So they're able to get rid of a lot of that very early in the funnel for relatively minimal cost.
00:02:32.400 --> 00:02:32.960
Absolutely.
00:02:33.039 --> 00:02:40.719
So there's of course a lot of different signals out there, but most of us treat our email address as you know, just something more mundane.
00:02:40.879 --> 00:02:42.639
We even barely even think about it.
00:02:42.800 --> 00:02:49.120
But your company tends to treat it as it's one of the most revealing signals of somebody who's really online.
00:02:49.919 --> 00:02:52.960
So, what does an email address actually tell you?
00:02:53.280 --> 00:02:58.879
Well, I I mean why data is important is effectively why is because email is so important, right?
00:02:59.120 --> 00:03:05.840
Because uh I've had I've had lots of different emails, but my primary email I've had since I was about 16 and I've kept it with me.
00:03:05.919 --> 00:03:13.759
But I've moved I've changed devices, of course, every couple of years, moved addresses, I've you know, uh uh people even change names, but they'll get their email.
00:03:13.919 --> 00:03:14.560
Yeah, right.
00:03:14.719 --> 00:03:21.280
So in a digital world, it it really is the crux of your your digital identity.
00:03:21.439 --> 00:03:27.280
In fact, most of it's you're surprised a lot of our uh clients refer to it as the email ID, not the email address.
00:03:27.680 --> 00:03:31.120
And that's a bit telling, I think, you know, because there's like it's it's so important.
00:03:31.280 --> 00:03:35.599
A lot of the time we'll see on account opening, you'll see uh email.
00:03:35.680 --> 00:03:35.840
Yeah.
00:03:36.000 --> 00:03:41.120
Nothing else, not your name, not your just enter your email and whatever you know, other data point they need.
00:03:41.199 --> 00:03:45.039
But that's it, just to keep the it it to a minimum and then they can follow up from there.
00:03:45.280 --> 00:03:56.639
So it's so important and it's so you know, has a such a long tenure uh within that that evolution of a profile and that it it makes it valuable from a you know customer point of view, and well, evaluation standpoint.
00:03:56.960 --> 00:03:57.520
Absolutely.
00:03:57.759 --> 00:03:58.080
Okay.
00:03:58.560 --> 00:04:02.080
So let's bring that into sort of like the real world, right?
00:04:02.159 --> 00:04:03.599
Like I want to imagine that.
00:04:03.759 --> 00:04:12.159
So imagine I'm an online lender, I'm approving loans, um, and a real share of that application, of course, comes from real people.
00:04:12.319 --> 00:04:21.839
Um, but without getting too technical, how do you know something about an email address that helps me tell who real customers are from the rest of them?
00:04:22.319 --> 00:04:25.360
So I suppose there's two scenarios, like the the good, the good path, right?
00:04:25.519 --> 00:04:31.360
The happy path, which is we hope most of your your client base, 80% hopefully, is is is good and or plus.
00:04:31.600 --> 00:04:35.839
Uh and what what when they come in and we get the email, it's all real time.
00:04:36.000 --> 00:04:39.199
Like so our environment, it's like uh it's almost like the Visa Credit Guard network.
00:04:39.279 --> 00:04:41.120
It's emails constantly interacting with us.
00:04:41.279 --> 00:04:44.639
We'll see, you know, we see billions obviously coming back, but we see all the interactions.
00:04:44.959 --> 00:04:49.279
So when a when a client queries us, we're querying it in right that second, right?
00:04:49.519 --> 00:04:51.040
What's happening with that email right now?
00:04:51.199 --> 00:04:52.720
So we give that context back.
00:04:52.800 --> 00:04:56.240
So it's all really fresh, really high high context data.
00:04:56.399 --> 00:05:01.120
But the what we're gonna say in um in the good flow is yeah, Dwayne.
00:05:01.199 --> 00:05:04.399
He we saw him first time for uh 16 years ago.
00:05:04.560 --> 00:05:10.240
He's uh the address that's the one we normally see, the name, that's the one we normally see with this email.
00:05:10.480 --> 00:05:15.360
And you know, IP looks good, he's not uh on a tour proxy or anything like that.
00:05:15.439 --> 00:05:20.319
Uh and the distance between his address that he gave and that IP is five miles or whatever, you know.
00:05:20.720 --> 00:05:21.839
So happy bath.
00:05:21.920 --> 00:05:22.480
Everything looks good.
00:05:22.560 --> 00:05:29.199
And then there's probably other checks that the the bank or finance company is going to do, and they hopefully align with that, right?
00:05:30.000 --> 00:05:35.519
Now, on the fraud side or you know, malicious side, it's a very different look, right?
00:05:35.600 --> 00:05:39.519
It's like the email, quite often actually for lending email is created that day.
00:05:39.600 --> 00:05:40.079
Yeah, right.
00:05:40.240 --> 00:05:41.120
So the email is new.
00:05:41.199 --> 00:05:48.480
And if you think about it just practically, if you're applying for a loan as a genuine person, you're not going to create a new email and then immediately go apply for a loan.
00:05:48.639 --> 00:05:50.879
It's a it's a period of a telling point, right?
00:05:50.959 --> 00:06:13.920
But it doesn't end there because you know, some of our B2B lenders, for example, we'd see the um them replicating, we've seen replications of Citibank, uh, McDonald's, uh, all these other you know, companies that they make, they make a domain, they use disposable domains, but they make a new domain to look like the company and then they apply for commercial loans based and pretending to be that company.
00:06:14.480 --> 00:06:19.439
Uh and that's just the start of it, and then they try and do, but we can catch it so early.
00:06:19.680 --> 00:06:24.240
So we look at that, how the domain is constructed, when it was constructed, how active is that domain.
00:06:24.399 --> 00:06:35.279
We even have the ability because we have such visibility to all that activity, we start to look at if a domain comes in, because there is new domain, so it's not just because it's new and doesn't mean it's bad.
00:06:35.360 --> 00:06:36.399
Uh it might be new business.
00:06:36.720 --> 00:06:40.160
So we say, All right, what are the usernames uh that have we seen?
00:06:40.240 --> 00:06:43.519
Let's say we've 10,000 in the last uh month.
00:06:43.759 --> 00:06:46.079
How what was construction pattern in that username?
00:06:46.160 --> 00:06:48.160
So was it Dwayne 101, Dane Two?
00:06:48.319 --> 00:06:49.519
We see that a lot, you know.
00:06:49.600 --> 00:06:51.600
So the sequential side of it, and we can detect that too.
00:06:51.680 --> 00:06:54.879
And we start looking back and pattern analysis and stuff like that.
00:06:55.120 --> 00:07:05.519
So from a very one single point or you know, a couple of single points, we can determine quite a lot about what's happening and what's you know what's going to happen with that with that client.
00:07:05.839 --> 00:07:06.399
Absolutely.
00:07:06.560 --> 00:07:21.920
Well, so when when let's just imagine when one of those fraudsters then sets out to pass as a real customer, um what about that long-standing email address is really genuine and is probably the hardest to fake?
00:07:22.319 --> 00:07:26.800
Yeah, I mean, it's it's actually really rare nowadays because you have different solutions.
00:07:26.959 --> 00:07:36.240
We're not necessarily the only company that you know offers email data, but we are the only company to do it the way we do it and and how we and you know, with the tenure we have.
00:07:36.800 --> 00:07:41.920
So what you know from a from a fraudsters, you know, they've of course they'll try and replicate it.
00:07:42.000 --> 00:07:49.279
They might maybe create an email uh six months ago and then leave it kind of you know settle or even use it a little bit to make it look look genuine.
00:07:49.519 --> 00:07:50.480
But it's very hard.
00:07:50.560 --> 00:07:59.759
Like we we s we have such high traffic volumes that we have uh uh you know, you break break the email out from um a single data point to a timeline.
00:07:59.920 --> 00:08:03.600
So it because it is that's what all effectively profiles are, right?
00:08:03.680 --> 00:08:10.560
They're a big long timeline and they they should evolve gradually, some changes, maybe an address, maybe a location, an IP, that kind of thing.
00:08:10.879 --> 00:08:15.360
They shouldn't change suddenly and they shouldn't just appear out of nowhere and then start transacting.
00:08:15.439 --> 00:08:19.759
So there's a few few telling points like that, and that's what's so hard to replicate.
00:08:20.000 --> 00:08:23.839
You know, people look at like we get asked frequently, oh, do we use you know social data?
00:08:23.920 --> 00:08:33.039
And we don't, because social data, I myself I used to work in the very early days on Facebook, and I can tell you that um there is a lot of fake accounts, right?
00:08:33.360 --> 00:08:42.240
And it's especially in this world, you know, with AI uh and and script and you know things like that, it's not that hard to create those social profiles.
00:08:42.399 --> 00:09:02.480
So just because they have a social profile with an email, it's not always now more and more, maybe it gets a bit better because I was you know, uh, but we have a effectively a walled garden in terms of our our environment in turn in terms of our data, and we're very protective over that to make sure that it has that keeps that integrity, you know, because it can't be replicated.
00:09:02.559 --> 00:09:05.279
They can't insert, and that's effective what you're doing with social.
00:09:05.600 --> 00:09:11.360
You're inserting a new data point into a database to make it look legit, but they can't insert it into ours like that.
00:09:11.600 --> 00:09:17.600
Um, so it does, it m maintains its integrity, and we have ability to kind of filter out anything else.
00:09:17.840 --> 00:09:18.399
Absolutely.
00:09:18.639 --> 00:09:30.879
So I think especially now, you know, when you hear data points, when you hear email addresses, when you social data, um, everybody, of course, thinks of AI and sort of like the impact of what AI will have.
00:09:31.039 --> 00:09:36.480
Um, we're already seeing AI being able to spin up a convincing fake person, right?
00:09:36.559 --> 00:09:46.559
From the deep fake, fake documents, fake background stories, um, and just being able to create that and actually do damage by causing a lot of fraud with that, of course.
00:09:46.799 --> 00:09:51.440
So, what is the one thing that frauds that can still not really easily fake?
00:09:51.759 --> 00:09:52.559
History.
00:09:53.200 --> 00:09:55.919
History is the hardest, right?
00:09:56.000 --> 00:09:59.519
So you can backdate documents, so you can fake those, right?
00:09:59.600 --> 00:10:03.519
But then if you exist, it's it's it's hard if you're a new company to do that.
00:10:03.679 --> 00:10:08.639
But if you're in this withstanding company, you know, who's been there, we know they weren't around.
00:10:08.720 --> 00:10:14.159
This particular profile or this particular email in our case wasn't there two years ago, three years ago, whatever.
00:10:14.320 --> 00:10:24.960
So even though now, yes, all of this has appeared as a perfect looking profile with all the verifications, it's possible they they moved and you know, they now move to a country like we see that with you know immigration.
00:10:25.039 --> 00:10:26.799
We'll see like a whole new profile coming in.
00:10:26.879 --> 00:10:33.120
But again, we treat them initially as correctly, in my view, you know, as suspicious initially until they prove, you know, tenure.
00:10:33.360 --> 00:10:37.519
Um so we're still kind of insulated from that.
00:10:37.600 --> 00:10:48.559
Uh and and and in fact, maybe even become more important uh in that because uh the verification steps required in that in that become, you know, yeah, like the quality, as you say, is is amazing, uh some of the stuff.
00:10:48.720 --> 00:10:52.240
But we managed to kind of again keep that, keep our integrity, I guess, again.
00:10:52.879 --> 00:11:06.000
So I think, yeah, history sort of like this two-dimension thing where you you have probably static data, you have an email address, you have a name, you have other data points that you're collecting, but then history makes it another dimension of correct.
00:11:06.480 --> 00:11:06.639
Correct.
00:11:06.720 --> 00:11:06.879
Yeah.
00:11:07.039 --> 00:11:18.559
Do you think that's an advantage that will actually last, or is it just a matter of time before we'll have AI learning about that and creating history for those profiles, or just using AI for those?
00:11:18.879 --> 00:11:19.840
It's an interesting one.
00:11:20.000 --> 00:11:27.519
I mean, uh it everything there's a lot, like uh I've described it before as the this is the industrial revolution for the internet, right?
00:11:27.679 --> 00:11:34.320
AI, the agentic side of it, and it's going to fundamentally change quite a lot of things, but not everything, right?
00:11:34.399 --> 00:11:37.519
There's some fundamentals that just can't be changed, right?
00:11:37.600 --> 00:11:49.279
Which is one, for example, when you're going on to buy some sneakers, well, yeah, now your AI agent that's gonna do it, but the difference is you're still getting those shipped to your house.
00:11:49.440 --> 00:12:00.799
Okay, so now your shipping address is still important, and you're probably going to have like uh, you know, want to get the receipt for your purchase, even though the AI agent, and that's probably gonna go to your email.
00:12:00.960 --> 00:12:09.200
And you know, so there's a lot of that, you know, um, I was just talking Visa earlier, and they and talk about the agentic flow, and then our email was there and all this.
00:12:09.600 --> 00:12:10.960
I said, Well, does that go back to the merchant?
00:12:11.120 --> 00:12:12.240
Oh, it does, yeah, of course.
00:12:12.399 --> 00:12:20.000
So a lot of this, those core fundamentals will will kind of transcend through post AI and still will be there.
00:12:20.159 --> 00:12:29.039
So it's what it's changing is, you know, and I've tried to explain to people is that the difference with agentic commerce is not you know, you're not directly dealing with the merchant anymore.
00:12:29.200 --> 00:12:32.080
There's an in there's a there's a go-between bit and that wasn't there before.
00:12:32.159 --> 00:12:35.519
And that's fundamentally, you know, if you try and simplify it, that's the difference.
00:12:35.600 --> 00:12:38.879
Yeah, you have a you know a butler that's doing it for you now, you know.
00:12:39.120 --> 00:12:48.320
And so, and the likes of um, you know, what Experian are doing into that is is just verifying that connection, you know, and and stuff like that.
00:12:48.399 --> 00:12:51.039
So that's that's primarily, but I do, you know, I do to answer your question.
00:12:51.120 --> 00:12:55.440
I think we're gonna, you know, be as important or even more so going into the future.
00:12:55.519 --> 00:12:56.960
Um, because we're still expanding.
00:12:57.120 --> 00:12:59.039
We've more, we've more and more customers every year.
00:12:59.120 --> 00:13:03.039
And we've been going 26 years and we're processing billions, but it's getting bigger every year.
00:13:03.519 --> 00:13:06.080
So at the moment, there's no indication it's gonna change.
00:13:06.320 --> 00:13:06.639
Yeah.
00:13:06.879 --> 00:13:21.360
And of course, you know, you have plenty of companies who are already in that fraud space, like even though you're in the email space, but then applicing applying that to payments companies for fraud prevention, of course, and identity verification.
00:13:21.440 --> 00:13:24.639
Um, and we see, of course, a lot of acquisitions also in that space.
00:13:24.799 --> 00:13:29.360
You have MasterCard and LexisNexis acquiring different companies for that.
00:13:29.600 --> 00:13:36.480
So to really set you apart, like what does add data do that others don't or can't?
00:13:37.120 --> 00:13:41.840
Uh well, I what one of the things in fraud, like we're not a fraud platformer, right?
00:13:41.919 --> 00:13:53.120
We didn't we and and I intentionally didn't want to add like device capabilities and all the other, you know, stuff that does work is just that I wanted this product, it's it's very effective.
00:13:53.279 --> 00:13:59.360
And I actually used to be a customer because I used to when I was in fraud product for transunion, I used to use the data and I knew how valuable it was.
00:13:59.600 --> 00:14:02.399
So it was one of the big reasons why I wanted to come here.
00:14:02.799 --> 00:14:07.759
But uh it was very easy to use, it was very understandable, it was very easy to deploy.
00:14:07.919 --> 00:14:11.519
And it's like that's it's that's its strength, one of its strengths.
00:14:11.759 --> 00:14:17.360
The other thing as well is, you know, when we when we look at, you know, who do we compete with?
00:14:17.519 --> 00:14:18.480
Well, not many.
00:14:18.559 --> 00:14:27.440
I mean, there's others that have the data, but there's nobody that has our tenure, that has our, you know, um our quality, and um, again, just the easy deployment.
00:14:27.759 --> 00:14:34.240
So it's it's an in, you know, again, it's changing so much right now, and there's a lot of companies out there.
00:14:34.480 --> 00:14:42.080
Um, but we don't, you know, we we plan on expanding our our our reach um in in terms of our different countries and the data.
00:14:42.240 --> 00:14:48.000
We've been we have we've generally launched about two new features to the product every year since we've developed it.
00:14:48.159 --> 00:14:52.799
Um so we just plan on kind of enhancing it, make it more, let's say, AI suitable.
00:14:52.879 --> 00:14:58.240
And we've really already released three that have that direct uh connection uh features within the product.
00:14:58.320 --> 00:15:01.519
So and that's it, really, is just to make it more you know, more available.
00:15:01.840 --> 00:15:02.799
More available to others.
00:15:02.960 --> 00:15:03.919
Yeah, absolutely.
00:15:04.240 --> 00:15:10.000
So Experian uh is a name that most people probably associate with credit scores.
00:15:10.399 --> 00:15:16.320
Um but they acquired At Data, a 26-year-old email data company.
00:15:16.559 --> 00:15:18.799
Um what do they get out of the deal?
00:15:18.879 --> 00:15:22.240
Like how does that expand their reach or the other way around?
00:15:22.320 --> 00:15:28.639
How does that work for At Data and how can you leverage Experience um assets and what I mean you provide?
00:15:29.039 --> 00:15:32.559
We work we were quite excited about this deal for multiple reasons, right?
00:15:32.799 --> 00:15:36.639
Because one, they were actually a customer of ours for 18 years.
00:15:37.039 --> 00:15:41.519
So we we knew them so well already, and they knew the you know the value.
00:15:41.840 --> 00:15:51.360
And they knew what we had was unique because in the fraud industry, what you'll find when you talk to different companies is most of them are buying from each other.
00:15:51.759 --> 00:15:52.000
Right.
00:15:52.080 --> 00:15:56.240
And what they they their uniqueness is they their combination, whatever combination they end up with.
00:15:56.799 --> 00:15:58.159
But we were buying from no one.
00:15:58.320 --> 00:16:03.279
We owned entirely owned at the end of the entire data set, and that was one of the things that made us unique.
00:16:03.360 --> 00:16:11.120
And of course, from an experience point of view, because they are effectively one of the largest identity companies in the world, um, but they didn't have that.
00:16:11.360 --> 00:16:11.679
Interesting.
00:16:11.840 --> 00:16:12.000
Right.
00:16:12.080 --> 00:16:17.840
So they didn't have the email, they didn't have the history that we have, they didn't have the the kind of deployment ability we have.
00:16:18.080 --> 00:16:22.480
So for them it was it was a pretty, you know, yeah, pretty obvious fit.
00:16:22.639 --> 00:16:25.039
Uh for us, what do we get out of it?
00:16:25.200 --> 00:16:28.879
Um you know, simply their reach, right?
00:16:29.039 --> 00:16:32.879
They have, you know, I've I'm I'm still only kind of getting to grips.
00:16:32.960 --> 00:16:38.879
It's four months in, and now every finance, anything to do with finance, oh yeah, we're an experienced customer.
00:16:39.039 --> 00:16:41.039
Every single one of them, or you know, without fail.
00:16:41.200 --> 00:16:43.200
So now we go, well, do you want to use it?
00:16:43.279 --> 00:16:45.440
And they go, Yeah, well, we have a contract already, so it's easy.
00:16:45.600 --> 00:16:53.440
You know, so the the market just opened up to some of the biggest companies in the world, uh, to us, you know, so it is definitely mutually beneficial.
00:16:53.600 --> 00:16:59.600
And and what they like as well is that it's not a regional product, it's not something just for North America, it's not Southern Europe.
00:16:59.679 --> 00:17:03.360
It's we can deploy anywhere, like I said, and we have 190 countries running now.
00:17:03.600 --> 00:17:06.079
So um that's also a bit unique for them.
00:17:06.319 --> 00:17:06.799
Yeah, absolutely.
00:17:06.960 --> 00:17:10.480
It also looks to me like they have different data sets.
00:17:10.559 --> 00:17:19.519
Like I think you already mentioned, like because you have that Vault data set actually that you've been building with a lot of that historical data that's unique, that's your moat.
00:17:19.680 --> 00:17:21.279
And then you have all these other things.
00:17:21.440 --> 00:17:26.720
You have uh credit scores, you have insights into businesses information, all of these other things as well.
00:17:26.880 --> 00:17:38.079
Like, what are what is like a use case that you can share that that that a company has been benefiting from ad data, but also the experience data sets that actually allows them to do more than just the standalone service.
00:17:38.400 --> 00:17:49.839
Well, I suppose one of the examples, what we're going to do is like we're gonna be taking some of the experience, like their one of the things I love about them is that the data quality they have is exceptional, right?
00:17:49.920 --> 00:17:54.799
Like their procurement of their data, so their identity data, I mean, from my perspective, is what I'm looking at.
00:17:54.960 --> 00:17:59.359
So is so we're gonna take some of that and enrich the fraud product that we have, right?
00:17:59.519 --> 00:18:01.200
Some of the some of their global data sets.
00:18:01.359 --> 00:18:05.759
So from an identity perspective, we're going from good to amazing, is the goal, right?
00:18:06.240 --> 00:18:27.839
Um, so what that means um is we'll be able to tie, you know, like with stuff like credit header data, we'd tie that to the email, tie that to so it becomes incredibly accurate because all that data is really is high quality, but not only high quality, it's it's refreshed very regularly, updated, and quite accurate because it's the credit reporting system in the US, it's incredibly important.
00:18:28.000 --> 00:18:29.359
Most most countries it is.
00:18:29.759 --> 00:18:39.200
So we get the access and benefit of that, and that just makes um because our accuracy is generally pretty high, but now we're moving into a different league with um different data, you know, data points.
00:18:39.440 --> 00:18:49.039
And we are now moving into there, because um you may have seen in the press um Experiende launching a suite of uh agenc products.
00:18:49.119 --> 00:18:49.279
Yeah.
00:18:49.359 --> 00:18:54.400
So now the email now is being is will be integrated as part of those agenc products as well.
00:18:54.640 --> 00:18:57.680
So there's a lot of different things happening right now.
00:18:57.759 --> 00:19:07.279
Um come back to you in a year to say, you know what exactly how it's all worked out, but um, but there's a lot of opportunity in you know different regions and different, you know, for different companies.
00:19:07.599 --> 00:19:18.240
Um and we we have new opportunities like so um, you know, companies like Booking.com, which wouldn't traditionally be um an experienced customer.
00:19:18.319 --> 00:19:19.839
Yeah, you know, that's not that's out of side.
00:19:19.920 --> 00:19:21.200
So they're excited about that, right?
00:19:21.279 --> 00:19:24.640
Because we get to go in and have a big impact very fast and you know, and all that.
00:19:24.720 --> 00:19:40.000
Like, and we've like great companies like that, like DLocal and a few others that that we work with uh that would be traditionally outside of their wheelhouse, but we we've made a big impact with them uh in terms of their KYC and fraud prevention stuff in in you know, again, many different markets.
00:19:40.240 --> 00:19:43.599
So yeah, I think that's it's it it very much is mutual.
00:19:43.759 --> 00:19:49.039
I mean, we we're now we were a company 55 and they're a company of nearly 22,000 people.
00:19:49.200 --> 00:19:55.039
So we're not gonna change their world now necessarily, but we'll we'll probably punch above our weight in terms of what we can do for them, all right?
00:19:55.359 --> 00:19:55.920
Absolutely.
00:19:56.079 --> 00:20:04.400
So um I think more and more people, of course, are expecting AI assistants to do their shopping or their booking for them as well.
00:20:04.559 --> 00:20:12.480
Um, and when it's a bob placing that exact order, how does a business know if that's real or trusted person behind that?
00:20:12.640 --> 00:20:16.640
Is that does that link back to the email addresses again or is there a different way?
00:20:16.880 --> 00:20:18.000
Yeah, well, like that.
00:20:18.079 --> 00:20:21.759
I mean, on the experience side, that's why they've launched the agent trust stuff.
00:20:22.400 --> 00:20:26.960
Basically, that is um it it is I wish I was more part of it.
00:20:27.039 --> 00:20:38.799
I'm I'm getting into it now, but uh so that is where they are um bringing in uh the link between the uh agent to the basically the person who's ordering, right?
00:20:38.960 --> 00:20:53.920
And that means that they can uh assign an ID and that ID gets passed through um as you know, when through the flow, and then basically from the you know on the on the on the user customer side they can see that it has a an approved uh experience ID, right?
00:20:54.000 --> 00:20:55.599
So it's a trusted customer.
00:20:55.839 --> 00:20:58.880
So it's a really cool system, very efficient, very low touch.
00:20:59.039 --> 00:21:09.200
Um and but that happens because we're like we were saying earlier, is that there's there's a go between you know that wasn't there before and how do you link all three, because there's three parties, right?
00:21:09.279 --> 00:21:12.000
You link them all, how do you link them all together and trust that?
00:21:12.400 --> 00:21:20.240
The issue I do see at the moment there's very little traffic, to be honest, because most of the the um You mean agentic traffic.
00:21:20.799 --> 00:21:25.599
Yeah, exactly, true agentic traffic because most of it's been blocked by the website's malware and stuff like that.
00:21:25.680 --> 00:21:27.680
Like because traditionally speaking, bots are bad.
00:21:27.759 --> 00:21:29.680
Yeah, you know, stop, period, nothing, you know.
00:21:29.759 --> 00:21:32.240
So we didn't want any bot activity, but these are all bot activity.
00:21:32.319 --> 00:21:39.680
So customers they have to, or companies will have to change their view of that because it'll nearly 90% will be good after in a short while.
00:21:39.920 --> 00:21:42.480
And the challenge will be who's good and who's bad, you know.
00:21:42.559 --> 00:21:47.440
Um and it's that that is going to be an issue, you know, with some of the the new AI models that are being launched now.
00:21:47.759 --> 00:21:48.000
Okay.
00:21:48.480 --> 00:21:54.160
So I think you already touched upon it, but how much of that, how much of that traffic, right?
00:21:54.319 --> 00:21:57.440
How much of the transactions is already actively happening?
00:21:57.680 --> 00:22:05.119
Are we still looking at very small single Single digit or a couple of thousand transactions per day or per month.
00:22:05.680 --> 00:22:07.119
So it's is it growing?
00:22:07.440 --> 00:22:16.400
It is growing, uh, but like most of it's still being blocked because of the that that, but like so there's a difference between agentic commerce and then AI being used.
00:22:16.480 --> 00:22:18.400
And so AI being used is very different, right?
00:22:18.720 --> 00:22:19.440
It's prolific.
00:22:19.599 --> 00:22:21.920
We see, like I said, we've launched features because of this.
00:22:22.000 --> 00:22:27.279
So we've launched um uh gibberish detection, sequential email detection, and tumbling.
00:22:27.440 --> 00:22:33.440
And those three features are basically a direct result of uh massive script creation of emails, right?
00:22:33.519 --> 00:22:41.359
So we see, like I said, Dwayne 101, Dwayne 106, 106, and they goes up like thousands of them gets created, and then they get spread across different websites.
00:22:41.519 --> 00:22:45.839
So the AI side of it and the AI malicious use of it is not new.
00:22:46.079 --> 00:22:48.160
Now, this is this is commerce, so it's different.
00:22:48.240 --> 00:22:50.799
There's an agent and it's it's designed to the purchase of it.
00:22:50.960 --> 00:22:58.400
Uh yeah, we we haven't seen it, but like it, I would say it's months uh before that's gonna massively shift and shift quickly.
00:22:58.480 --> 00:23:12.559
So most merchants have to create a start start working now, basically, to be prepared, make sure that they have the right solutions in place, make sure that they have solutions that have the durability to withstand, you know, because a lot of them you take, let's say, device ID.
00:23:12.799 --> 00:23:22.480
Well, it has no relevance in an in in AI because the device ID wasn't that device, it wasn't your device, it was it was the with the exception of things like the comet browser.
00:23:22.559 --> 00:23:26.480
I don't know if you're from Perplexity, that that actually uses your computer to do it.
00:23:26.640 --> 00:23:28.480
So maybe that becomes popular, I'm not sure.
00:23:28.559 --> 00:23:29.839
We'll have to see how that evolves.
00:23:30.000 --> 00:23:48.559
But um so yeah, but I I think all of that infrastructure is being created right now, uh, in both both from the agentico side and actually the provider side, you're starting to see like no then you know conferences and stuff that's all these uh you know agentic-based kind of solutions are starting to merge.
00:23:48.640 --> 00:23:49.599
They weren't there last year.
00:23:49.839 --> 00:23:59.200
So that all that's kind of starting to come together, and as soon as they all come together, that's when you'll see an explosion in terms of the traffic because it is quite ready, I think, um, and people are ready for it.
00:23:59.519 --> 00:24:01.440
So just have to see how it evolves.
00:24:01.680 --> 00:24:03.039
See how that evolves.
00:24:03.359 --> 00:24:17.200
Um, I think another thing that I continue to see, but I think probably a lot of people are also noticing, is that you have some of the bigger tech companies who are providing a lot of email services, like an Apple with their iCloud services, Gmail, of course.
00:24:17.359 --> 00:24:27.920
Um, more and more are letting people giving people the option to hide their email addresses, you know, and privacy rules, of course, due to regulations, GDPR and other things that keep tightening, right?
00:24:28.240 --> 00:24:30.720
Much more uh is being done on that way.
00:24:30.960 --> 00:24:36.240
As more people are trying to amass their email, does that actually hurt your business?
00:24:36.720 --> 00:24:42.240
I mean, the so there's only one major one, really, and it's Apple who've done it.
00:24:42.319 --> 00:24:50.720
Gmail allow change, but actually it's quite a cumbersome thing in that they they only allow one change a year, and it's uh it means you'd have to change all your accounts, right?
00:24:50.880 --> 00:24:53.440
Your access accounts, and people aren't inclined to do that.
00:24:53.680 --> 00:24:58.880
Uh but Apple have done it differently where they they have um Hide My Email service, right?
00:24:59.039 --> 00:25:09.680
So predominantly more on apps than it is on the website, but uh for us it represents about 2% of our total traffic, so it's still relatively low, and growth rate is limited.
00:25:09.920 --> 00:25:18.240
Um so with that one, it's interesting because we actually see because it's it goes back to a website and it keeps using the same email.
00:25:18.319 --> 00:25:21.119
So we actually do have we see returning, even though they've highten it.
00:25:21.279 --> 00:25:22.319
We actually see the return rate.
00:25:22.480 --> 00:25:26.799
So instead of one email, a person has 40 because they're going to 40 different websites.
00:25:26.880 --> 00:25:30.000
So we see all of those 40 and potentially could merge them back, you know.
00:25:30.079 --> 00:25:36.400
But um, so no, um, you know, for now it hasn't had any impact from a compliance perspective.
00:25:36.640 --> 00:25:51.920
GDPR largely from a you know, GDPR has provisions for fraud prevention, you know, so it allows a lot more, you know, and rightly so, you know, to to for companies to do to have more access to data to perform fraud prevention checks.
00:25:52.160 --> 00:25:57.599
So we haven't had um like we do our GDPR compliance and doing one at the moment with Experian as well.
00:25:57.839 --> 00:26:01.599
So uh yeah, it hasn't been a big issue on that front either.
00:26:01.920 --> 00:26:02.079
Okay.
00:26:02.880 --> 00:26:17.039
So as more shopping is being done by those AI agents and more people are hiding those details, does email still stay at the center of proving who somebody is, or does it become just another signals among many?
00:26:17.200 --> 00:26:18.960
Uh so what do you think?
00:26:19.359 --> 00:26:22.960
I suppose it's uh it's always a signal among many, I guess.
00:26:23.039 --> 00:26:27.440
Uh it's a case of um in data science, like we have like a ranking, right?
00:26:27.599 --> 00:26:30.880
So so all your signals, what how how does it rank?
00:26:31.039 --> 00:26:37.519
And it is it is top, like it's starting top three, and it very, very rarely deviates from that.
00:26:37.680 --> 00:26:41.599
Um so do I see it having that into the future?
00:26:41.839 --> 00:26:42.799
Almost certainly.
00:26:43.039 --> 00:26:50.960
Um there is no, like I said earlier, there's no indication right now that it's going away tomorrow because it does there's nothing that replaces it necessarily.
00:26:51.119 --> 00:26:57.279
The shift with email actually has been that it's not used so much for communication anymore.
00:26:57.440 --> 00:26:57.759
Yeah.
00:26:57.920 --> 00:26:59.200
You know, that's been the shift.
00:26:59.279 --> 00:27:02.160
So it's it's it's use, but that doesn't necessarily matter to us.
00:27:02.319 --> 00:27:08.400
But the use of it as a as a literally a back and forth communication tool, and I mean I'm private, still used to certain degree in work.
00:27:08.480 --> 00:27:15.359
Um, but IMs and WhatsApp and stuff is is more predominant now, you know, where we would have sent an email and I used to pay someone and someone like that.
00:27:15.599 --> 00:27:20.000
But but it's use in terms of you know checkout and as a kind of the your address.
00:27:20.319 --> 00:27:28.880
Uh because again, it is that it is it is an ID slash address because if you're buying digital goods, your physical address doesn't matter anymore.
00:27:29.119 --> 00:27:34.799
So but you still need an address and it's something for those confirmations to go to, and that's email.
00:27:34.960 --> 00:27:39.119
And I now I don't see where that it would go other than that, you know.
00:27:39.920 --> 00:27:40.720
Absolutely.
00:27:40.960 --> 00:27:45.279
So I tend to be a very positive future-focused person.
00:27:45.440 --> 00:27:52.559
Um, but however, sometimes when you're building something, you're building a new company, there's always roadblocks uh ahead of it.
00:27:52.720 --> 00:27:56.559
Uh, what is something that you believe that needs to drastically change, right?
00:27:56.640 --> 00:28:12.880
It could be regulations, it could be technology itself, it could be just the way that companies are operating or maybe utilizing emails that you believe like if we can change that, we can actually get a lot more innovation, we can, you know, grow substantially, not just as a company, but overall as businesses.
00:28:13.119 --> 00:28:14.000
Yeah, yeah.
00:28:14.240 --> 00:28:18.799
Um I mean, like the rate, the rate of progress is phenomenal.
00:28:18.880 --> 00:28:26.880
And it's, you know, and right now my my issue has been is actually around with AI, is that it's that transparency.
00:28:26.960 --> 00:28:30.319
There's a lot of fear-mongering with it, uh, you know, it's particularly that.
00:28:30.480 --> 00:28:34.000
And a lot of people they're they're afraid of what they don't know.
00:28:34.319 --> 00:28:42.559
So it's AI, AI, AI, and it's gonna either, depending on your viewpoint, it's gonna destroy everything or create everything, you know, it depends on what you look at.
00:28:42.720 --> 00:28:52.559
So that transparency and that insight is is a big issue right now for for consumers, businesses, everything, because they're they're petrified, most of them, uh, as to what happens.
00:28:52.880 --> 00:29:04.079
So I don't have an answer for it, unfortunately, but like the uh is that transparency into like you know, if we were able to do it nicely and have a nice clean okay, over the next five years, this is what's going to happen.
00:29:04.240 --> 00:29:07.680
Everybody can kind of settle down and and and work towards that.
00:29:07.920 --> 00:29:12.240
Um, but the other issue I think is the confusion.
00:29:12.319 --> 00:29:27.599
For I like I'm on the I you know, okay, I build these products and stuff like that, but I do sympathize with the buyers, the the companies out there, because you go to conferences and there's thousands of companies, hundreds and thousands or whatever of companies, and for the most part, they all look the same.
00:29:27.680 --> 00:29:27.759
Yeah.
00:29:27.920 --> 00:29:30.559
And you're like, how if I was a how do I pick one of these?
00:29:30.640 --> 00:29:32.160
And how do I know I got the best one?
00:29:32.240 --> 00:29:40.160
How do I it's a massive challenge for for for companies to find a correct vendor for them.
00:29:40.319 --> 00:29:45.119
Um, and a lot of the time I think they're just sick of people, you know, pounding their door and stuff like that.
00:29:45.359 --> 00:29:53.680
Um so I think they're the two for me, if you could, if they were those two issues, you know, it would make things a life a lot easier.
00:29:53.839 --> 00:29:58.880
Um, you know, the transparency and and it I suppose they're they're both effectively about transparency.
00:29:59.039 --> 00:30:08.079
Um so yeah, um being able to filter out some of the vendors that are suitable for your industry and that noise, and then you know, transparency into the AI.
00:30:08.160 --> 00:30:11.440
Now, hopefully the transparency into AI is kind of self-fulfilling, right?
00:30:11.599 --> 00:30:14.240
That over time we'll start to get used to it, like we did with the internet.
00:30:14.559 --> 00:30:18.319
I mean, you could argue it was the same going back about 30 years now with the internet.
00:30:18.480 --> 00:30:22.799
We're like, oh my god, what now, you know, all the bricks and mortars, this the high street will never change.
00:30:22.880 --> 00:30:24.960
And yeah, it has, it has changed dramatically.
00:30:25.119 --> 00:30:26.240
So we're probably seeing it.
00:30:26.319 --> 00:30:32.960
I just hope that it doesn't happen too fast that people, you know, catch them unaware that they have time to prepare and and and and kind of get ready for it.
00:30:33.279 --> 00:30:33.839
Absolutely.
00:30:34.000 --> 00:30:36.880
Well, thank you very much for joining Executive Conversity Race.
00:30:37.039 --> 00:30:39.599
It was uh great having you on as again, and uh till next time.
00:30:39.759 --> 00:30:40.400
Yeah, thank you.
00:30:41.279 --> 00:30:45.440
If you like this episode, hit like, follow, and subscribe for more.
