VideoWeek spoke to their co-founder and CPO Tanya Field about how Novatiq allows publishers to identify “ghost users”, the importance of interoperability, and just how many ID solutions will emerge after the demise of the third-party cookie.
Telcos have often struggled with their ventures into ad tech – how does Novatiq view the opportunity for telcos in advertising?
Telcos have a massive role to play, but in a different way than just buying up ad tech companies or looking to effectively ingest ad tech processes within their own space. For me, it’s much more about the fact that the telco network sits at the centre of every digital transaction that there is. And therefore, what they should look to do is leverage the service layer application that they can deliver into the ecosystem on a broader basis. So that means engaging with all of the ad tech ecosystem, and assisting by delivering a fundamental ID infrastructure, by virtue of the fact that they sit in the centre of each digital transaction.
This is very different to buying up ad tech companies and then trying to integrate them into a telco. It’s more about the service layer, delivering an ID infrastructure that everybody in the ecosystem can benefit from. And they can do that because they have relationships with the customers. And they can leverage the benefit of the relationships that they have with the customers to deliver a verified ID solution, because they sit in the centre of all digital transactions.
It would be a different proposition to what’s come before. I think they’re extremely well placed to do that, especially as we progress with the rollout of 5g. It would seem logical that they could extend ID solutions that they’ve already been delivering for a number of years, for example, in banking and payments, to extend that out into the martech and adtech environment.
Can you outline what exactly a ‘telco-verified identity service’ is? How does it work, and how do you recognise consumers across different devices?
It’s actually a really simple premise. Currently, we have authenticated IDs, which are about people logging in, and linking it into an email or a phone number from a login basis. The expectation is that 25, maybe 30 percent of users, for different websites will log in. But then you’ve got a vast array of ghost or anonymized users who can only be recognised from the first-party cookies that are served, and they’re limited in duration.
When you have a first-party cookie, and it runs out, another one is launched. Linking the two together without any login mechanism is very difficult. The premise of using the telco network is really simple. It’s about publishers and brands being able to link together those first-party cookies that they’re generating to create their own first-party ID. So the idea is the first-party cookie is sent into the network, and the network is able to link that to a user just by using standard network protocols.
Then we’re able to hold that first-party cookie against that user in the cloud, no personal information exchanges. But we create a record where we’ve got the first-party cookie, and the user. And then every time we get a first-party cookie from the publisher, we look it up again. If it’s the same cookie we send it back and if another cookie from the same publisher comes in, and it reconciles with the same user, we’re able to say to the publisher, that cookie equals this cookie, update your records.
It enables them to string together the same user, but they don’t have any personal details. They may have the profile details that they’ve gathered for their own website usage, which enables them to have robust datasets that are against joined up users. It also enables them to have a much greater view of the total number of users that they have using their website because they’re not double and triple counting
So simply, all it is is mapping the user together to give a 360 degree view of them, even if they’re generating different cookies in different environments. The reason we can do that is because behind the network file, whether it’s mobile, or whether it’s fixed, we work across both behind the network firewall, we’re able to put together the reconciliation between the user that’s generated the cookie, and what they map to. So we create an ID graph that is based on network transactions.
We have seen moves from Apple and Google to increase privacy at device and browser level, given this environment how will Novatiq operate? Will Novatiq be operable on Apple devices?
Yes, because the premise of how the Novatiq ID works is very different to any other IDs in the ecosystem. It doesn’t rely on any information that’s stored in the browser. So all of the information points that are being shut down in the browser, don’t affect the way that we generate our ID. Because we use the network as a source of truth, we don’t need any of the information that’s distributed in the browser itself. All we need to receive is the first party cookie that’s generated for the Novatiq Zenith ID so that we can create a robust identity structure for the publishers and the brands.
Then we have a second ID, which is called the Hyper ID, which is generated in the ad request. Because that’s a transient ID, it doesn’t have any requirement to be carried over. So it’s generated per ad request. And that’s, again, a first-party ID because it’s generated by the publisher and picked up by the SSP.
So we don’t see ourselves as being affected by the current changes, because by virtue of the type of ID, it’s not relying on any of the information that’s distributed in the browser. And that’s the reason why we use the network. It’s also the reason why we believe that the network is key to enable progress on IDs recognising the ghost users.
It won’t have the depth of logged-in users, because there you’ll have personal information, you’ll know a lot more about that 25 to 30 percent who it applies to. But actually, if you want to have the breadth and to be able to understand that it’s the same user, and therefore you can have a continuous dialogue with that user, then it’s really important to be able to link together the identification of users in the ghost environment. And for us, that’s key because we believe in advertising as a service.
For me, as a customer, I don’t want to be bombarded with the same spam advertising for things that I bought three months ago. So it’s really important that we know I’m the same user or that a transaction has occurred so that I don’t need to see that anymore. And I think the day that we can crack that will make the difference between consumers recognising that you can have advertising as a service, it doesn’t have to be a disruption.
We are seeing many ID solutions come to market in the current climate. Do you think there is space for all these solutions?
I’m a great believer in interoperability, I don’t think we’re going to see a one size fits all solution. We firmly believe in interoperability, and we work with a number of the authenticated ID players in helping them stitch together the IDs in their own domains.
We’re very different in that ours isn’t a universal ID. Our IDs are specific to each publisher or brand. But obviously behind the network, we’re able to aggregate that up so that we can run harmonious advertising campaigns.
I think there will be maybe five to 10 ID solutions when it all settles down. We have to also remember that there are different levels of privacy in different regions around the world. And the ID solutions will need to be able to cater for those layers of complexity as well. So I think there will probably be up to 10 ID solutions working together. Interoperability will be key to be able to deliver an overarching solution for the martech and adtech environment.
Telcos often worry that using customer data will impact upon their trust with customers? How do you avoid this?
So they’re not really sharing any data, they’re just stitching information together for the ID. It’s about providing a service to the consumers. To say this is a service that will enable you to carry on using the open web in an anonymous fashion, if you don’t want them to know your name, address, your email details, but you still want the services.
We can help facilitate that by safeguarding your privacy, but making sure that you’re still able to get a continuity of service, because the only way that the open web survives is by having advertising revenues.
I think it’s a delicate message, to recognise that it’s providing a service layer, not utilising the data. There is always, on top of that, the possibility that the telco can use their own first-party data to fuel advertising profiling. But that’s a different consent dialogue with the consumer which is handedly separately.
I think this is a dialogue that telcos are already very familiar with, and have been going through with their customers for a long time. You’ve only got to look at initiatives like providing priority tickets, where there is a value exchange.
If consumers trust anyone to keep their data safe it’s the telcos, because they already have a trusted relationship with them.
There are those in our industry who say that all ad targeting that uses personal data (i.e anything other than contextual data) is on its way out, and that efforts to develop alternatives are just delaying the inevitable. Do you think solutions like your own are safe from current and future privacy regulation? And do you think consumers will understand and accept this new kind of targeting?
Our solution enables people to reach audiences, those audiences are cohorts because you never know, on a one on one basis who those individuals are. I don’t at this moment believe that reaching a cohort or an audience is going to go away.The reason I don’t believe that’s going to go away is because I think it’s a requirement from the advertisers to be able to streamline who they’re reaching into audiences.
Running an ID solution, doesn’t mean that you’re enabling people to know who those individuals are and sharing data. At Novatiq, what we do is link users together and say, this is user A, and they broadly fit into this audience. I think that these approaches are very privacy-first focused. Even if there are other changes that occur, everything that’s done is with the consent of the user.
I do think we’re moving away from an approach where individuals are identified and targeted specifically but I do think that people still want to reach defined audiences. And they can reach those either through a level of contextual or through reaching them by defined audiences that are activated with our transactional ID. I struggle to see an environment where we won’t have a mix of audience cohorts, as well as contextual.