Crumbling cookies and the challenges of personalisation and remarketing

As third-party cookies slowly become a thing of the past, StopPress chats to one of the country’s largest digital publishers, NZME, about their data strategy, remarketing and how to best engage with readers on a personal level.

NZME is home to some of New Zealand’s biggest digital brands; The Herald, OneRoof, GrabOne and Driven. It’s not surprising that their marketing team is well-ahead of the changes to the way in which digital brands collect and manage consumer data.

We chat to Richard Wright (RW), Digital Analytics Manager and Charlotte Bell (CB), Head of Digital Revenue at NZME about the media company’s approach to data collection.

With third-party cookies being phased out by 2022, what implication does this have for marketers collecting data of visitors to their sites?

CB: There are huge implications for publishers and marketers and at NZME, we are embracing this change. It’s a step towards a privacy first future and that’s a good thing for everyone. However, it’s important to understand that this change impacts third-party cookies only and when we collect data from visitors to our websites, we generally use first party cookies. Our focus as a publisher is to deliver the best possible user experience and content and we don’t need third-party cookies to do that.

The biggest impact will be to a marketer’s ability to use the myriad of existing third-party audience segments when trying to attract new visitors to their sites. This third-party data is often used without any real understanding of how audiences are qualified into these segments, if the people targeted have consented to this data use, or if they are even aware of it. This will no longer be possible without third-party cookies.  

Google are promoting FLoC (Federation of Learning Cohorts) as a replacement to the existing third-party cookie audience segments, and although FLoC is unlikely to offer the same flexibility and granularity as the segments available today. Everyone will have access to the same segments however savvy marketers will realise that in order to gain a competitive advantage, they need to maximise the value from their own first party data and work with partners who have valuable first-party data. 

Is relying on first-party data going to be enough? Or what other collection tools/ methods will marketers use?

RW: At NZME, our first party data is expanding, through strong growth in registered and authenticated audiences. There are a number of initiatives underway to leverage this audience data strength, and also deliver effective advertising solutions.

Some of the tools that marketers might want to add to their toolbox include:

  • Contextual ads based on specific context, topics, and the meaning of articles (using techniques such as Natural Language Processing and auto-tagging of content).
  • Identity based advertising using first party data matches with publishers or via services such as LiveRamp. In both cases, data matches can be executed using secure and privacy compliant methods.    
  • Smarter user segmentation (using first party data) based on Machine Learning and data science.  Modelled user segments can be created for ad targeting through deep analysis of on-site user behaviour.

Of course, FLoC provides an alternative to third-party cookies to determine audience segment or interests, when targeting Google Chrome users. FLoC will determine and store the users’ segments and preferences in the browser and provide these to advertisers to enable ad targeting. FLoC enabled browsers will be available in March for testing.

Do we (marketers) still need third-party cookies?

CB: Third-party cookies have been around for the last 30 years and this is a pretty archaic technology that has ended up not serving consumers well. The original purpose was to provide a more personalised experience for users; however, they are now mainly used to track user behaviour’s as they move around the internet. This practice is very opaque to the consumer and can feel intrusive and creepy. The inventor of the internet, Tim Berners-Lee has suggested recently that, if he had his time again, he would not have third party cookies as a core part of the internet infrastructure.

There is a real opportunity now for publishers and advertisers to get it right and put the consumer first in terms of privacy, user experience and personalisation.

Retargeting is important in getting re-visits, how will/ should marketers approach this when handling audiences and segmentation?

RW: Retargeting will definitely be challenging with the deprecation of third-party cookies. If someone visits your website, you will not be able to target these users on other sites using third party cookies. Google is considering a system called FLEDGE which is a concept of a ‘trusted server’ where you may provide your first party data to a trusted third-party entity and they can deliver remarketing on your behalf. Remarketing services such as Criteo, NextRoll, Magnite and RTB House would become ‘trusted’ and be able to serve remarketing after going through some level of certification.

As stated earlier, smarter segmentation and contextual advertising is gaining increased importance in answering these challenges. At NZME we are doubling down on Machine Learning techniques to accurately segment users.  An example is first home buyers on Oneroof – how can we determine an accurate First Home Buyer segment on Oneroof based on the users behaviours, profile etc and then scale this segment to the wider NZME network? It sounds easy, but it is really quite difficult – the home buying journey is complex, influenced by market factors and can be long and complicated. Machine Learning is really the only way of finding these sorts of high value, high complexity segments and is where we are focusing our segmentation efforts.

CB: Native advertising and content are also a great way to engage with niche, targeted audiences. For example, in a recent campaign NZME developed a bespoke solution to personalise not just the advertising, but also the content itself, based on first party data. We’re using dynamic content to create literally thousands of content variants with different headlines, copy and images to deliver each user the most engaging and effective content, tailored to them. NZME is working with advertisers to bring their data ideas to life in ways that the walled gardens just can’t do. 

When collecting data, how is NZME ensuring that it’s done ensuring the privacy and security of customers?

RW: NZME has a very big focus on the privacy of our customers and we have stringent processes in place to ensure that our customer’s privacy and security is at the heart of what we do.

We work closely with advertisers and partners around utilising first party data and this is always conducted in a privacy compliant way. This will become increasingly common in this new age. Brands and advertisers will need to ensure that they have the processes in place to share data in a privacy safe way. This could include using a data safe haven or platforms such as LiveRamp to ensure privacy and data security.

Additionally, as we roll out Machine Learning and AI initiatives in our platforms we have to be sure that we are not introducing bias into our targeting and segmentation. The goal is to develop “explainable AI” solutions rather than “black boxes” to ensure our solutions are as transparent as possible. 

What of consent management platforms? How important will these become?

RW: The changes to privacy legislation in December did not explicitly have cookie consent as a requirement, which surprised some people, given GDPR and CCPA’s implementation of this as a core tenant. This omission does not however absolve publishers of consent responsibility. 

Our goal is to ensure that all of our users will always be aware of what they are opting into on our platforms, and our privacy policy is regularly reviewed as our technology platforms evolve. We also have stringent consent and sunsetting mechanisms for email to ensure not only privacy compliance, but also that we’re doing the right thing by our audience. We certainly embrace informed consent.

How important is it to create content that readers want to engage with at a personal level? How does this play into data collection/ sharing?

RW: One of our key strategic priorities is to be ‘New Zealand’s Herald’ – and content and personalisation is key to this. We have recently built two Content Recommendation Engines (for web and app) with the goal of providing users with a Machine Learning driven content recommendation experience. These have proven to be extremely successful at engaging our audience and are being constantly improved. The Herald publishes hundreds of articles every day so it’s crucial for us to be able to surface the articles that our users want to see. It’s a balancing act however and we’re wary of the echo chamber effect that can be created by algorithmically driven news feeds.

We’re also working on providing a more personalised and engaging experience via email, push notifications and elsewhere on our sites. Research suggests that users are happy to share data, provided they are getting value back in return, and delivering a more valuable and engaging experience is the primary reason that we collect data from our audience. 

For more on our Cookies series, click here.

In association with NZME.

About Author

Comments are closed.