8 insights about the future of addressability & measurement post-cookie from The Drum Live

While conversations spanned a wide gamut of topics, these are eight of the top takeaways from the sessions spearheaded by adtech and media experts from Warner Bros Discovery, Reddit, Lotame, Comcast, Kantar, Quantcast and Comscore.

1. The increasing importance of first-party data strategies

As the curtains close on third-party cookies, first-party data is becoming an increasingly valuable asset for marketers and advertisers. Experts urged brands to foster direct relationships with customers to gather actionable insights, transforming every interaction into a data-gathering opportunity.

“Having … an idea what [kind of] data you want to use or not use is going to be very important. The more honest conversations that a brand can have with consumers and with its customers – on what you’re doing and why you’re doing it – creates a sense of trust between the brand and the consumer. And that will encourage consumers to share more with brands that are doing that,” said Steve Silvers, executive vice-president of global creative and media and ecosystem at Kantar.

However, collecting data is just the beginning. The real challenge lies in activating this data in meaningful ways that drive customer engagement without intrusion. Activating on that data begins with gaining a deeper understanding of it, argued Michael Sandor, sales director of adtech firm Quantcast.

As Sandor put it: “As the amount of first-party data and deterministic datasets gets smaller and smaller as privacy regulation increases and the cookie goes away, [it’s important] for brands to place additional emphasis on really getting an understanding of that first-party dataset,” said Michael Sandor, sales director at adtech firm Quantcast.

2. The proliferation of privacy regulation and privacy as a driver of change

Amid an influx of data privacy regulations across the globe, data signal loss across the web and the growth of consent-based models of data collection, consumers have come to expect more of a say over how their data is collected and used by advertisers and publishers. Obtaining explicit permission from consumers to use their data for targeting purposes is becoming a cornerstone of trustworthy digital marketing practices.

Marketers must stay agile and adapt to evolving legal frameworks, especially in the US, where states are taking privacy legislation into their own hands in the absence of a comprehensive federal law like the EU’s General Data Protection Regulation.

“Privacy legislation is going to continue to increase and there are going to be less identifiers … It just magnifies the importance of getting ahead and having a multi-signal approach,” said Quantcast’s Sandor.

3. Incrementality as a problem worth solving

​Testing incrementality – assessing the lift that advertising spend provides – helps marketers justify budgets and refine strategies. But it’s easier said than done, experts agreed.

“It’s not enough any more to just measure exposure and an outcome,” said Seth Goren, group senior vice-president of subscriber growth and media at Warner Bros Discover. “You’ve got to figure out how to do experimental design at scale and in realtime and it’s an easy thing to agree to … but it’s a very challenging thing to do in practice.”

4. The critical role of data collaboration post-cookie

Isolation is becoming increasingly infeasible in an environment with limited data signals, making collaboration all the more critical.

Lotame’s Theriault also highlighted the importance of data collaboration by bringing together first-, second- and third-party data to enhance insights and decision-making capabilities. This collaboration extends beyond the confines of individual organizations, involving external partners to deepen the data pool available for analysis and activation.

But clean rooms are only one tool for effective data collaboration, Theriault said. “In [data collaboration’s] simplest form, there are cleanrooms – the likes of Snowflake [and] AWS. And they’re simply trying to join together two different datasets, [like] a brand’s first-party data and a publisher’s first-party data.” Advertisers who team with a range of partners, however, can unlock different levels of data insights that enable them to move beyond simple privacy-safe ID resolution.

5. The danger of over-indexing on walled gardens

The data-rich, logged-in environments of the world’s largest social platforms have, in some ways, become a crutch for advertisers as they face an uncertain, cookieless future.

As Lotame’s Theriault explained: “Because there’s no real clear path [for addressability beyond third-party cookies], a lot of brands are migrating to social media and walled gardens [because] it’s just easy and the industry is really complicated. [Today,] 66% of digital ad dollars are spent within walled gardens, but, interestingly and ironically, 66% of consumer time is spent outside of gardens. So, there’s a ton of competition within the walled gardens for those eyeballs, but there’s all of this green field outside of it. We’ve relied too heavily [on the walled gardens] historically. And so I don’t envy the brand marketers who are trying to figure out how to retain addressability outside of the walled gardens.”

6. Witnessing contextual advertising’s renaissance

With cookies crumbling, contextual advertising is having a renaissance. Understanding the environment where ads are placed can often be as powerful as who is seeing them.

“We’re having a ‘what’s old is new again, moment – like it's 2003. We’re talking about contextual, we’re talking about [marking] mix models – this is bleeding edge stuff from the early aughts,” said Silvers.

At Reddit, said Belliveau, the platform has invested significantly in contextual and interest-based targeting – a decision the company has found useful from both ad performance and measurement perspectives. “It also helps us have a better connection with our advertisers and communities because they’re open [to receiving branded messages] What we find is, if we’re connecting them with brands that make sense for them – that’s part of their journey in the world, or they’re discovering new products – they’re receptive to that. And so those signals are strong and they work. And then we’re able to showcase success and measure things.”

7. Leveraging AI for targeting and measurement post-cookie

Artificial intelligence and machine learning are beginning to play key roles in predicting consumer behavior and refining targeting strategies, offering a glimpse into the future of automated marketing.

“AI is going to be so transformational – in ways that we don’t even understand yet,” said Mary Ann Belliveau, vice-president of large customer sales at Reddit. “It will be able to read, interpret, digest and extrapolate. We know that. [In that case], the details of some of what marketers are doing now [will] become less important because AI will understand smaller datasets, and [AI] will understand how to work with it.”

Reddit itself, Belliveau said, has made significant investments in AI and machine learning in the last couple of years across a variety of applications, including determining “how we can make sure that, if we’re really anchored on contextual and interest-based targeting, machine learning and AI can help us match the ads better to that situation.”

8. Combining measurement tactics for maximum impact

Measuring the impact of advertising spend becomes trickier without cookies. Marketers are exploring new models to attribute conversions and understand their media performance. While tried-and-true measurement approaches such as multi-touch attribution will remain front-and-center for many advertisers, the resurgence of contextual and the debut of a range of both deterministic and probabilistic ID solutions are introducing new means of measurement in lieu of the third-party cookie.

As Kantar’s Silvers said: “We have a ‘Swiss Army knife’ approach to both targeting and measurement because the signals that are available in a particular channel or on a particular event will be different. And if you limit yourself to a single method, you will be blind to those opportunities.”

The idea was echoed by Quantcast’s Sandor, who said: “We’re constantly thinking about how we can appropriately balance deterministic [ID data] – which is limited in scale but can be very effective – with some of the other multi-signal approaches like probabilistic [ID data], cohorts and certainly contextual, which can be applied in a smart and relevant way. I don’t think there will be one solution that replaces the cookie.”

For more, sign up for The Drum’s daily newsletter here.