In truth, it doesn’t matter when – or if – third-party cookies go away for good.
Despite the industry’s collective hand-wringing, we’re in a moment rife with possibility. The advertising industry has never been better positioned to think differently about how to delight consumers with useful and relevant advertising while continuing to prioritise their trust. Of course, this requires making use of different, sometimes new technologies and tactics and embracing an open mind.
Here are a few recommendations for your advertising strategy as we inch closer to 2024 – regardless of whether cookies disappear.
1. Maximise first-party signals
At first, this may sound like nothing new. After all, brands have been ingesting first-party signals for years now. But as we move away from third-party cookies, the necessary shift is in understanding that it’s not just about the number of insights and signals advertisers have.
Instead, brands should be exploring and maximising the utility of their signals and insights. For many, figuring out how to do this is still a work in progress.
In the coming months, confirm that you have the ability to better understand your customers by working with ad tech providers that can help you make sense of and use your first-party signals in a privacy-safe manner. This can take many shapes.
At Amazon Ads, we have worked to empower brands to analyse their pseudonymised signals. They can discover new insights that inform activation strategies or combine them with Amazon Ads pseudonymised signals to create better connections with consumers in the Amazon Marketing Cloud (AMC), our cloud-based clean room solution. They can take action on those audiences in the Amazon DSP to reach their audience both on and off Amazon properties
And it doesn’t begin and end with audience activation. Your first-party signals are your primary source of truth. You leave opportunities on the table by not fully putting first-party signals to work.
2. Combine with and learn from other sources of truth
A brand’s first-party signals are like a seed that can blossom into something greater with a little help. That’s why we’re encouraging brands to enrich their first-party signals with additional quality signals. It’s essential to get into this habit now, regardless of what happens to cookies.
Working with ad tech providers that have their own quality, differentiated first-party signals can help to make your signals even more powerful. As a recent report from Forrester stated, “differentiated access” to first-party signals sets a DSP apart. When we think about what that means at Amazon Ads, we consider the browsing, streaming and shopping signals populated by the breadth of insights available only to us.
And the quality of signals matters. The combination of your first-party signals with those available through a DSP with differentiated access can help enrich campaign recommendations and optimisations and expand reach with lookalike audiences based on your signals.
The “seed” of a brand’s first-party signals enriched with additional quality signals can inform machine learning models, for example, in a DSP that can better predict ad relevance.
3. Lean into modelled signals
This begins with a mindset shift. If you’re constantly trying to “keep pace” – in this case, by maintaining the same strategies to meet performance metrics made possible by cookies you’re setting yourself up to stay afloat, not succeed. Brands should instead be focused on determining the best ways to use probabilistic methodologies to both reach audiences and measure ad performance accurately.
Brands should also be mindful of matching the moment to the message in their ads. Content focused on an upcoming basketball game that’s accessed on mobile devices during rush hour may indicate readers have an interest in comparing ticket prices. There are methods of reaching audiences during these moments that don’t involve first-party signals, but rather the context of an opportunity as understood by probabilistic signals.
You should still focus on your objectives, but consider other methods of getting there. Brands poised to weather the death of the cookie will benefit by being flexible and taking advantage of predictive-based modelling methodologies. That might mean adjusting your CPA expectations to achieve scale or perhaps even exploring new ways of measuring success.
Certainly, first-party signals are critical. But as technology advances, probabilistic signals are increasingly valuable. That is, as long as they are built on top of sound technology and consumer understanding, with a provider that has always put customer privacy first.
