The Future of Marketing is a monthly sponsored content series by Amazon Ads exploring martech, shopper insights and consumer trends.

By Steve Pinto­, worldwide GTM lead, addressability and privacy

Today, almost 40% of web and 37% of app traffic is unaddressable via traditional advertising methods, and by the end of 2024 that number is expected to climb to 95% across the board as third-party cookies deprecate. 

But there’s one important fact that hasn’t changed with the evolution of technology: brands still need to connect with customers and measure the impact of their efforts.

Increasingly, that means finding smarter solutions when it comes to understanding your audience – solutions like Amazon’s signal-based marketing capabilities. Signal-based marketing is Amazon’s approach to leveraging available signals along with machine learning to deliver relevant messages without the need to rely on third-party cookies. It’s an approach that respects user privacy, while still offering options they’re likely to be interested in. In other words it’s a win win.

And what are those signals? You can find them in a wide range of customer events and behaviors that can indicate interests and affinities, observed at particular moments in time. Assembled together, they create a map that allows Amazon DSP’s machine learning models to predict interests and affinities. 

Consider the insights you can gather via Amazon’s purchase, streaming and contextual signals. Interest signals from Amazon’s global retail and streaming properties helps the platform understand the ways customers discover, consider and purchase products and services. These include contextual signals that help advertisers understand the content customers engage with along the way. Together, these signals inform ad-serving and optimization decisions, both on Amazon properties and across the open web. 

Then there are advertiser first-party signals, including website engagement, and conversion signals, and specific audiences stored across advertisers’ marketing stack (based on hashed records, third-party ad identifiers, third-party cookies or mobile IDs). The combination of Amazon Ads signals and an advertiser’s own signals feed into models enhancing campaign recommendations and optimizations.

Then we have publisher-direct signals. Data from Amazon Publisher Direct provides Amazon DSP with direct access to publisher inventory. These signals further inform modeling – and because Amazon minimizes intermediaries in the supply chain, they are received faster and with higher fidelity. 

Finally, Amazon maximizes the use of RTB signals including device type, time of day, software version and page URL – all of which help aggregated machine learning models understand ad opportunities with finer granularity. 

Combine all of these together and advertisers make better ad-serving decisions at scale. 

“Cookies are, and always have been, false precision,” Neal Richter, director of bidding science and engineering at Amazon Ads, recently explained in AdExchanger. “Signal-based marketing, by contrast, allows brands using Amazon DSP to engage customers with relevant, useful advertising, as opposed to repeatedly showing them an ad for a product they had considered and not purchased.”

Signal-based marketing combines all of these signals to help brands reach customers without sacrificing reach, relevancy or ad performance – or relying on privacy-invasive identifiers. In fact, the approach was specifically created to lessen the necessity of third-party identifiers when matching an ad to the right place and moment. With signal-based marketing, customers see more relevant messages in more locations – and they’re not followed around the web by the same ad on repeat. So far, brands using Amazon DSP in the US have seen a 20-30% increase in addressability on browsers like Safari, Firefox, and iOS. 

People-based marketing, an alternative to signal-based marketing, is about targeting an individual, but it can put a customer’s privacy at risk and can often be wildly inaccurate about what customers are actually interested in. According to Amazon, this manufactured concept of an individual can vary widely, from buyer to buyer, marketer to marketer, platform to platform and site to site, because people-based marketing works by tying multiple identifiers together to create one profile of a customer.

By contrast, signal-based marketing doesn’t depend on the past and isn’t about targeting an individual. Instead, it relies on in-the-moment signals and predictive methodologies made possible by Amazon’s machine learning models. It enables brands’ advertising efforts to be relevant without identifiers and without sacrificing scale. 

In other words: No cookies? No problem.

To learn more about Amazon DSP and how signal-based marketing empowers brands to connect with customers while prioritizing trust and privacy, click here. 

———–

Steve Pinto is the Global Lead – Addressability & Privacy GTM for Amazon AdTech. He keeps Amazon Ads’ sales team informed and trained on the latest ad addressability and privacy updates while working directly with clients to bring our latest targeting products to market. He recently launched the Amazon Contextual Targeting product to Amazon DSP’s global customer base.

Prior to joining Amazon Ads, Steve was the Head of Supply Quality at the Meta Audience Network and responsible for Data Product Partnerships at Visa Inc.

———–

Enjoyed this article? Learn more about how you can stay ahead of the curve in The Future of Marketing