Kepler’s Global Amazon Lead Hannah Grobmyer breaks down the key takeaways from Amazon’s largest advertising event.
Amazon hosted its largest advertising event, unBoxed, in NYC in October. The event featured a surplus of new product and feature announcements, but many of the key talking points resonated with us as cross-channel marketers. Below are some of the themes that we heard repeated throughout the three day conference and ways in which we think advertisers should be thinking about them now.
Cookie deprecation and addressability
Joe Mitchell, Go-to-Market lead for the Amazon DSP, talked about the difference between precision and accuracy in targeting. While changes to the ecosystem based on cookie deprecation may lead to loss of precision targeting, if we’re smart digital marketers, we’re not actually sacrificing our accuracy in reaching our audience.
One solution Amazon is providing retailers is to expand their Contextual targeting. Using advanced signals to identify users who are browsing on aligned content, or content deemed to be relevant to their shopper journey, Amazon is able to locate target audiences on and off the site.
For all advertisers, Amazon has expanded their existing “Amazon Audiences” to include additional event-based signals. This reduces the need to view exactly what has happened, via cookies, and focuses more on what is likely to happen—e.g. a future conversion.
Amazon is also helping to address gaps in conversion reporting caused by the loss of cookies by including modeled conversions as well as modeled reach/frequency within the DSP. This eases the transition away from cookies and helps provide greater visibility into the business impact of digital campaigns.
Machine learning and automation
Machine learning came up in several sessions. While it may still be an amorphous concept for many of us, it’s actually already at work in more ways than you may expect.
Machine learning helps power the increased audience signals leveraged in Amazon Audiences. When used within the Amazon Marketing Cloud ecosystem, machine learning helps identify new segments to push back to the DSP for targeting—e.g. previously unknown cross-sell opportunities.
That said, one panelist noted that the “anchor” to all analytics work should be the insights, and insights are inherently human. As much as we may rely on automation or machine learning to help us move faster and smarter, these tools aren’t a replacement for hands on keyboards.
Measurement and data clean rooms
Traditional measurement methods like MMMs and Brand Lift Studies have found their way back into vogue as it becomes more challenging to view 1:1 attribution—both with cookie deprecation and walled-garden platforms. That said, the primary measurement tool discussed across sessions at unBoxed this year was Amazon’s data clean room solution, Amazon Marketing Cloud (AMC) and the many ways it can help unlock value for advertisers.
Kepler was the first AMC Certified partner globally, so we’re no stranger to the value AMC offers. That said, adoption of DCRs, and AMC in particular, continues to be lower than expected across the industry. We’ve learned that the best way “in” is to focus on the business questions, and then look to derive insights and opportunities with AMC.
Keerat Sharma, Director of AMC & Ad Tech Solutions, spoke about how AMC is not a point solution. Learnings derived from AMC can support campaign planning and optimization, creating a feedback loop for improved performance, and often increased investment.
Connectivity and interoperability
Many panelists spoke about the need to connect technology systems to create the right flow of customer information, activation, reporting, results and planning. With Amazon this could be the connection between leveraging AWS and S3 for your data storage, Amazon DSP, Console and Sizmek for your ads, and Amazon Athena or AMC for your insight generation.
As marketers we should all be considering how our systems will talk to each other in the future. Do we have the appropriate tech in place to both understand and activate against our first-party data? What “holes” do we have?
With the dominance of walled-garden marketing, the decline of third-party tracking, and the abundance of devices in the household, it will be increasingly difficult to tie marketing systems and measurement tools together. Providers like Amazon that have a multi-tool solution may see traction among savvy marketers.
Streaming and video
No advertising discussion is complete without an angle on creative, and the formats that are engaging consumers now. Video and Streaming TV were a central theme across sessions with the introduction of Video for Sponsored Brand and Sponsored Display, and more upper-funnel activations with opportunities like Thursday Night Football, virtual product placement and Freevee pause screens.
Amazon is doubling down on the relevance and importance of video, noting that consumers are 17x more likely to click on a video than a standard banner, and that 88% of consumers say they were influenced to buy a product from a video. They’re also creating solutions to help advertisers more quickly and easily create videos for themselves—leveraging Amazon’s support with Video Builder and Amazon Creative Services.
I personally enjoyed speaking with Hanesbrands’ Katie Tripodi, and Amazon product leaders Victoria Ruff and Andrew Cole about the future of Streaming TV. We discussed Amazon’s new and interactive feature releases, as well as where we see the future of Streaming TV. Our key takeaways focused on the need for smarter planning, buying and measuring to navigate a divided ecosystem.
Diversity, equity and inclusion
In the Day Two Keynote Latasha Gillespie, Global Head of DEI for Amazon, spoke on the importance of DEI efforts in advertising. It was a poignant and fitting way to close out the week. She shared two resources on how to improve representation in messaging, including the Amazon Studios Inclusion Policy and Kantar's Inclusion and Diversity in Advertising (Global playbook).
DEI is inextricably linked to everything that we do. Not only is it important for us to consider positive and genuine representation within our creative messaging, it’s also important to think about how and where we are spending our ad dollars, and the hidden complexities of targeting.
For instance, machine learning tactics have notable issues with bias. While it may be “machine” driven, it was built by a human. Unconscious bias extends into the way we build and activate models. This makes it even more critical for us to think strategically and as humans first as we activate marketing efforts.