Google Marketing Platform
Case Study

Enterprise Brand SAS embraces custom bidding to strategically reach connected TV viewers

SAS, a trailblazer in data analytics for nearly five decades, has cemented its position as an industry leader in analytics, AI, and data management. To demonstrate its prowess in empowering businesses through effective data utilization, SAS crafted a compelling "Get More Done" campaign emphasizing the significance of uncovering data-driven insights.

This philosophy permeated SAS's approach to its own CTV campaign. Instead of employing traditional TV advertising's blanket approach, SAS sought a more data-centric and impactful method to engage its B2B audience. To effectively reach its target audience, SAS needed an ad buying tactic that could prioritize the designated market areas (DMAs) where its enterprise clients were most concentrated. This would enable SAS to focus its CTV budget on the areas that held the greatest potential for business growth.

Embracing auction-based CTV inventory for enhanced efficiency

The SAS team recognized the potential of programmatic advertising to streamline the complex process of purchasing CTV ad inventory across multiple markets, particularly on an international scale. Traditionally, this process involved establishing close partnerships with publishers in each market, a step SAS was eager to overcome.

In Display & Video 360, advertisers typically resort to Programmatic Guaranteed for their CTV ad purchases. In a bold move, SAS opted to purchase auction-based CTV inventory, bypassing the deal-making process. This decision not only accelerated the campaign launch but also enhanced the team's flexibility in optimizing their bidding strategy with real-time data.

Creating custom bidding algorithm to reach relevant DMAs

To that effect, SAS created a custom bidding algorithm in Display & Video 360. This algorithm went beyond leveraging Google AI's default bidding solutions; it intelligently adapted them to SAS's specific needs. Instead of relying on Display & Video 360's pre-configured bidding strategies, the custom bidding algorithm was able to dynamically adjust bids to serve more impressions in priority markets while reducing cost per completed view.

Unlike traditional geo-targeting, which simply excludes all locations outside of a specified region, SAS's weighted geo-targeting approach allowed for granular control over ad distribution. The only way to match this level of control would have been through hundreds of separate line items, specific to each market. Instead, SAS added priority market insights in the custom bidding script, utilizing a single line item for optimization. As a result, SAS could prioritize serving CTV ads in the most promising areas while maintaining a broad reach across all regions.

“Display & Video 360 empowered us to bring in our customer insights and tailor our CTV investment, reaching the viewers we care about most.”

Mibbie Majors, Global Head of Programmatic, SAS

Testing the impact of custom bidding strategy using the experiments tool

SAS wanted to ensure that the custom script effectively prioritized reaching decision-makers in top DMAs. They A/B tested their custom bidding model against other bidding strategies using the Display & Video 360 experiments tool in order to measure which version drove the most completed views.

The results of the 10-week campaign proved to be a success. Not only did using a custom bidding script allow SAS campaign managers to deliver 39% more impressions in priority markets, it did so with greater efficiency, reducing cost per completed view by 24%.

With custom weighted scripts, SAS was able to create a more efficient and cost-effective CTV campaign. More importantly, it laid the groundwork for scaling this approach across other campaigns. Now, SAS is exploring the use of recently launched custom bidding signals for CTV like content duration, genre, and delivery with its Google Marketing Platform certified partner, DELVE to further optimize their CTV spend on Display & Video 360.

You might also like: