How AppNexus Customers Use Data Science And Programmable Technology To Build Hyper-Effective Marketing Campaigns

In an increasingly competitive digital arena, the only way for brands and agencies to gain a real advantage in the marketplace is to take matters into their own hands. Rather than relying on human intuition and the opaque decisioning of black-box platforms, today’s most successful marketers are leveraging their own data to build unique algorithms that optimize creative and spending choices based on their individual business needs.

As we push forward into the Programmable Era, a time of intelligent marketing and a personalized internet, it will only become more important for buy-side professionals to be adept at combining customizable technology tools with robust data science to optimize their campaigns.

At the AppNexus Programmable Marketing Forum in London earlier this month, three of our advanced customers -- Media iQ Digital, Mindshare and Wayfair -- shared the tactics they’re using right now to get ahead of the curve.

Media iQ Digital Is Using Complex Data Science To Identify Likely Customers And Serve Them The Right Message

The analytics technology company Media iQ is working hard to move its customers beyond the traditional demographic approach to segmentation and into a more predictive approach based on deep insights into consumer behavior.


Onstage, the company’s head of product management John Goulding shared case studies of how Media iQ has used data science to accomplish this goal. In a particularly powerful example, the firm worked with a home furnishing client to personalize its messaging to different customer groups -- even though those customers did not exhibit similar behaviors based on demographic characteristics like age, gender or location.

In order to predict the rates at which a certain consumer would spend money on home goods, Media iQ created individual customer profiles by connecting data stored in its client’s adserver with information from its CRM and on-site pixels. With a user-level view in hand, the data scientists used a technique called clustering to break down hundreds of millions of rows of customer information into user groups that demonstrated similar buying behavior despite their disparate demographic attributes. The client then used dynamic creative optimization to show consumers different ads and a personalized homepage experience based on whether the user was in the Just Browsing, Professional Designers, or DIY Renovators cluster.

The result? Higher revenues for the client and a better ad experience for its customers.

“Programmatic isn’t just about bid optimization or efficiency anymore,” Goulding said. “It’s also about driving relevance and impact.”

Mindshare Worldwide Is Helping Clients Own The Big Moment With Personalized Bidding Logic.

With the Rio Olympics on the horizon, marketers have a chance to heavily influence consumers by showing them the perfect message during the emotionally stirring moments of the games. At the Programmable Marketing Forum, Mindshare Worldwide head of programmatic Adam Ray explained how his media agency is using the AppNexus Programmable Bidder to build creative and bidding strategies that capitalize on the moment marketing opportunity.

APB runs on a programming language called Bonsai that allows marketers to easily construct decision trees that determine how much a brand will bid and which creative they will use for different audiences at different times. Rather than having to discuss strategy with clients using industry jargon about “DSPs” and “ad exchanges,” Mindshare is using APB to build campaigns grounded in the basics of who, what, when, where, and how.

As an example, Ray walked the audience through a bidding tree Mindshare built with a client brand to reach consumers during the final of the men’s 100 meter sprint this August. Together, they segmented customers based on things like their location, past spend with the brand and whether they were being targeted before or during the big race.

The chart below demonstrates a small piece of the bidding tree Mindshare constructed with its client, which ultimately included 8,000 variations of how the brand will use its voice during the Olympics.


Crucially, APB has not only allowed to Mindshare to create highly targeted campaign strategies, but to do so using simple terms everyone can understand.

“APB gives us great opportunities to be creative on things which we simply couldn’t have done before,” Ray said. “We had the language to do it, but we didn’t have the technology to deliver it.”

Wayfair Partnered With AppNexus To Build A Highly Intelligent In-House Bidder

Wayfair is one of the world's largest online destinations for "all things home" with a fairly sophisticated marketing team that runs all of its ad campaigns in-house. Recently, the firm rolled out Magellan, a proprietary display advertising platform that it built on top of the AppNexus Programmable Bidder technology.

In his presentation, the company’s associate director for marketing platforms Matthew Herman explained how Wayfair went about choosing a partner that would allow it to retain a high degree of control over its campaigns, with complete transparency into reporting and the ability to iterate quickly based on marketing performance data.


Herman compared off-the-shelf, vendor-style solutions to a Model T and said that what he wanted instead was a Ferrari with lots of levers to pull however he sees fit. While AppNexus provides data insights, Wayfair retained complete control of how they’re applied by uploading its own complex decision trees onto the platform. The company also built its own dynamic creative solution to serve ads with product recommendations that are exactly the same as those it puts on its website and in its emails.

Still, even with all of this control, the company needed a degree of support. For starters, Wayfair didn’t want to build the bidder itself, as doing so is time-intensive and requires heavy maintenance after the fact. In addition, Herman said his company benefitted from sending team members to learn the finer points of ad-tech from AppNexus’ experts, as well as from the insights generated through APB’s log-level reporting and working sessions with AppNexus data scientists. In fact, recommendations made by Wayfair were later integrated into the APB platform itself.

As for the effect on Wayfair’s business, well, we’ll let the customer speak for itself.

“We have much more control over our campaigns, the performance is outstanding, and working with the AppNexus team has been great, as well,” Herman said.

To learn more about how our customers are achieving success in programmable marketing, view the complete video of their presentations here (they start at the 19:20 mark).


  • AppNexus Events