A New Standard for Audience Targeting

Audience targeting is essential for every advertiser — if the message isn’t reaching the right set of consumers, the campaign cannot make the desired impact.
But sophisticated audience targeting is complex, time consuming, and difficult to scale. The vast array of targets available on and across platforms makes linking them together difficult for the human to accomplish manually. Human error can sometimes lead to the neglect of some high-performing targets when campaign managers select platform specific targets. On each platform, target sets must be built from scratch as there’s no way to utilize previous work. As a result, account managers are forced to build campaigns slowly and piecemeal, lacking the precision, speed, and efficiency that automation promises.
Pixability’s data science team knew there had to be a way to improve efficiency without sacrificing granularity. To extend the intuition and expertise of our account managers, and to deliver better campaign performance to our clients faster, our data scientists are developing a streamlined solution to target audiences across platforms (YouTube, Facebook, Instagram and Twitter) at scale. Using data science, we’re creating a shared ontology of how target audiences relate on and across platforms. By mapping keywords to targets, establishing relationships between these targets, and learning from past selections, Pixability’s targeting system accelerates the precision and efficiency of audience targeting for our account managers.

Here’s how: when an account manager enters a keyword, say, beauty, the system displays related targets (makeup, hair care, skin care) grouped by hierarchical relevance, across each platform. Related targets are grouped together, even when the targets are on separate platforms, such as a “hair care” group for both Facebook and YouTube targets. The system also displays high-performing adjacent targets, relying on the account manager’s intuition to select adjacencies based on relevance to the specific campaign — for example, beauty adjacencies include sports, fitness, and health food. This shared ontology of related and adjacent targets allows for intuitive navigation at various levels of granularity. The system learns from the account manager’s selections to display even more relevant target groups and their adjacencies in the future, and the account manager can both export and save their target selections to streamline future campaigns.
Equipped with this new solution, our account managers can plan and execute ad campaigns faster and with greater precision, as well as perform more optimizations in-flight. Campaign setup time is decreased by 95%, and the time required per optimization is down 90%.
By creating a shared ontology of cross-platform targets, displaying these targets interactively, and using machine learning to improve upon past selections, Pixability’s audience targeting system is revolutionizing how we execute video ad campaigns. By combining human intuition and machine learning, our data scientists are setting a new standard in audience targeting — and ultimately, campaign performance — accelerating the speed and precision of connecting relevant ads to the right viewer.
If you’d like to see what Pixability’s revolutionary video advertising solution can do for you or your clients, contact us today.