Nexxen Connect: Q&A on Political Advertising with Andrew Mullins

What are examples of how you tested new strategies with the direct response and paid media campaigns?

At the end of the day, voters (and donors) support a candidate because they want something from the candidate: Be it a policy position, or delivering a win for their Party.

For us, we wanted to offer our supporters something valuable and tangible in return for their donations or data. We never tried to trick or deceive our supporters (we never did those fake 1500% matches, for example). One way this manifested itself is in the way we handled debates.

For nearly every debate, we in-house executed a “trip to the debate” contest. Essentially, we’d offer our audience a chance to get flown out to a presidential debate in return for their email address, phone number, or a small donation. We executed everything we promised! When the contests would wrap up, our team woke up early on Saturdays to conduct an actual drawing (per the contest terms) and one of us would call the winner to share the happy news. This took on a gameshow “you just won a million dollars!!!!”-type vibe. Our audience loved these contests, and they really became a valuable way we expressed gratitude to our supporters.

As a follow up can you describe the process of creative iteration – why is it important, which parts are manual vs. automated?

We took a lot of pride in tracking every single data point of our paid marketing, and letting performance guide all of our decisions. For every user we acquired on ads, we knew exactly the tone, message, and targeting strategy that led to the user acquisition. This guided our entire creative process. For example – we learned pretty early on that border-centric ads brought in high-quality users that donated to our campaign weeks and months after acquisition. We weren’t so much looking at immediate metrics (like CPA or ROAS) as we were lifetime value metrics (like future donor conversion rate). Those metrics gave us the insights we needed to focus on QUALITY user acquisition, not just high-quantity. This is something most campaigns don’t have a data backbone to execute, and we did it well, and at scale.

We also tested literally everything. At one point in the cycle (like September I think), I tallied up the number of creatives we had been running on Facebook compared to other campaigns, and the results weren’t even close… (and this was just Facebook). We focused on iterative testing. We found components of an ad that drove performance (everything from message to font to background color), and iterated on the ad until we found the perfect mixture for performance.