THE LAB: The Attention Problem
Attention and persuasion aren't a tradeoff. They're unrelated.
March 2026
What grabs attention has no relationship to what persuades. Get it right, and your content can grab voters and move them — at the same time. That should fundamentally change how you approach your ads.
Conventional wisdom in political communications holds that engagement and persuasion are a trade-off. What goes viral rarely changes minds.
We've found that the conventional wisdom is wrong.
What We Found
We ran approximately 150 randomized trials over the last six months. The same three results kept showing up regardless of election, geography, or message type. Together, they challenge everything we thought we knew about paid media.
FIRST: Attention and persuasion are uncorrelated. There are things that fascinate you and bore you. There are things that change your mind and things that don't. Those aren't the same set.
SECOND: Attention is predictable. Not in a trivial "optimize the color of the carpet in the powerwashing video" way, but in a practical one. If something isn't skipped, it probably won't be skipped again. Attention patterns repeat. You can learn them, reuse them, and let them compound.
THIRD: Persuasion lives in audio. We learned this accidentally years ago, then confirmed it this year. The video component of a produced political ad contributes basically nothing to its persuasive impact.
Voters don't pay attention to the ominous red-and-gray color grading they've seen 100,000 times, or the down-on-their-luck family staring at bills at the kitchen table. Our minds have been trained to ignore stock-video slop. It doesn't inform voters, and it doesn't move votes. It just signals "this looks like an ad."
The voice talking to the voter, and the argument the voice is making, does the work.
What We Did About It
If persuasion lives in audio and attention is predictable, you're free to be weird with video.
So we got weird.
We built ads that are 30 seconds of someone cleaning a carpet, pressure-washing a driveway, or playing with kinetic sand. No candidate. No b-roll. Just hypnotic visuals with a message running across the bottom.
We call it "mesmerizing content."
The theory: people skip ads, but they don't skip content they genuinely want to watch. So we stopped interrupting and started blending in. We paired our strongest messages with content tailored to voter interests — gardening, gaming, sports — to capture attention with hypnotic visuals.
The Results
It worked. Completion rates jumped. Persuasion held. When we matched videos to voter interests, we saw a +4% treatment effect on completion rates — and a -4.3% change in Trump approval among those who completed.
These ads performed in the two most important Democratic races in 2025: Abigail Spanberger in Virginia and Mikie Sherrill in New Jersey.
The Implication
This test was limited to 30 interests. In practice, we can create tens of thousands of tailored ads that precisely match niche interests and increase competitiveness in the attention economy.
Voters will scroll past your typical spots. They'll watch ours. And they'll be persuaded.
Want to see mesmerizing content in action? Email data@tavernresearch.com →
