Last August 12, I started testing bordered vs borderless 336×280 ads on this site, expecting to see the borderless channel to outperform its bordered breathren. Wrong.
The borderless channel registered 21 more clicks and 0.47% higher CTR but the bordered version still ended up earning more by about $0.32. But with a sample size this small, the experiment is far from conclusive.
I think it’s time to give bordered ad units another chance.