I post a lot about bad technology – I’m looking at you Zillow and Redfin – so I felt like it was time for a look at some really good software that we’ve started using at Virgent.  We’re big believers in Facebook advertising, and we’ve always struggled with the best way to optimize our ads. The standard Facebook interface just isn’t very good and helping marketers identify which ads are performing poorly and which factors are driving the efficacy of each ad group.

Enter AdEspresso. It’s changed the way the way we set up and optimize Facebook ad campaigns. In just the week we’ve been using it (We’re still on the free trial period), my CTRs have increased 8x. Before you ask, I’m not being paid for this post, and can’t even remember how I stumbled on their site in the first place, but the results are clear:

Guess when we started used AdEspresso

Guess when we started using AdEspresso

AdEspresso allows you to A/B test hundreds of versions of an ad with almost the same ease of setup as the standard Facebook interface. The software allows you to enter multiple versions of every piece of a Facebook ad – the headline, the text, the images, etc – and then it automatically create hundreds of distinct ads with all the possible combinations of those factors, along with other factors like gender, ad placement, and other targeting criteria. A typically AdEspresso campaign may result in more then 200 distinct ads.

Optimization is the real benefit. Because each ad only targets a specific subset of your criteria, AdEspresso can statistically compare what factors lead to higher clickthroughs and automatically optimize your ads over time. The results may surprise you. Within two days, AdEspresso was able to quadruple our CTR by showing us that one ad we were running performed best with men, with a specific headline, when not placed on the right column of Facebook.

I’d encourage anyone who does Facebook advertising to give it a look. They offer a two week free trial, but we’re definitely planning on paying for a subscription once the trial ends.