Facebook Social Ads: 4 approaches to improved CTR

Social Ads are the Facebook equivalent to a classified ad in your local paper, when it comes to look and feel.  You don't have much control over the format or general appearance of your ad - but you do get to drop in an image, and you do get to craft your own demographic recipe.  

With so few variables under your control, you want to make sure you are using them as effectively as possible.  In a recent campaign, I tested campaign creative versus the school's standard logo image. The ad ran to a very narrow geographic area, to individuals 16 years and older. One version ran with campaign creative and the other ran with the logo - all other aspects were the same.  To get a little deeper, I added a third version that included a related keyword and a fourth version that targeted only high school students.

I was guessing that the creative would pull in more click-throughs than the school logo. I also had medium hopes for the keyword targeting and low hopes for the high school filter.  We know from our SkoolPool analytics that Facebook users leave a lot of profile fields blank, so I thought that from that I'd actually end up missing many high school students who hadn't clicked the official "high school" box on their profile info.

But that's why we do testing! Before I get into the results, here are a few key terms I'll be using:

• Click-through rate (CTR): clicks divided by impressions/displays
• Cost-per-click (CPC): average cost per click
• Cost-per-thousand (CPM): average cost of 1,000 impressions

After 4 days, the school logo version of the ad had a CTR of 4% and the campaign creative version was only achieving 3%.  These were identical ads, with identical targeting, bidding etc - the only difference was the visual.  Therefore with this campaign, the logo achieved 50% more CTR than the campaign creative version. 

The keyword targetting in the third version of my ad narrowed the audience from almost 70,000 users to less than 200.  After 4 days this version had only won 398 impressions - and zero click-throughs.  

The fourth version of the ad was filtered to show only to self-identified high school students.  The high school field is a completely optional profile field on Facebook that some students may choose to select, and others may not.  For example, the starting pool for this campaign was almost 70,000 (16+ years old).  Filtered for high school self-identifiers, the pool dropped to 9,600 (16+ years old).  Take the high school filter off and limit age to 16 - 18 years old and the pool is 12,380. Conclusion: there are about 2,500 or so users between 16 to 18 years of age that are not self-identifying as high school students but may actually very well be high school students.  For the record, the campaign's geography was within Ontario where drop outs lose their driving privileges if they drop out before turning 18. 

The high school filtered version of the ad achieved a 3% click-through rate over the first 4 days of the campaign and had a CPM that was almost half of the unfiltered ads. This ad ran with the campaign creative. Therefore, filtering for "high school" only yielded a 33% increase in CTR (3% vs 2%).  Note: Despite the higher CTR, the "high school" ad had fewer clicks because it had a smaller overall audience size. 

Stats:

Thanks for reading, Thanks to @cfast for keeping me on my toes! Updated 08.27.08

Academica Group Inc.
Full Cycle Marketing for Higher Education™