How to Mine for Quality, Profitable Users on Facebook
Published: June 25, 2014
Author: Bill Cleere
I heard a telling question regarding Mobile App Install ads on Facebook at a recent trade show. An advertiser perusing the booths asked a rep from a prominent PMD:
“So do you just bid OCPM and let Facebook do the work?” The PMD rep’s answer was dead-on:
“Sometimes that’s where you start, then you begin to mine the gold.”
That’s exactly the way I look at mobile install ads. Let’s face it, Facebook experts can do wizardly tricks with spilt-testing, drill-downs, iterations, and creative swapping. However, the first serve of an ad may not be ready for all that yet, because on Facebook, sometimes the chicken comes before the egg. Many Advertisers have proprietary algorithms, predictive tools, and targeting experts to optimize their campaigns, but sometimes it’s best to let Facebook chop through the brush for you a little before you start pulling the fruit.
Recently we ran a campaign for a branded game that yielded over 3,000 installs for an average CPI of $1.40. Our first target was simply all genders, all ages, and a couple of very broad targets based around the brand being advertised, with an audience size of around 40 million. We let Facebook use its secret-agent intelligence to go out and find likely installers and grab us some cheap installs to start.
Then, as a little more budget was spent, the CPI started to rise. At that point we jumped in to find where the real winners were. Using 3rd-party analytics, we started to see who was spending in the game on in-app upgrades. Combining that with gender, location, creative, and age analysis, we iterated the second wave of ads. We analyzed the targeting, used our algos to find similar audiences, and drilled down to an audience of about 15 million. A couple more waves like this, and we began to see the CPI come under $2 (client’s target was $2.50), and ultimately to the $1.40 previously mentioned.
It all began by handing the keys to Facebook and letting it grab impressions, likely installers, and most importantly…data! With just a little bit of insight from Facebook’s initial dive, we were able to start pulling out the big fish, and ultimately, those in-app spending whales.