Bid performance on Facebook: oCPM v. CPA, revisited
I love the saying, “There aren’t any expert marketers, only experts in testing.” What worked yesterday in digital marketing may not perform today. This is especially true with Facebook as the advertising landscape sees huge changes from year to year. This is a point I always drive home to my clients. Test, analyze, and retest, and eventually you will end up with a killer strategy. Then, do it again…
In the spirit of testing, I want to revisit an old subject; oCPM v. CPA bidding on mobile app install ads. Facebook is no doubt tweaking these algos. Mobile app install ads are obviously key to Facebook’s strategy for the future, especially in light of the recent unveiling of the Audience Network.
For this test, we ran 2 ads with identical creatives, targeting, and budgets. The only factor we tested was the bid type, one with oCPM and the other with CPA. I expected to find similar performance between the two. However, I was a little surprised by the results:
We have a clear winner here. 43% savings on an install simply by switching the bid type is the reason we continue to test everything. This data tells me two things: 1) Facebook still has some work to do before their ad products are ready for amateurs to completely manage their own campaigns (efficiently); and 2) we should anticipate further tweaking as Facebook’s mobile ad offerings continue to evolve.
The point is simple: as performance marketers, we must continue to test and leverage stats to our advantage.
What are some factors where you have found that most affect performance?