has anyone else experimented with FB's new option to "manually bid on conversions"? i have tested some different payout offers over the last few days and seen very mixed results. also if their algo realizes/thinks the CPA bid is too low i have often stopped getting traffic altogether on a campaign.
would like to hear any other experiences if there any.
I'd also be interested to hear if anyone's tested this out yet, and if so how it did.
I'm no expert with Facebook or media buying in general, however I do have some data to share. Not sure if it's helpful.
I'm the business/funnel owner here, so I know we make about $1.30 per FB opt in over the first 30 days. That's my target CPA.
I used oCPM with the action being the pixel fired on the opt in thank you page. Here is some data:
http://imgur.com/1wtwEaP
The different rows are for different ad and placement combos (we tested a couple ad variations and both newsfeed and rhs)
Any critique would be welcomed.
As a side note I changed some campaigns to optimize CPM for the initial purchase (not opt in.) I figure this could save me the effort of looking at the value per opt in per ad...but now I'm realizing that if sales are trickling in for the first 2 weeks someone gets on the list, there's a delay on that data and I'm not sure how that reflects in FB's algo.
ocpm is used when you ALREADY have conversions.
then fb will auto optimize for THOSE ppl.
start with cpc.
get conversions.
then switch to oCPM
I wanted to try this but then again I dont want Facebook knowing where im ultimately going to get conversions so I have not tested it.
My hope is that it weighs cpm bids in ways that we're not able to manually target for (eg psychographic factors) depending on conversion data.
I also assumed that before conversions reach a significant amount they would use website clicks as a proxy.
I'm testing will let you guys know