I’ve spent over $9400 dollars on Facebook Paid Ads. Here are some notes that answer many, many questions you and people might have in regards to what happens IF. What happens WHEN. I’ve done the work, invested my own cash, money (no credit card, no loan, not agency money) to get the answers.
I feel like MYTHBUSTERS. Confirmed, Plausible, or a Myth.
By the way, a few ‘smartiepants’ love to say “OH you need at least 25, 50 conversions to be able to really tell if an adset is a winner or loser.”
That is utter BULLOCKS. (I love that Old-English word btw)
In my 6 years of buying paid ads, every single time when I had a winner, it was a winner from day one -and all my affiliate friends that run campaigns will confirm this. Don’t listen to fake fake gurus that don’t run any campaigns just say they do. DATA doesn’t lie. Here are the screenshots, and below you will find what the data says.
-Watched 95% of video LLA’s
-Targeting Broad + Engaged Shoppers
1. Lowest CPA was acquired when we used AUTO BID, and not bidding high as fuck with offer payout x3 target CPA.
2. CPM was lowest when the audience was the biggest (3 million or more)
3. Having the cheapest CPM had zero to do with getting the lowest CPA
4. Going super broad (entire usa, 40+ F) with product related keywords and breaking it down by device brand (iphone, android, etc) got the lowest CPA
5. The average video watch time was NOT the highest for the top 3 best converting ads sets with lowest CPA (28s,22,28s)
6. The average video watch time with the worst adsets were (25s,22s,27s)
Therefore there’s no coorelation that avg watch time has anything to do with more conversions.
7. Cost per 3 seconds video view for the top 3 best adsets with lowest cpa – 9c,10c,7c
8. Cost per 3 second video view for the worst 3 adsets with highest cpa – 5c,7c,6c
The cost of video views was lower on the lower performing adsets.
9. CTR (ALL) was 8.5% or better for the top performers vs below that for the worst performers.
10. Out of the $9000+ spent, only 350$ was spent during the time the target audience was most likely sleeping. And it rendered one conversion at a super low CPA of $7. Impressions were very few in between, but interestingly enough CTR jumped 20-40% during the night on average.
11. Post Comments/Shares counts were inconclusive as the top 3 performers, and the worst 3 performers got very similar ammount of post shares/likes.
12. Age groups, it seems like they are truly random.
13. Best method was targeting keywords, followed by 1% LLA of profits/video views.
PS: I have also attached screenshot showing how bidding 3x,4x,5x etc offer payout will mean you’r going to pay very high CPA whereas if you use Auto get the lowest price per conversion full get the lowest CPA — i ran a LOT of data for 3 days didn’t just do a $5 test as you can see