Some real-life experiences of Facebook ads


1. Warm up your account


A fan page that has never posted and has a high level of interaction with users, the cost of a thousand impressions can be much higher.


Facebook wants you to leave users on the site


So the best way to warm up your account is to do something that won't get people off Facebook, and it's highly engaged.


Enter a niche related to your product to Youtube, such as cars and sort the most look...
Or the funniest car video... and then sort by most of the time


Upload the video to your Facebook fan page, post as a post, and then post PPE, and you can also watch the video: Wide lying like car people:


you can send pictures to do the same thing


2. Some people who sell courses like to say, "You need at least 25,50 conversions to really tell whether an ad is a winner or a loser."


That's completely Bushit.


In the years I bought paid advertising, whenever I had a winner, I was almost the winner from day one, and if the first three days didn't translate, the chances of a successful turnaround were almost zero, unless you wanted to make a brand that could lose money for a long time.


3. If the ad runs well effect suddenly waist cut, look at your fan page customer ratingsIf less than 2 is more than half this reason.So far no solution, want to do long-term business, control product quality and service, open an account when more than a few fan pages.



An experiment cost $9,000.


Ecom Products


United States


1. When we use AUTO BID, we get the lowest CPA instead of Target CPA.
2. When the largest audience (3 million or more) is attended, the cost per thousand presentations is the lowest.
3. The cheapest cost per thousand shows is not related to getting the minimum cost per conversion.
4. Super wide (whole U.S., 40 plus F) product-related keywords subdivided into device-by-device brands (iPhone, android, etc.) to get the lowest CPA
5. The average video viewing time is not as long as the better, the top three ads average video viewing time (28s, 22s, 25s) the average video viewing time (25s, 22s, 27s) so the average viewing time has little to do with more conversions.
6. Of the $9,000 or more spent, only $350 was spent on the most likely sleep time for the target audience, a time period that showed very little, but interestingly, CTR increased by an average of 20-40% at night.
7. The impact of the number of comments/shares on ad performance is uncertain, as the top 3 best performers and the worst performing top 3 receive very similar share/likes.
8. Age groups, it seems that they are really random.
9. The best targeting method is keyword, followed by 1% LLA profit/video views.
10. Don't use manual bidding on Facebook, in the case of CBO, unless my target is Lookalikes, when pixel sales exceed 1000, or small budget tests can use CBO.Otherwise there is no chance of winning an automatic bid.