Home > Paid Traffic Sources > Facebook & Instagram

Facebook Traffic Question (12)

12-15-2011 05:24 PM #1 minh1204 (Member)
Facebook Traffic Question

I know this has been discussed to death, but i still don't get it.

1 ad .06 CTR has gotten 100,000 impressions today SUGGESTED $0.67 - 1.71 / CURRENT BID $1.88 / CPC .68
1 ad .09 CTR has gotten 500 impressions today SUGGESTED $0.91 - 2.26 / CURRENT BID $1.88 / CPC .69
1 ad .11 CTR has gotten 900 impressions today SUGGESTED $0.67 - 1.62 / CURRENT BID $1.88 / CPC .96

Seems like the lowest CTR ad has the most impressions I'm not concerned with the CPC so much because I know that will go down over time, but how can I get the .09 and .11 ads to get more impressions?

12-15-2011 05:47 PM #2 mattconlon (Member)

clone them multiple times, and hope that Facebook lets one of them take off. With FB it's more or less luck which ad will take off, you can run 5 of the exact same ads and each will perform differently / receive a different amount of impressions. Put more campaigns up as opposed to more ads within individual campaign

12-15-2011 05:53 PM #3 minh1204 (Member)

so even with an initial decent ctr and bid, fb could choose to not let an ad run?

12-15-2011 06:31 PM #4 polarbacon (Moderator)

Quote Originally Posted by minh1204 View Post
so even with an initial decent ctr and bid, fb could choose to not let an ad run?
first thing is you don't need to get it....you just need a way to get around it and it has been discussed at length (including a case study i did just to remove all doubt from peoples minds)

the exact same ad 4-5 times in the same camp....simple....

second your stats are delayed and what you don't see is that ad had the best adrank of that camp....and as it dumped imps the ctr dropped.....but once fb favors that ad in that camp.....thats the ad thats gonna get 80-90% of the imps.....

12-15-2011 07:42 PM #5 minh1204 (Member)

i should have mentioned that i have each ad in its own campaign and already have 3 duped ads in each campaign. the stats i gave were for the ads that took off in each campaign.

my setup is like this:
camp 1:
ad1
ad2
ad3

camp 2:
ad1
ad2
ad3

camp 3:
ad1
ad2
ad3

12-16-2011 12:25 AM #6 zeno (Administrator)

So, to confirm, you have the following setup:
camp 1:
ad1
ad1'
ad1''

camp 2:
ad2
ad2'
ad2''

etc?

With FB, from my experience, it is the initial CTR an ad gets that causes it to stick and take volume. By that I mean the first click being giving a decent CTR, and the next half dozen clicks keeping the CTR high. When that happens, boom goes the volume. Over time the ad drops in CTR because of the impression volume, so will often drop to lower CTR than some of it's duplicates but retains that lion share of impressions. The other ads would likely do the same if they stuck. The fact that they didn't just means that one specific duplicate satisfied the right conditions to be given the green light by FB.

12-16-2011 12:41 AM #7 minh1204 (Member)

yes that's how i have my ads setup.

so if you had 3-5 dupe ads and 1 took off, do you pause the others that didn't stick or do you just let them continue to run?

12-16-2011 01:05 AM #8 zeno (Administrator)

The fact that they didn't stick will mean they (probably) have a higher CPC so I would kill the weaker ones. I tend to leave a second one sitting there if it has decent CTR. Why? I believe it helps minimise the CTR drop in your 'stuck' ad, and that CTR dropping away can have a big impact on profitability. Basically that 2nd copy takes a bit of the impressions and stops your campaign relying on a single ad. If that good ad falls away, CTR drops and CPC goes back up, the 2nd ad could potentially take over as the more profitable.

12-16-2011 02:27 AM #9 hd2010 (Member)

Quote Originally Posted by zeno View Post
So, to confirm, you have the following setup:
camp 1:
ad1
ad1'
ad1''

camp 2:
ad2
ad2'
ad2''

etc?

With FB, from my experience, it is the initial CTR an ad gets that causes it to stick and take volume. By that I mean the first click being giving a decent CTR, and the next half dozen clicks keeping the CTR high. When that happens, boom goes the volume. Over time the ad drops in CTR because of the impression volume, so will often drop to lower CTR than some of it's duplicates but retains that lion share of impressions. The other ads would likely do the same if they stuck. The fact that they didn't just means that one specific duplicate satisfied the right conditions to be given the green light by FB.
For this kind of setup, should it be done after finding the winning ads ?

12-16-2011 03:08 AM #10 zeno (Administrator)

Probably. There's nothing wrong with it for testing, but I would prefer a setup like this:
Camp 1-6 (split demos, e.g. 3 age ranges + M/F):
ad1
ad1'
ad1''
ad2
ad2'
ad2''
ad3
ad3'
ad3''
ad4
ad4'
ad4''

I.e. test a few images within 1 campaign with duplicates of each. Then once you find the winners you can pause/change bids/scale budget. I prefer to have them all in one tight campaign per demographic at the start because you can open up a bigger budget to the campaign as a whole, whereas if you have separate campaigns for each image and are hitting the same demo, then you will probably need smaller budgets on each. Bigger budget encourages greater delivery, offsets some of FB's ad delivery system crazyness.

12-20-2011 06:04 PM #11 hd2010 (Member)

Quote Originally Posted by zeno View Post
Probably. There's nothing wrong with it for testing, but I would prefer a setup like this:
Camp 1-6 (split demos, e.g. 3 age ranges + M/F):
ad1
ad1'
ad1''
ad2
ad2'
ad2''
ad3
ad3'
ad3''
ad4
ad4'
ad4''

I.e. test a few images within 1 campaign with duplicates of each. Then once you find the winners you can pause/change bids/scale budget. I prefer to have them all in one tight campaign per demographic at the start because you can open up a bigger budget to the campaign as a whole, whereas if you have separate campaigns for each image and are hitting the same demo, then you will probably need smaller budgets on each. Bigger budget encourages greater delivery, offsets some of FB's ad delivery system crazyness.
@zeno : for this kind of setup, won't you think ad1 or ad2 will be taken all the impressions which might leave ad3 and ad4 untested, how you determine which images to test ?

12-20-2011 07:52 PM #12 zeno (Administrator)

Hey, I suppose I do things that way because I use an API and it's easy for me to adjust bids to alleviate the whole impressions skew,so i'm somewhat biased. If you have ~12 ads in a campaign, high bids and a reasonable budget you do usually get a decent distribution across them.

If you really want to do even testing across different images at the same time definitely split it at the campaign level to start off with, i.e. 12 image per campaign as others have suggested.

Home > Paid Traffic Sources > Facebook & Instagram