Hi Karen,

My apologies for the slow response. I've been mostly offline for the weekend.

In short, I tend to turn off elements if it's just an image or line of text that is not doing well in a campaign that otherwise is performing well.

I tend to turn off and start again with the best performing elements when things are too expensive, or I need to make changes to the demographic info that I am targeting. It sounds like you should be turning things off and starting again with a much more refined campaign.

Keep in mind that you can still split test multiple images without using dynamic creative. To do this you just add multiple images without the dynamic creative toggle turned on and that will effectively create identical ads within the same ad set, with all variables being equal except the images. Facebook will favour the better performer.

I think you;re on the right track. Just pick the best performer from each creative element and refine your targeting based on performance.

If you can comfortably go up to $10/day then you can try conversions. You will likely be in range. But it will probably take 3 - 5 days to find out depending on how well the first day or two performs. If two days pass and you have something dismal like $10/sub, Then turn the campaign off.

Otherwise, I think you're on the right track.

Hey Karen,

Just to be clear, while it might be a tiny bit easier to do the traditional image split testing, I just want to point out that it is a totally different thing. 

With Dynamic Creative you are shuffling a ton of different elements into many different combinations and getting quick data that you can use to refine your ads. With traditional image split testing and running different ads within the same ad set, they are fixed combinations. So the data you get about which ad is winning is more accurate, but it also takes longer to get the data.

Long live the silly!

Hey Karen,

It's faster in the sense that when you are starting a new campaign with Dynamic Creative you get a ton of data very quickly and you can hone in on high performing headlines, images, body copy, call to action buttons, and link descriptions. And within a few days you are usually on track with an ad that is at least doing ok. 

When you use more traditional split testing you are typically just changing out one element at a time. Usually images or targeting. So it takes longer to start from scratch and find yourself with a campaign that is doing reasonably well. However, the traditional split testing is more accurate because you are not shuffling the elements all manner of ways. You are truly testing complete ads with one variable changed. 

The idea is to use Dynamic Creative to get you in the ballpark, and then traditional split testing to trim the remaining fat. That make sense?