The one about testing process for Facebook ads

I’ve been thinking about the best process to test new ads on Facebook and Instagram. I mean the practical set-up of building new campaigns, ads sets and ads meanwhile keeping the best-performing ads alive.

Like I wrote about in previous posts, there is a process to keep well-performing ads at a good price by changing audiences or even updating creatives. After spending thousands of dollars / day on ads, I found there are some combinations of creatives and audiences that just outperform everything else over a lifetime of the set-up.

Yes, the prices do rise and sometimes even get too high, but on average, certain combinations are better. I allocate 80% of my budget to these ads. Then I set a 20% of the budget to test anything and everything.

Not sure if this is the best method, but what I do right now is:

  1. Set up new campaigns, ads-sets and ads using AdEspresso. It allows me to create 250 variation of let’s say 5 ads by testing different countries, genders, placements, etc.
  2. If any combination gets a great price at statistically-significant impression-rate, I duplicate the ad-set with the ads and create a new campaign. This new campaign is no longer part of the “testing” budget, but it gets a place in the 80% “good budget”.
  3. Pause the original ad-set in the “testing budget” so I’m not competing with myself.
  4. I keep a close eye on these ad sets for the next few hours because there is a problem. Sometimes when I duplicate an ad-set with ads and everything is the same as the original, prices are not the same.

I’m not sure what’s the purpose for this price disparity. Everything is identical, except for the campaign that holds the ads-set which should not have an effect on the price because all price-related variables are set up at the ad-set or ad level.

I don’t think the prices are so volatile or so random that it would just change for no reason. Rather, I am still trying to learn what causes this. (Please do let me know if you figure it out)

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s