Facebook Ad Test

Written by Peter Day

Lead Generation Strategies

How To Do Facebook Ad Tests Effectively For Affiliate Marketing Lead Generation

There are many ways to run Facebook ads. Any time you hear someone talking about Facebook strategies, keep in mind that there are always multiple strategiesso do your research!

This is one of mine that I’ve named “Creatives -> Audiences -> Creatives”

The example I’ll use is this: You heard about a good $20 lead gen offer in car loans, and you want to try it for the first time.

Phase 1: This phase is about Creatives.

Pick audiences that you think make the most sense (broad, age group, interest, lookalikes, etc) and structure your campaign the following way:

Create 1 Campaign -> Create 1 Adset (1 audience structure) -> Create 3 completely different Facebook ads (different images, headlines, copy, etc).

Basically, you’re not choosing an audience as a final decision.

You just pick one audience to test 3 Facebook ads with.

Set a budget of $100-$150/day to start with.

This is a guideline.

Any lower than $100 will just prolong the testing time for too long.

Later, you can raise it if you see good results, or if you are bold, you can run 3-6 ads or with a higher budget to start with.

After 4-5 days you will see which of the 3 ads wins overall in the Facebook Pixel conversion rate.

We will mark that one ad as the winner of the 3.

Phase 2: This phase is about Audiences.

We will do a bunch of audience testing in this phase at the adset level.

Raise the budget, e.g.: $200/day

Take your winning ad from Phase 1 and create about 10 different audiences with $20/day for each.

For example:

  • Adset 1: Broad audience (meaning: all genders, age 18-65+)
  • Adset 2: 1% lookalike audience
  • Adset 3: 1 interest audience
  • Adset 4: 2 interest audience
  • Adset 5: 3 interest audience
  • Adset 6: 5% lookalike audience
  • Adset 7: Narrow audience (meaning: two or more criteria). On Facebook, under “Interests” click on “Include.”
  • Adset 8: Ages 20-35, all genders, specific location
  • Adset 9: Ages 40-65+, all genders, specific location
  • Adset 10: All women
  • Adset 11: All men

Notice that I ended up adding an 11th audience category; 10 is used as a recommended guideline, however, it can be as fluid as the offer requires.

Run it for between 1 & 2 weeks.

As you had selected a single ad from Phase 1, this acts as a control. You did not run any new ads, only added multiple audiences for that one ad.

Phase 3: This phase is about Creatives again.

From Phase 2, now you have the top 3 best performing audience categories. Let’s say it was Adset 4,9,11. Pause all the other adsets.

Important: Don’t delete adsets (or anything) on Facebook as you will need those stats.

Just pause them.

Now you can create multiple ads for each of these 3 audiences (4, 9, & 11).

Natural selection is at play again: over time, you will see which ads work best for which specific audience.

Select the top one in each ad for each of the 3 audiences that perform the best and pause the rest.

Process Flow overview:

In Phase 1 you got one winning ad.

In Phase 2 you used that ad and created many adset variations.

In Phase 3 you got the top 3 winning adsets and created multiple ads for each.

In the end, you will have a winning ad in each adset.

e.g. in bold, giving you the final set to keep and run:

Adset 4: 2 interest audience (Ad 1, Ad 2, Ad 3, Ad 4, Ad 5)

Adset 9: Ages 40-65+, all genders, specific location (Ad 1, Ad 2, Ad 3, Ad 4, Ad 5)

Adset 11: All men (Ad 1, Ad 2, Ad 3, Ad 4, Ad 5)

It all boils down to offers.

The strategy itself can be perfect, but it means nothing if your offer is not good enough in the first place.

I.e.: a good car loan offer should get at least $1-$2+ EPC. If you only have a $0.42 EPC offer, then it’s not even worth experimenting and blowing all this money.

A good offer should start to break even in Phase 2 or so.

Sometimes the opposite happens: You get a really good offer that runs well even with a terrible ad.

In that case, you may not need to do this campaign structure at all, or do it partially (Phase 1 and 2).

The basic premise of the whole process is to always consider the offer first. Make sure you partner with networks that have good offers!

More questions?