Ad targeting is probably the most important and least understood component of Facebook/Meta ads. Here’s a break down of how it actually works.
🎯 Targeting Setup
Targeting defines the audience who will see your Meta ads. Think about who you want to reach, but also think about how Meta can best identify those people.
You have traditional targeting parameters like demographics (age, gender, etc), interests, and location. But you also have data-derived audiences like Look-a-Likes and dynamic audience lists based off of your 1st party data.
Targeting parameters can be combined using AND/OR Boolean logic. Combine settings with AND to narrow your audience. Combine settings with OR to expand your audience.
You want your audiences to be sufficiently large so that you only reach 10% of your audience each month. More than that, and ad costs will go up.
🎯 CBO: Head-to-Head Targeting Tests
Campaign Budget Optimization (CBO) lets you set up multiple audiences in a single campaign (each in its own ad set) and have them compete over a shared budget. This ensures that the best performing audience gets the most spend.
That’s like playing multiple hands of blackjack at once, getting to see all the cards for each hand, and THEN placing your bet. This would ensure you always bet more on winning hands. It’s cheating at the casino, but simply smart budget management within Meta.
Meta will decide how much budget each ad set should get based on how efficiently it meets the campaign goal. As one audience fatigues, dollars will automatically shift to the next best performer.
🎯 Audience Optimization
Whatever objective you set as your campaign goal will be the singular focus of how a campaign optimizes. Set it to a traffic objective, and the ad will find the cheapest clicks. Set it to a sales objective, and the ad will find people with the lowest cost per conversion.
Every conversion you get trains Meta’s learning model. Once you get 50 conversions, the learning model knows exactly who to find more of. Then it will accurately predict who is most likely to convert and will show more impressions to those people.
If your audience is too small, Meta will be forced to reach people who aren’t very likely to convert. Or performance will rapidly decline when Meta runs out of people to find.
If you get too few conversions, Meta won’t have a clear picture of who to optimize to.
🎯 Summary
Ad targeting is probably the most important and least understood component of Facebook/Meta ads. Here’s a break down of how it actually works.
🎯 Targeting Setup
Targeting defines the audience who will see your Meta ads. Think about who you want to reach, but also think about how Meta can best identify those people.
You have traditional targeting parameters like demographics (age, gender, etc), interests, and location. But you also have data-derived audiences like Look-a-Likes and dynamic audience lists based off of your 1st party data.
Targeting parameters can be combined using AND/OR Boolean logic. Combine settings with AND to narrow your audience. Combine settings with OR to expand your audience.
You want your audiences to be sufficiently large so that you only reach 10% of your audience each month. More than that, and ad costs will go up.
🎯 CBO: Head-to-Head Targeting Tests
Campaign Budget Optimization (CBO) lets you set up multiple audiences in a single campaign (each in its own ad set) and have them compete over a shared budget. This ensures that the best performing audience gets the most spend.
That’s like playing multiple hands of blackjack at once, getting to see all the cards for each hand, and THEN placing your bet. This would ensure you always bet more on winning hands. It’s cheating at the casino, but simply smart budget management within Meta.
Meta will decide how much budget each ad set should get based on how efficiently it meets the campaign goal. As one audience fatigues, dollars will automatically shift to the next best performer.
🎯 Audience Optimization
Whatever objective you set as your campaign goal will be the singular focus of how a campaign optimizes. Set it to a traffic objective, and the ad will find the cheapest clicks. Set it to a sales objective, and the ad will find people with the lowest cost per conversion.
Every conversion you get trains Meta’s learning model. Once you get 50 conversions, the learning model knows exactly who to find more of. Then it will accurately predict who is most likely to convert and will show more impressions to those people.
If your audience is too small, Meta will be forced to reach people who aren’t very likely to convert. Or performance will rapidly decline when Meta runs out of people to find.
If you get too few conversions, Meta won’t have a clear picture of who to optimize to.
🎯 Summary
The secret to setting up high performing audiences is to strike a balance between efficiency and scale. Feed the system variety with sufficiently large/diverse audiences so you have more chances to find a high performing target. Make sure you have a high quality data connection with high match rates to ensure your conversions get matched to people and train the learning model.The secret to setting up high performing audiences is to strike a balance between efficiency and scale. Feed the system variety with sufficiently large/diverse audiences so you have more chances to find a high performing target. Make sure you have a high quality data connection with high match rates to ensure your conversions get matched to people and train the learning model.