Two Meta campaigns with the same targeting reach the same people only if they also optimize for the same thing. Targeting decides who is eligible to see your ad. The optimization event decides who Meta actually picks from that pool. Give two ad sets different goals and they can end up in front of largely different people.
On a mentoring call, an operator selling four online programs to the same market asked a sharp question: if one campaign runs videos and another runs images, both with identical targeting, are they shown to the same people? Our answer: yes if everything else is the same. But if one optimizes for lead submissions and the other for video views, expect very little overlap.
Targeting builds the pool. The auction picks the person.
Meta runs an auction every time there's a chance to show someone an ad. Every ad whose target audience includes that person is eligible to compete. The winner is the ad with the highest total value, which Meta describes as a combination of three things:
- Bid. What the advertiser is willing to pay for their outcome.
- Estimated action rates. The probability that showing this ad to this person leads to the outcome the advertiser wants.
- Ad quality. Signals such as feedback from people viewing or hiding the ad.
The outcome in the middle is the optimization event. Meta calls it the outcome its system bids on in the auction, and it can differ from your campaign objective. The system learns to show ads to the people most likely to produce it. So the same person can be a strong match for your video view ad set and a weak one for your lead ad set. Same audience, two different predictions, two different sets of people.
What it means for testing and structure
- Don't compare across goals. A lead campaign beating a video view campaign tells you about the goal and the people it reached, not the creative. To test images against videos, keep the optimization event, audience, budget, destination URL and call to action the same.
- Use the A/B test tool for a clean read. Meta says its A/B tests show each version to a separate segment of the audience so nobody sees both, while informal tests can lead to overlapping audiences. More in why two ad sets in one campaign is not an A/B test.
- Same audience and same goal means competing with yourself. Meta lists high auction overlap as one cause of learning limited ad sets, and suggests combining ad sets and campaigns. See why one Meta campaign beats five.
Four programs, one market
For the four programs, our suggestion was one campaign with an ad set per program, but with targeting that differs a little for each, so each program's creative goes to the people most likely to want it: sports and fitness interests for one, yoga and health interests for another. If the targeting really is identical, use one ad set with all the creatives instead. Even then, Meta shows the ad most likely to get the lowest cost per optimization event for each person, so the creatives won't get equal delivery.
The takeaways
- Targeting decides who is eligible. The auction, driven by the optimization event, decides who sees the ad.
- Same audience with different optimization goals can reach largely different people.
- Never judge creative by comparing campaigns with different goals.
- For a clean comparison, change one variable and use Meta's A/B test tool.
- Same audience and same goal across ad sets means you're bidding against yourself.
Common questions
Do two Meta campaigns with the same targeting reach the same people?
Not necessarily. Targeting decides which people are eligible for your ads. If the two ad sets optimize for different events, Meta ranks each against a different prediction of who will take that action, so they can reach largely different parts of the audience.
What does the optimization event change in Meta delivery?
Meta calls the optimization event the outcome its system bids on in the auction. It learns to show ads to the people most likely to produce that outcome, which can differ from your campaign objective.
Can I compare a lead campaign with a video view campaign to test creative?
Not fairly. Each is shown to different people, chosen for different actions, so the result reflects the goal as much as the creative. Keep the optimization event, audience and budget the same, and use Meta's A/B test tool, which shows each version to a separate segment so nobody sees both.
What if two ad sets have the same audience and the same goal?
Then they compete for the same people. Meta's learning limited guidance lists high auction overlap as a cause and suggests combining ad sets and campaigns. If the targeting is identical, one ad set with all the creatives is usually simpler.