Manual Placements

Manual placements is the option in a Meta ad set where you choose exactly which surfaces the ad may appear on. Meta's help center lists the groups: feeds (Facebook Feed, Instagram Feed, Threads feed, Marketplace, and others), Stories and Reels surfaces, in-stream ads for reels, search results, and Audience Network rewarded videos. Meta recommends Advantage+ placements, but states that you can choose your own placements manually. Not every placement is available for every ad, since objective, format, and region limit the list. Placement groups include feeds, Stories and Reels, in-stream ads for reels, search results, and Audience Network, and each has its own creative needs.

In Ads Manager you turn off Advantage+ placements and tick the surfaces you want. In the Marketing API, the same control sits in the targeting spec, with fields such as publisher_platforms and positions for each platform. Meta's documentation says that if you leave a field empty, it considers all default positions, and that it may add new positions as they become available, so the only way to be sure an ad never appears somewhere is to select placements explicitly. Audience Network cannot be used alone. There is also a soft opt-out in the API that lets up to 5 percent of spend reach excluded placements. For example, a Reels-only test needs vertical video built for that surface, and a feed-only test can use a square or 4:5 image.

Manual selection interacts with creative, budget, and learning. Each exclusion removes inventory from the auction, so the delivery system has fewer chances to find results. That tends to raise cost per result when budgets are small, because the system cannot shift spend to the cheapest surfaces. It also helps when you want a controlled test of one format, such as Reels only, or when creative is made for a single aspect ratio. Features can depend on placements too: Meta's help pages note that some creative options require Facebook Feed to be included if you choose manual placements. Review the setting after any large creative change, because a placement that failed with one ad may work with another.

The mistake people underestimate is narrowing too fast. Excluding Audience Network or Stories after a week of thin data may remove inventory that would have delivered cheap results. Another mistake is using manual placements as a style choice, because Feed feels more professional, without testing the cost. A third is forgetting that excluded placements stay excluded: when Meta adds new placements, a manual ad set will not use them. A fourth is judging placements on last-click conversions while ignoring that Stories and Reels often influence people who convert later in a different place. Exclusions based on brand safety are a different case. If a placement could harm your reputation, cost per result is not the only thing to weigh.

This week, use manual placements for a clear reason: brand safety, a legal restriction, a format test, or a placement with enough results to show it loses. Duplicate the ad set rather than editing a running one, change only the placements, and run both for the same dates with equal budget. Compare cost per result and, where you can, a downstream metric such as cost per qualified lead. Keep the manual version only if it wins by more than normal week-to-week noise. Document which placements are on and why, and revisit the list every quarter, since Meta adds and changes surfaces over time.