A hand reaching to press one glowing amber switch in a wall of identical dark switches
Meta

Meta Wants to Choose Where Your Ad Runs. Let It.

Advantage+ placements or manual? Your Meta placement report is a receipt, not a price list, and switching off the expensive placement usually raises your blended cost per result. When restricting delivery is actually right.

Nora BennettPaid Media Strategist, BrandRocket14 min read · September 18, 2026

You open the placements breakdown in Ads Manager, and there it is in black and white. Audience Network is bringing in leads at $1.80. Instagram Stories is charging you $14 a piece. Facebook Feed sits somewhere in between.

The conclusion feels obvious. Stories is the problem. Uncheck it, move that budget to the placements that are working, and the whole ad set gets cheaper.

It is one of the most reasonable-looking moves in the entire platform, and it usually makes things worse. Not because the numbers in that report are wrong. They are accurate. It is because of what those numbers actually are, which is not what almost everybody assumes.

The Checkbox Meta Buried Under "Show More Settings"

Every Meta ad set has to answer one question before it spends a dollar: where can this ad appear?

Meta's answer, by default, is everywhere it can. That setting is called Advantage+ placements, and it used to be called automatic placements, which was the clearer name. With it on, your ad is eligible to run across Facebook, Instagram, Messenger, WhatsApp, Threads, and Meta Audience Network, which is Meta's network of third-party apps and websites. Inside those platforms sit dozens of individual surfaces: the main feed, profile feeds, Marketplace, Reels, Stories, Explore, search results on both Facebook and Instagram, the notifications panel, the desktop right column.

The alternative still exists, despite a run of videos this year announcing that it does not. Click Show more settings in the Placements section and you can pick surfaces by hand. Meta's own help center is direct about which one it wants you on: "It is recommended that you choose Advantage+ placements for your ads because it allows our delivery system to try to make the most of your budget. However, you can also manually choose your own ad placements if you want your ads to appear in specific placements only."

So the control is there. The question is whether reaching for it helps you, and that turns entirely on understanding what your placement report is telling you.

Your Placement Report Is a Receipt, Not a Scorecard

Here is the assumption almost every owner makes, and it is the whole problem: that the cost per result next to each placement is a property of that placement. That Instagram Stories is a $14 place and Audience Network is a $1.80 place, the way one supplier charges more than another.

That is not what those numbers are. They are a record of what the delivery system chose to buy on your behalf, which is a completely different thing.

Your placement report is not a price list. It is a receipt for decisions the system already made, and the prices on it change the moment you take an option away.

Meta's bidding system is not trying to get you the cheapest result in each placement. It is trying to get you the most results overall for the budget you gave it. Those are different goals, and the difference is where the money is. The system looks at every available opportunity across every placement you allowed, ranks them, and buys the cheap ones first. Whatever mix of placements that produces is an output, not a plan.

Meta publishes the arithmetic, and it is worth walking slowly because it explains everything else.

Say there are eleven chances to show your ad. Three are on Facebook and cost $3 each. Three are on Instagram and cost $5 each. Five are on Audience Network, and those are not uniform: three of them cost $1 and two cost $7. You have $27 to spend.

With everything switched on, the system does the sensible thing. It buys the three $1 opportunities on Audience Network, then the three $3 opportunities on Facebook, then the three $5 opportunities on Instagram. That is nine results for $27, an average of $3 each.

Now look at what your report says the next morning. Audience Network: $1. Facebook: $3. Instagram: $5.

Instagram looks like your worst performer by a distance. It costs five times what Audience Network costs. Every instinct you have about running a business says cut it.

Turn Off the Expensive One. Watch the Cheap One Get Expensive.

So you uncheck Instagram. The three $5 opportunities disappear from the pool. Your budget is unchanged.

The money still has to go somewhere. The system buys the three $1 opportunities on Audience Network and the three $3 opportunities on Facebook, same as before. That is $12 spent and six results. It still has $15 in hand and no cheap inventory left, so it starts buying the Audience Network slots it was deliberately skipping before, the ones that cost $7.

You end up with eight results for $26, an average of $3.25. You paid more per result and got fewer of them.

And here is the part that makes people check the report twice: Audience Network, your star performer, now shows a cost per result of $3.40 instead of $1. It did not get worse. You forced the budget into its expensive corner by removing the alternative.

Meta states the lesson in one line, and it is the single most useful sentence in their documentation on this: "one placement's average cost per optimization event being higher than another's doesn't necessarily mean it's inefficient."

The expensive placement was not a leak. It was the thing keeping your budget away from worse inventory somewhere else.

The placement that looked expensive was not costing you money. It was the reason the rest of your spend looked cheap.

This is the same trap as a cost cap set too low, and it fails for the same underlying reason: a constraint you place on the auction does not make the auction cheaper, it just shrinks the set of things your money is allowed to buy. Fewer options is not a discount.

The 11.7% Number, and Who Ran the Test

Meta's pitch for all this is a specific figure. On their Advantage+ placements page they state that "in an experiment, ad sets using Advantage+ placements delivered an 11.7% lower cost per action (CPA) on average compared to ad sets using manual placement settings."

You should know exactly what that number is before you lean on it. It is Meta's own experiment, reported by Meta, with no published methodology, no sample size, no date, and no description of what the manual ad sets were doing. It comes from the company that benefits when you open your ads to more of its inventory. That is not an accusation of dishonesty. It is a description of who is holding the measuring tape.

So do not believe the article because of the 11.7%. Believe it because of the arithmetic in the last two sections, which you can check yourself and which does not require trusting anybody. Nine results at $3 versus eight at $3.25 is not marketing material. It is the mechanical consequence of how a bidding system spends a fixed budget, and it would be true whether or not Meta had ever run a study.

The stat is corroboration. The auction logic is the argument.

When Turning a Placement Off Is Actually the Right Call

None of this makes manual placements wrong. It makes them a tool with a narrow job, and there are three situations where reaching for it is correct.

You have a hard constraint, not a preference. Some businesses genuinely cannot appear in certain contexts. A regulated financial services firm, a healthcare provider with compliance obligations, a brand with contractual placement restrictions. That is a real reason, and "my ad must not run in third-party mobile games" is a business rule, not an optimization. Treat it as one.

One surface genuinely breaks your creative. Not underperforms. Breaks. If your only asset is a wide horizontal image, it will be cropped into uselessness in a vertical Stories slot, and no amount of delivery intelligence fixes a picture that has lost its subject. But read the next section before you conclude this is you, because it usually is not.

Your product is device-specific. If you sell an iOS app, showing the ad to Android users is not inefficiency, it is waste with a zero percent ceiling. The settings under placements let you edit devices and operating systems directly. Worth knowing before you do: Meta's documentation states that opting out of mobile or desktop devices will itself trigger Advantage+ off.

Notice what is not on that list: "this placement has a high cost per result." That is the one reason that does not survive the arithmetic.

A hard constraint is a business rule. A high cost per result is a hunch wearing a business rule's clothes.

Meta is unusually candid about the trade-off in their best-practices doc. Their recommendation is to remain "opted into all placements," and where they do describe using placement controls, they attach the warning directly: doing so "may limit the effectiveness of Advantage+ placements." The platform is telling you the cost of the control while handing you the control.

Unchecking One Box Turns the Whole Thing Off

This is the detail that changes the decision, and it is stated plainly in Meta's own instructions for choosing placements: "If you exclude a placement in any ad set, you'll trigger Advantage+ off."

It is not a dial. It is a switch. You do not get Advantage+ placements minus Audience Network. You get manual placements, with whatever you happened to leave checked, and the delivery system stops optimizing across the full set. People assume they are making a small adjustment to a smart system. They are turning the smart system off and inheriting the job themselves.

There is one middle setting almost nobody knows about, and it is worth finding. When you exclude placements, Meta offers a checkbox called Allow limited spend to excluded placements. Per their documentation, it "allows up to 5% of your budget to be spent for each excluded placement when it's likely to improve performance," and you can manage it placement by placement.

That is a genuinely sensible compromise for the nervous. You keep a surface mostly switched off, but you leave the system a small window to prove you wrong with real money rather than leaving you to argue with a report. If you are going to restrict something on a hunch, restrict it like that.

The Control Almost Nobody Knows Sits at the Account Level

If you do have a genuine hard constraint, there is a much better way to enforce it than unchecking boxes in every ad set you will ever build.

Meta's advertising settings include account-level audience and placement controls. Set an exclusion there and it applies to all current and future campaigns in that ad account, automatically, without anybody having to remember. For a compliance requirement, this is the correct instrument by a wide margin. A rule that depends on a human remembering to uncheck a box at 6pm on a Friday is not a rule, it is a hope.

And here is the part that makes it strictly better than unchecking boxes in the ad set, which Meta says quietly in a single line: these account-level settings "are applied even if you're using Advantage+ placements." You get the hard constraint and you keep the optimization. Enforce it at the account and the delivery system goes on doing its job everywhere you have not fenced off. Enforce it in the ad set and you switch the whole thing off to solve a problem in one corner.

If your constraint is narrower than a whole placement, look at brand safety controls too. Facebook in-stream video can run inside partner live streams, and those can be excluded for specific ads and campaigns or across the entire ad account without touching your placement settings at all.

It is also the honest test of whether your constraint is real. If you are willing to enforce it account-wide and permanently, it is a business rule. If that feels too drastic, then what you actually have is a hunch about performance, and the arithmetic above already told you what to do with it.

Most "Bad Placement" Is Just a Wrong-Shaped Ad

Here is where the real money is, and it is not in the placements panel at all.

When a placement shows a genuinely terrible cost per result, the most common cause by far is not that the audience there is worthless. It is that your ad arrived in a shape it was never built for. A 4:5 feed image dropped into a 9:16 Stories slot gets its top and bottom filled or cropped. Text that was comfortable in the feed ends up under the interface. The ad technically ran. Nobody could read it.

That produces exactly the symptom that makes owners uncheck the box, and unchecking the box treats the symptom. The placement was fine. The asset was wrong. We have written the whole playbook on this: a story ad is not your feed ad resized, and the fixes are specific and cheap.

Meta's own guidance points the same direction. Their best-practice advice is to pair all placements with diverse assets, supplying your creative in multiple aspect ratios so the system has a properly shaped version for whatever surface it wins. Give it a square, a vertical, and a horizontal, and the question of whether Stories "works" mostly answers itself.

Before you decide a placement failed you, check whether you ever gave it an ad it could display.

This is also why the instinct misfires specifically for people who know Google Ads well. On Google Display and Performance Max, going into the placements report and excluding junk sites is correct, necessary, and something almost nobody ever does. Same word, opposite advice, because the mechanics are not the same: there you are excluding genuinely low-quality third-party inventory from an open web network, here you are removing surfaces from a closed system that is already optimizing across them. Carrying the Google habit into Meta is one of the most common ways a competent advertiser makes their Meta account worse.

What to Do Monday Morning

Open one ad set and work through this in about ten minutes.

  1. Check whether Advantage+ placements is on. Placements section, look for the Advantage+ label. If someone has been hand-picking surfaces, find out why, and whether the reason was ever anything more than a report that looked bad.
  2. Stop reading the placement breakdown as a scorecard. Read your blended cost per result for the whole ad set instead. That is the number attached to your bank account. The per-placement split is diagnostic color, not a to-do list.
  3. Write down any genuine hard constraints. If you have one, set it at the account level in advertising settings so it holds everywhere, then stop thinking about it.
  4. Upload your creative in three shapes. Square, vertical, horizontal, for every ad. This is the highest-value thing on this list and it has nothing to do with placements settings.
  5. If you turned placements off previously, turn them back on and give it two weeks. Compare blended cost per result across the whole period, not day two, and not placement by placement.

The broader pattern is worth holding onto, because it keeps showing up in this platform. Meta has spent several years removing dials and replacing them with systems, and the reflex of a careful operator is to grab whatever dials remain and hold on tight. Sometimes that reflex is right. With placements it is not, because the dial does not control price. It controls how many things your budget is allowed to consider, and every option you remove is one the system can no longer buy cheaply on your behalf. The same logic governs the bigger version of this decision, handing Meta the entire campaign, where the answer is genuinely more complicated.

If you would rather have someone watch the blended numbers and tell you when a constraint is actually earning its keep, that is what we do all day. And if you would rather run it yourself, the arithmetic above is the whole secret. We are around either way. If it would help to have another set of eyes on the account, we run Facebook and Instagram ads for small businesses and we are happy to look.

Nora Bennett · Paid Media Strategist, BrandRocket

Paid media strategist at BrandRocket. Spends her days inside Google Ads and Meta accounts, helping small businesses get more out of every dollar they spend.