Meta is taking placement control away from advertisers. Soon the system decides where your ads run.

What happened

Inside every Meta ad campaign there is a unit called an Ad Set. That is where you define who to advertise to, how much budget, and in which places the ad appears. Those places are called Placements: the Facebook and Instagram feed, Stories, short Reels videos, and also a network of external apps called Audience Network where Meta shows your ads outside its own apps.

Until now you could choose: this place yes, that place no. If you saw Reels bringing you weak leads, you turned them off. Meta's new notice says that option is going away. You will no longer be able to exclude a place, a platform, a device type or an operating system. The automated system will decide on its own where to show you.

Tap to enlarge

What they offer instead is a tool called Value Rules. It is a rule that tells the system "I'm willing to pay a bit more for this place, a bit less for that one." Note the difference: it affects the budget, but it doesn't turn any place off. If until now you simply closed a bad place, that switch no longer exists. You are left only with the ability to hint to the system, not to decide.

On the side, Meta is also removing Messenger Stories as a placement starting August 27, and the budget that was there will be redistributed automatically to other places.

We looked for an official, public notice from Meta. There is no visible help page announcing this: the notice arrives quietly, inside the account itself. And that in itself is part of the story. No date, no orderly announcement, just a change rolling out slowly. Jon Loomer, a well-known ads expert, reported it first, followed by eMarketer and Social Media Today. They all agree on the same thing: this is not a glitch, it is policy.

Back to contents

How much time you have, and why not to wait

First, you can relax: this isn't happening tomorrow. There is no date, and the rollout is slow. But don't confuse "not immediate" with "no need to prepare." There is one more point worth understanding: it isn't clear at all that a manual way to pick placements yourself will survive outside Advantage+‎ (Meta's automated campaign suite, where the system runs almost everything on its own). The automated default might simply become the only option.

And exactly because of that, if you are a decision maker, a founder or a digital person, preparing is not a recommendation. Once the change lands, it will be very hard to rework your whole plan midstream. Better to build now what stays in your control, while there is time and quiet to do it right.

Back to contents

Why this is happening, and it won't be the last time

It's easy to read this as a one-off. That's a mistake. Removing control over ad placement is another step in one direction that has been clear for two years. Meta keeps handing more decisions to its automated system: the choice of who to advertise to, the choice of which ad to show, and now the choice of where. The goal they don't hide is that one day you'll upload a link to your product, and the AI will do the rest. We saw the same pattern in miniature: a catalog that switches on when you duplicate, Autofill filling landing-page forms, and links that send visitors away from your site.

For you the meaning is not "how do I beat this change." The meaning is to stop and think differently. As long as you thought of advertising as a series of campaigns you steer by hand, you depend on buttons that keep disappearing one after another. What stays in your control, and what will decide the outcome, is three things: the quality and quantity of your messages (the creatives you produce), the landing pages you send people to, and the quality of the signals you feed the system. The more the machine decides, the more these are the only elements truly left in your hands, so that is where to invest your thinking, over time, not toward next week's launch.

Back to contents

How it affects you

Not everyone will feel it the same. If you sell a simple product where every buyer is worth roughly the same, say an app with a purchase on day one, the change hurts less. The system brings you buyers, and that's it.

The real pain is for anyone working on leads (inquiries from potential customers) or on SAAS systems (software sold on a monthly subscription). There the difference in quality is huge. The same signup form can bring a serious customer who pays tens of thousands of dollars a year, or a curious visitor who comes once and never returns. Until now, if a certain place brought you mostly curious visitors, you could turn it off. Now that switch is gone too, and the system can fill you with weak inquiries with no way for you to stop it by hand.

And it's true at both ends. A small startup with no big budget, and a large company with many departments each advertising something different, reach the same problem: they don't have enough high-quality "success signals" in each campaign for the system to figure out on its own where to go.

Back to contents

What to do about it: one signal, high quality and fast

You need to understand one thing about Meta's system, and it's simple. You tell it what counts as "success," and it chases that. The professional term for it is a conversion, the action you want to happen. If you define a cheap success that's easy to get, the system will bring you tons of it, at low quality. And once you no longer have the button to turn off bad places, a mistake like that is doubly dangerous.

So the "signal" you report to the system as success has to meet two conditions together:

High quality. Not just a signup or an easy inquiry, but an action that truly predicts a customer who will pay in the end. Two real-world examples:

• A user who signed up and actually performed the core action inside (in product terms this is called activation), not just opened an account and vanished.
• An inquiry that a sales rep (an SDR, the rep who does the first screening) already spoke to and confirmed is relevant. Such an inquiry is called a Sales Qualified Lead, or SQL.

Fast. The success has to happen within a few days. This is the part most people miss. Even if the signal is perfect on quality, if it arrives only 45 days after someone clicked the ad, the system has almost nothing to learn from. It tries to connect the ad to the outcome, and when the outcome comes a month and a half later, the link is lost. A high-quality but slow signal is worth very little to the system.

The biggest problem starts when there aren't enough of these signals. For a small startup it's a matter of volume, there simply aren't many customers a month. For a large, fragmented company it's a matter of spread, each campaign gets a trickle of quality successes, too few for the system to stabilize. The fix is the same in both cases: define the quality success correctly, shorten the time it takes to arrive (for example find an earlier signal that still predicts a good customer), and consolidate small campaigns into one big one to give the system enough examples to learn from.

Back to contents

What to do now

1. Check whether the notice has already reached your account, and write down which places you turn off today. Those are the points that will be exposed when the control is taken away.
2. Define explicitly what your quality success is. Not a "signup," but the action that predicts a paying customer.
3. Measure how long that success takes to arrive. If it's more than a week, look for an earlier signal you can report.
4. Connect your systems (the CRM, where customers and inquiries are recorded) to Meta, so you can report the real success and not just the initial signup.
5. Consolidate small, fragmented campaigns, so each one has enough quality successes for the system to learn from.
6. Until the change lands, use the control that still exists. At the account level (Advertising Settings > Placement Controls) you can still fully block Audience Network, Marketplace and Right Column, and it overrides every campaign. (TheOptimizer)
7. If you still exclude placements at the campaign level, turn off the 5% checkbox by hand. Otherwise up to 5% of the budget per excluded placement keeps leaking exactly there. (TheOptimizer)

Back to contents