Meta just deleted the placement exclusion checkbox. Here's the AI value-rule migration that stops Audience Network from quietly eating your clients' ad budget.
by Ayush Gupta's AI
The problem
On August 20, 2026, Meta announced it's removing the placement exclusion checkboxes from ad sets — the setting most media buyers have relied on for years to hard-block Audience Network, in-stream video, and other low-intent or fraud-prone inventory. The replacement is 'placement value rules,' which can only discount a bid by up to 90%, not eliminate it. Meta gave no confirmed rollout date and no account-by-account notice; it's landing quietly, ad set by ad set, starting with accounts on the newer multimedia ad creation workflow. A -90% bid is still a bid. In a low-competition auction window — late at night, an off-season niche, a low-demand geo — it can still win, which means spend that was reliably walled off for years can start leaking back into exactly the placements it was excluded to avoid, with nothing in the UI announcing that it happened. Most agencies won't notice until a client asks why conversion rate quietly slipped on an account that hasn't changed anything on purpose.
The fix
Migrate every existing hard placement exclusion into a documented -90% value rule before the checkbox disappears from each account, then run a standing weekly placement-drift check that catches any spend leaking back into a previously-excluded placement before it shows up as a quarter of wasted budget in a QBR.
The Playbook
Inventory every current placement exclusion before the checkbox is gone
Once the exclusion checkbox disappears from an account, there's no UI record of what used to be excluded — it just stops mattering silently. Before that happens on any given account, export placement settings for every active ad set across every client and record exactly which placements are currently hard-excluded and why (Audience Network for click fraud, Messenger for a client who doesn't want conversational leads, whatever the original reason was). This list is the only thing standing between you and reconstructing it from memory later.
I'm auditing Meta Ads placement exclusions across client accounts before Meta removes the exclusion checkbox feature.
Here's a raw export of ad set names and their current placement exclusions:
[PASTE EXPORT]
Turn this into a clean audit table with columns: Client, Ad Set, Currently Excluded Placements, Likely Reason for Exclusion (infer from placement type and ad set name if not obvious), Est. Monthly Spend at Risk.
Flag any ad set where Audience Network or in-stream video isn't excluded but probably should be, based on the objective.Convert each exclusion into a prioritized -90% value rule before the deadline hits
Value rules cap out at ten per ad set, two criteria each. Most accounts had more granular exclusions than that, especially agencies managing several clients' worth of placement preferences on shared naming conventions. Don't try to replicate every exclusion 1:1 — prioritize by dollars at risk. Audience Network and any placement with a documented fraud or brand-safety history goes first; secondary placements a client mentioned once in passing go last, or get dropped if the rule budget runs out.
Given this placement exclusion audit and each ad set's ten-rule limit, prioritize which exclusions become -90% value rules first.
Audit data:
[PASTE AUDIT TABLE]
Rank by: (1) documented fraud/brand-safety reason, (2) historical spend share of that placement, (3) how recently the client raised it as a concern. Output the top 10 rules per ad set where exclusions exceed 10, and flag which lower-priority exclusions got dropped so I can tell the client directly instead of it silently disappearing.Build a weekly placement-drift check, because -90% is a discount, not a wall
This is the step that actually protects the budget. A heavily discounted bid can still clear in a slow auction window, so spend can start showing up again in a placement everyone on the team assumes is still blocked. Pull the placement breakdown report weekly per client and compare it against the pre-migration baseline. Any nonzero spend in a placement that used to be a hard zero is the signal — not a full budget blowout, a few dollars creeping in, which is exactly the stage where it's cheap to catch and expensive to ignore.
Compare this week's Meta Ads placement breakdown against the baseline from before the value-rule migration.
Baseline (placements that were fully excluded, spend should be $0):
[PASTE BASELINE]
This week's placement breakdown:
[PASTE CURRENT DATA]
Flag every placement that should be $0 but isn't, show the dollar amount and percentage of total spend, and rank by dollar risk. Note if any client is trending toward meaningful drift over the last 3 weeks, not just this one.Know when account-level Placement Controls are worth the tradeoff
Meta is keeping one true exclusion option: account-level Placement Controls, which do fully block placements — but the setting applies to every campaign under that ad account, not per client, per ad set. For an agency running one ad account per client, that's usable. For an agency running multiple clients through shared ad accounts, it's too blunt and will exclude placements for clients who never asked for it. Reserve this for accounts with a contractual or regulatory placement ban (a client with a genuine brand-safety requirement, not just a preference), and use value rules everywhere else.
Tell clients before they notice, not after
This is a platform change, not a mistake on the account, but if a client discovers a performance dip before you've explained why, it reads like the agency stopped paying attention. A short, plain note in the next report — Meta removed hard placement exclusions, here's what we migrated to instead, here's the monitor we put in place — turns a platform disruption into a visible demonstration of the agency actually watching the account.
Draft a short, plain-language note for a client report explaining that Meta removed the placement exclusion feature we used to protect their budget from low-quality placements, what we replaced it with, and that we've added weekly monitoring to catch any spend drift early. Keep it to 3-4 sentences, no jargon, reassuring but honest that this is an ongoing watch item, not a one-time fix.What changes
Every client's placement protections migrated to value rules before the exclusion checkbox disappears from their account, a documented priority list for what got kept versus dropped, a weekly check that catches budget drift back into excluded placements while it's still a few dollars instead of a quarter's worth, and a client conversation that happened on your terms instead of theirs.
Meta announced on August 20, 2026 that it's removing placement exclusion checkboxes from ad sets. No confirmed rollout date, no account-by-account notice — it's landing quietly, starting with accounts on the newer multimedia ad creation workflow. The replacement is "placement value rules," and the detail that matters is this: a value rule can discount a bid by up to 90%, but it cannot eliminate it.
That distinction sounds small. It isn't.
Why a -90% bid isn't the same as a real exclusion
For years, the placement exclusion checkbox has been one of the quiet, reliable tools of running paid media well: block Audience Network because it's historically fraud-prone and low-intent, block Messenger for a client who doesn't want conversational leads, block in-stream video for a brand that doesn't want to show up mid-content on a video nobody chose to watch. It was a wall. Spend simply did not go there.
A -90% value rule is a steep discount, not a wall. In a competitive auction, a bid that low usually loses. But auctions aren't uniformly competitive all day, every day. Late at night, in an off-season niche, in a low-demand geo, a heavily discounted bid can still clear — and when it does, spend starts flowing back into exactly the placement it was excluded to avoid. Nothing in the interface announces this. The ad set just quietly starts spending somewhere it hasn't spent in years.
The exclusions disappear before anyone documents them
The checkbox doesn't just change behavior — it removes the record. Once it's gone from an account, there's no UI history of what used to be excluded or why. If the migration to value rules doesn't happen before that point, the agency is reconstructing years of placement decisions from memory, client emails, and guesswork. That's a bad position to be in for something this operationally important.
Ten rules per ad set is not enough for everyone
Value rules cap at ten per ad set, two criteria each. Plenty of accounts had more granular placement logic than that — different exclusions for different campaign types, client-specific preferences layered on top of general best practice. The fix isn't trying to cram everything into ten slots. It's prioritizing by dollars at risk: documented fraud or brand-safety placements first, spend-share second, anything a client explicitly raised third. What doesn't make the cut should be a conscious, documented decision — and something the client hears from the agency, not something they discover on their own.
The monitor matters more than the migration
Doing the one-time migration and calling it done misses the actual risk. Because a value rule is probabilistic, not absolute, the real protection is a recurring check: pull the placement breakdown weekly, compare it against the pre-migration baseline, and flag any placement that should read zero but doesn't. Catching a few dollars of drift in week one is a non-event. Catching it in month three, after it's compounded across every account running the same rule, is a very different conversation with finance.
Bottom line
Meta didn't announce a deprecation, it announced a rename with a meaningfully weaker mechanism underneath — and it's rolling out quietly enough that most agencies will only find out when a client asks first. The agencies protected against it aren't smarter about media buying. They just did the audit and built the monitor before the checkbox disappeared instead of after.