Meta took away placement exclusions. How far should a value rule turn a placement down?

KEY TAKEAWAYS

1. Sort every old exclusion by its reason first. Habit gets deleted, policy moves to account controls, performance becomes a value rule.
2. The value rule is a price. Set it to the ratio of a result's worth in that placement to its worth elsewhere.
3. Breakdown results are estimated and attributed averages. Judge the ratio on CRM outcomes, with a range.
4. If you run bid cap or ROAS goal, value rules are not available. Check before you rely on them.

On 12 October 2026 Social Media Today flagged a new Meta help page that confirms what advertisers had been seeing since late August. Ad sets using the Engagement, Leads, App promotion and Sales objectives no longer let you pick placements by hand. Awareness and Traffic keep the control for now. Meta says the change is rolling out gradually and gives no end date.

Most coverage lists the replacements and stops there: use value rules, use account controls, upload more aspect ratios. The harder question is the number. A value rule asks how far to turn a placement down, and the guides that answer it offer starting points from 10% to 90% with no reason attached. This piece covers how to sort the exclusions you had, and how to set the percentage from your own data.

What changed, and what stayed?

Meta's help page lists four alternatives, as read by Relevant Audience on 13 October:

• Per-placement media. Up to ten images and videos in one ad, with aspect ratio, text, URL and delivery settings set per asset.
• Placement value rules. Raise or lower bids for specific placements.
• Account controls. Exclusions that apply to every campaign in the account, covering Audience Network placements, Facebook Marketplace and the Facebook right column. A new option sets these per objective.
• Partnership ads. You can limit placements only if the partner's post is itself limited.

Meta's case for the change is that Advantage+ placements delivered a 10% lower cost per result on average in an experiment. An earlier version of Meta's guidance, quoted by PPC Land in August, put it at 11.7% lower cost per action. Neither figure comes with a date, sample or setup. Advertisers who chose manual placements also differ from those who did not, so the comparison partly measures who chose what. Treat both as direction, not evidence.

The scope also shifted. Jon Loomer's first report on 20 August described a test limited to Sales and Leads. Meta's October page names four objectives. The later, official page is the one to plan around.

Is a value rule an exclusion?

No. Meta's API documentation caps a decrease at 90% and an increase at 1,000%. At -90% a placement still competes, it just bids a tenth as much. AdRiseLab's point is that turning a placement off has become turning it down, and that brand promises written around an off switch no longer hold.

Three limits in Meta's developer docs get less attention than the floor:

1. Bid strategy. Value rules work only with highest volume and cost per result goal. Bid cap is not planned. ROAS goal support is expected later in 2026. If your ad sets run on either, you have no placement lever inside the ad set at all.
2. Conversion setup. Sales ad sets with web and in-store conversions, and Leads ad sets with web and call conversions, are excluded.
3. Account limits. Meta's value rules page allows 6 rule sets per account, 10 rules per set and 4 criteria per rule through the API. Its bid multiplier page says 20 sets. Plan for 6 until Meta reconciles them.

Placement coverage also differs by tool. Ads Manager offers seven placements: Facebook Feed, Stories and Reels, Instagram Feed, Stories and Reels, and Audience Network. The API adds Facebook Marketplace, Search and Video and Instagram Explore, but a rule set using those becomes read-only in Ads Manager.

Why did you exclude it? Sort before you replace

Every old exclusion had one of three reasons. Each reason has a different home now.

Before you trust the account list as a blanket fix, note it applies to every campaign. For an agency running several brands in one account, one client's rule becomes everyone's.

What should the percentage be?

Here is where the guides diverge. TheOptimizer suggests starting around 10% to 20%. Dataslayer says 20% to 30%. VIXI starts at 50% and moves to 75% or 90% after two weeks. A music-ads guide from song.so goes straight to -90% and +1,000%. None explains where its number comes from.

A value rule is a price. Meta's bid strategy page says highest volume raises your bid as needed with no set limit, and the bid multiplier page says the rule's weight multiplies the bid for people who match. Our reading: at a -50% rule, Meta should only buy a result in that placement when it costs roughly half what it costs elsewhere at the margin. So the rule should equal what a result there is worth relative to a result elsewhere.

Rule weight = downstream rate in the placement / downstream rate in the rest of the ad set

Downstream rate means the share of results that become what you actually sell: qualified leads, paid orders, retained installs. A weight of 0.5 is a -50% rule. A weight above 1 is an increase. Meta's own example in its documentation, +60% for one age band, reads naturally this way: that group is worth 1.6 times the default.

This is also why Meta warns that "your overall cost per result may increase." A correctly set rule should raise cost per lead and lower cost per qualified lead. If you judge it on cost per result in Ads Manager, a good rule will look like a failure.

A worked example, with the uncertainty left in

Hypothetical numbers. A B2B lead-gen ad set runs for 30 days.

Audience Network leads are worth a third of the others. The rule weight is 10 / 30 = 0.33, a rule of about -67%. In plain terms, the rule tells Meta an Audience Network lead is worth buying at around $13 when a feed lead costs $40.

Now the range. We calculated 95% Wilson intervals for each rate. Thirty of 300 gives 7.1% to 13.9%. 180 of 600 gives 26.5% to 33.8%. The honest ratio runs from about 0.21 to 0.53, so the right rule sits anywhere from about -47% to -79%. A -50% rule is the cautious end of what the data supports. -90% is outside it.

With smaller numbers the method stops working. If the placement had 40 leads and 4 qualified, the interval for that rate is 4.0% to 23.1%. Against 30%, the ratio could be anywhere from 0.13 to 0.77. That is not a finding. Leave the placement alone and collect more.

Two cautions on the inputs. Meta's breakdowns documentation states that breakdown values are estimated, and results in a breakdown are Meta-attributed, not caused. Take qualified counts from your CRM, joined by lead ID or a placement parameter in the URL, not from Ads Manager. If you are unsure how Meta's attribution settings shape these numbers, our piece on incremental attribution covers what each window counts.

How should you test the rule?

1. Export your current exclusions. For each one, write the reason. Use the table above to sort them.
2. Move policy exclusions first. Set account controls for placements on Meta's list. Turn off the 5% limited spend option on any ad set that still has manual exclusions; Meta's developer blog confirms it allows up to 5% of spend per excluded placement.
3. Check eligibility. Confirm each ad set uses highest volume or cost per result goal and a supported conversion setup.
4. Pull 60 to 90 days of downstream data by placement. Compute the ratio and its range. If the range spans no change, do not set a rule.
5. Set the rule at the cautious end of the range. Put protective rules first, since only the first matching rule applies.
6. Run it against a copy without the rule. Burtch and colleagues studied 181,890 Meta A/B tests and found audiences differed between cells far more often than chance, with automatic placement among the causes. That makes an ad set A/B a fair test of the whole setup, which is the question here, but not of placement quality inside it. Judge on cost per qualified result for the ad set, not per placement.
7. Review after enough downstream outcomes, not after a week. Sales cycles set the clock. Our sizing guide for creative tests uses the same conversions-first logic.

What about the placements value rules cannot reach?

Instagram Explore, Facebook search and in-stream video have no hard control and only an API-level rule. Per-placement media helps with fit, but do not treat a missing asset as an exclusion. Blip's guide notes Meta fills missing ratios with padding, expansion or AI crops. TikTok raised a similar placement question when it opened its off-platform network, covered in our TikTok Ad Network piece. In both cases the control has moved from a switch to a price.

The short version

Delete exclusions you cannot justify. Move policy ones to account controls. For the rest, set the rule to the value ratio from your CRM, at the cautious end of its range, and judge it on cost per qualified result. If you also want to know how your brand shows up when buyers ask AI assistants instead of scrolling a feed, the waitlist is open.

Questions

Which Meta campaigns lose manual placements?

Meta's help page, flagged on 12 October 2026, says Engagement, Leads, App promotion and Sales ad sets no longer choose placements manually. Awareness and Traffic keep the control for now. The rollout is gradual and Meta has given no end date.

Can a value rule switch a placement off?

No. Decreases stop at 90%, so a placement can be made expensive to win but not removed. A hard off switch exists only in account controls, which cover Audience Network, Facebook Marketplace and the Facebook right column across every campaign in the account.

Do value rules work with every bid strategy?

No. Meta's API documentation lists highest volume and cost per result goal only. Bid cap is not supported and is not planned, and ROAS goal support is expected later in 2026. Sales ad sets with web and in-store conversions and Leads ad sets with web and call conversions are also excluded.

How much data do I need before setting a placement rule?

Enough downstream outcomes that the ratio is not a guess. In our worked example, 30 qualified leads from one placement still left the right rule anywhere between about -47% and -79%. With four qualified leads the range is too wide to act on.

Sources

1. Social Media Today, Meta offers updated ad placement control guidance socialmediatoday.com
2. Relevant Audience, Meta documents what replaces ad set placement controls relevantaudience.com
3. Meta for Developers, Value Rules developers.facebook.com
4. Meta for Developers, Bid Multipliers developers.facebook.com
5. Meta for Developers, Bid Strategy developers.facebook.com
6. Meta for Developers, Introducing limited spend on excluded placements in Marketing API developers.facebook.com
7. Meta for Developers, Insights API Breakdowns developers.facebook.com
8. Meta for Developers, Placement targeting developers.facebook.com
9. Burtch et al., Characterizing and Minimizing Divergent Delivery in Meta Advertising Experiments arxiv.org
10. Jon Loomer, Meta is removing placement controls from ad sets jonloomer.com
11. Jon Loomer, Value rules jonloomer.com
12. Jon Loomer, Meta adds value rules for placements jonloomer.com
13. PPC Land, Some Meta advertisers lose placement controls, with bid cuts capped at 90% ppc.land
14. PPC Land, Meta's value rules might actually cost more than they're worth ppc.land
15. AdRiseLab, Meta removed placement controls from ad sets. Value rules are not a replacement adriselab.com
16. VIXI, Meta placement exclusions removed. What to do now vixi.agency
17. TheOptimizer, Meta ads value rules. How they work, when they help theoptimizer.io
18. Dataslayer, Meta value rules. What they are, how they work and why they increase costs dataslayer.ai
19. song.so, Meta placement update 2026. Value rules step by step app.song.so
20. Blip, Meta ads placement customization guide 2026 withblip.com

NEXT STEP

See where your brand stands.

Your Digital Authority Score, who AI recommends for your target questions, and the placements that close each gap. Your first report is free.

Get your free report →