Meta Ads Ad Set Budget Sharing: What It Is, When to Use It, and When to Turn It Off

If you’ve set up a new Meta campaign recently and excluded a placement like Audience Network or Facebook Right Column, you might have noticed something different. Meta now gives you a checkbox that says: Up to 5% of your budget is spent for each excluded placement when it’s likely to improve performance. And that checkbox is turned on by default. That means when you exclude a placement, Meta doesn’t fully exclude it anymore. Not unless you go back and manually uncheck that box. Your “excluded” placement can still receive up to 5% of your ad set budget. And that 5% applies per excluded placement, not total. If you’ve excluded 4 placements, that’s potentially 20% of your budget going to places you specifically said you didn’t want. As PPC Land documented when the feature first appeared: “Campaigns with multiple placement exclusions could see significantly more than 5% of total budget directed to placements advertisers intended to avoid.” This is part of Meta’s broader push toward algorithmic control over placement delivery. They’ve been steadily removing manual controls since 2024: detailed targeting exclusions were eliminated in January 2025, Dynamic Media was enabled by default for Advantage+ Catalog ads by October 2025, and now placement exclusions have this soft override baked in. The logic from Meta’s side makes sense. Their data shows that Advantage+ Placements (where Meta chooses everything) generally delivers lower cost per result because the algorithm has maximum flexibility to find cheap impressions wherever they exist. By sneaking 5% of spend into “excluded” placements, Meta is trying to prove that those placements can contribute to your results. The problem is that many advertisers exclude placements for good reasons: brand safety concerns with Audience Network, low-quality traffic from specific surfaces, or simply because they’ve tested those placements and they don’t convert for their offer. A default opt-in that overrides those decisions without clear notice is frustrating. Let me walk you through how to actually control your placements in 2026. How Placement Control Works at the Ad Set Level At the ad set level, you have two options: Advantage+ Placements (default). Meta decides where your ads run across all 25+ placement options. You give up control, and the algorithm finds the cheapest impressions. This is what Meta recommends for most advertisers, and honestly, for purchase-optimized campaigns with strong pixel data, it often works fine. The algorithm is good at finding cost-efficient impressions. Manual Placements. You choose exactly which platforms (Facebook, Instagram, Messenger, Audience Network) and which surfaces within them (Feed, Stories, Reels, Marketplace, Search Results, etc.) your ads appear on. This gives you full control. When you select Manual Placements and deselect specific placements, this is where the 5% spending feature kicks in. After you exclude placements, look for the checkbox that allows Meta to spend limited budget on those excluded surfaces. It may appear as a recommendation or as a checked option within the placement settings. The catch: This feature currently applies to Sales and Leads campaign objectives. If you’re running Traffic, Engagement, or Awareness campaigns, the behavior may differ. Check your specific campaign setup to confirm. For campaigns where you’re testing which placements work, leaving Advantage+ Placements on makes sense. You let Meta explore, collect data, and then review the breakdown reports to see which placements actually convert. But once you have that data and know that certain placements don’t work for you, switching to Manual Placements with genuine exclusions is a reasonable choice. Just make sure the 5% override isn’t silently undermining your exclusions. How Placement Control Works at the Account Level Meta also offers account-level placement controls. These apply to every campaign in the account, so you don’t need to remember to exclude specific placements each time you create a new campaign. To access account-level placement controls: Go to Advertising Settings in your Meta Ads Manager Select Placement Controls Toggle on “My business can only advertise on specific placements“ From here, you can exclude: Audience Network (ads on third-party apps and websites) Facebook Marketplace Facebook Right Column These account-level controls are separate from the ad set level placement selections. When you set an exclusion at the account level, it overrides any ad set level settings. Even if someone on your team creates a new campaign with Advantage+ Placements, the account-level exclusion will still apply. This is the cleanest way to permanently block a placement across your account. No checkboxes to worry about. No 5% overrides. The placement is simply off. However, if you want to exclude any specific placement like for example “Ads on Facebook Reels”, then you have to do this at the ad set placement control settings. When to use account-level exclusions: You’ve tested Audience Network extensively and it consistently delivers low-quality traffic for your business Your brand has content adjacency requirements that Audience Network can’t satisfy You have compliance or regulatory reasons that require restricting where your ads appear You want a “set it and forget it” solution that applies to all current and future campaigns Important: Account-level placement controls are only available for Auction campaigns. If you are running Reach and Frequency campaigns, then you need to manage your placement selection under your ad set placement control settings. Steps to Completely Remove a Placement If your goal is to fully block a placement with zero spend leaking through, here’s the process. Option A: Account-Level Block (Recommended for Permanent Exclusions) Open Meta Ads Manager Go to Advertising Settings (gear icon > Advertising Settings) Click Placement Controls Toggle on “My business can only advertise on specific placements“ Uncheck the placements you want to block (Audience Network, Marketplace, Right Column) Click Review Changes and then Apply Wait up to 48 hours for changes to take effect across existing campaigns This is the safest method to get rid of Audience Network, Marketplace, Right Column. It applies to everything in the account and isn’t affected by the 5% spend checkbox at the ad set level. For additional placements, consider option B. Option B: Ad Set Level Block (For Per-Campaign Control) Create or edit your campaign At the ad set level, scroll to Placements Click on Show more settings to make Placement controls visible. Click and expand Placement controls Uncheck the placements you want to exclude. Look for a checkbox saying “Allow limited spending to excluded placements“. This is the option that allows Meta to spend up to 5% on excluded placements and will appear only after excluding placements. Uncheck this box. Save and publish If you don’t uncheck the box on step 6, your “excluded” placements will still receive up to 5% of your ad set budget each. Option C: Combine Both for Maximum Protection Use account-level controls to permanently block placements you never want (like Audience Network), and use ad set level controls for per-campaign adjustments (like excluding Stories for a campaign that doesn’t have vertical creative). How TheOptimizer Handles Placement Optimization Automatically Here’s the approach I recommend for experienced buyers who want the best of both worlds: algorithmic flexibility for discovery, automated protection against waste. Instead of manually excluding placements upfront (which limits Meta’s ability to find cheap impressions), start with Advantage+ Placements or a broad manual placement selection. Let Meta explore. Then use automation rules to cut underperforming placements based on actual data. TheOptimizer connects to Meta’s API and lets you build rules that automatically block placements based on performance thresholds. Here’s what that looks like: Rule: Block Underperforming Placements IF Placement Spend > $X AND Placement CPA > Target CPA by 30% AND Placement Conversions < Y THEN Block Placement Run every 10 to 30 minutes This rule gives every placement a fair chance to prove itself. If a placement spends meaningful budget and doesn’t convert at an acceptable rate, it gets blocked automatically. No manual checking. No forgetting to review placement breakdowns. Why this is better than pre-excluding placements? You don’t miss hidden winners. Sometimes a placement you’d normally exclude turns out to work well for a specific creative or audience. Automated rules let it run until the data says otherwise. You respond to real data, not assumptions. Excluding Audience Network because “everyone says it’s bad” ignores the fact that for some offers and verticals, it converts at a very low CPC. Let the data decide. You handle the 5% problem automatically. Even if Meta’s 5% override sneaks spend into an excluded placement, your automation rules will catch it if that spend doesn’t convert. The placement gets blocked based […]

Before talking about bidding strategies, you need to understand what you’re actually bidding in. Every time there’s an opportunity to show an ad to someone, Meta runs an auction. Your ad competes against every other eligible ad targeting that same person. But the winner isn’t simply the highest bidder. Meta calculates a Total Value Score for each ad: Total Value = (Your Bid x Estimated Action Rate) + Ad Quality + User Value Three things matter. Your bid (how much you’re willing to pay). The estimated action rate (how likely this specific person is to take the action you’re optimizing for). And ad quality (how relevant and engaging your creative is based on past signals). This means a lower bid with a highly relevant ad can beat a higher bid with a poor one. Your creative quality and relevance score are multipliers on your bid. A $20 bid with strong relevance can outperform a $40 bid with weak relevance. Your bidding strategy controls one piece of this equation: how Meta decides what to bid on your behalf. Everything else (creative quality, audience relevance) is determined by your ads and your account history. The 5 Bidding Strategies Available in 2026 Meta offers five bidding strategies in 2026. Three are goal-based (they control cost), and two are spend-based (they control volume). 1. Highest Volume (formerly Lowest Cost) What it does: Meta bids whatever it takes to get you the most results within your budget. No cost control. No cap. It just spends your budget as efficiently as it can. The upside: Maximum delivery. Fastest exit from the learning phase. Zero setup. It’s Meta’s default for a reason. The downside: Your CPA fluctuates. Monday might be $25, Tuesday $55, Wednesday $30. You have no cap, so when competition spikes (holidays, weekends, industry events), your costs spike with it. Meta doesn’t care about your profit margins. It cares about spending your budget. Best for: New campaigns, new accounts, data collection phases, and situations where volume matters more than per-unit cost. 2. Highest Value What it does: Instead of maximizing the number of conversions, Meta maximizes the total conversion value. It finds people likely to make bigger purchases rather than more purchases. The upside: Higher average order value. Better for e-commerce with a wide product price range. The downside: Requires purchase value data sent through your pixel or Conversions API. Without it, Meta has nothing to optimize against. Best for: E-commerce stores where order values vary significantly ($20 t-shirt vs. $200 jacket). Useless for lead gen or flat-value conversions. 3. Cost Cap What it does: You set a target CPA, and Meta tries to keep your average cost per result at or below that target. Key word: average. Individual conversions might cost more or less, but the average should hover around your cap. The upside: Predictable costs over time. Meta still has flexibility to bid above your cap when it finds high-probability users, as long as the average stays in line. The downside: During the learning phase, costs can exceed your cap significantly before stabilizing. A lot of people burn their hands at cost caps because they expect it to behave like bid caps. Cost cap is an average, not a ceiling. Best for: Scaling campaigns where you want to maintain profitability without micromanaging bids. Ideal when you know your target CPA but want Meta to have room to find volume. 4. Bid Cap What it does: You set the maximum amount Meta can bid in any single auction. If winning an impression would require bidding above your cap, Meta doesn’t bid. Period. The upside: Hard cost control. Your CPA will never exceed your cap (on a per-auction basis). As Mathias Schrøder told Ads Uploader: “Last year was our most profitable year ever. We made a deliberate shift to prioritize profit over revenue. Bid caps were central to that strategy.” The downside: Your spend becomes the variable instead of your CPA. Some days you’ll only spend $200 of a $500 budget because Meta couldn’t find enough auctions to win at your price. Delivery can stall completely if your cap is too low. Best for: Advertisers who know their exact break-even CPA and prioritize profitability over volume. Requires data and experience. 5. Minimum ROAS (Return on Ad Spend) What it does: You set a minimum ROAS threshold (say, 2.5x), and Meta only bids on auctions where it predicts the purchase will meet or exceed that return. The upside: Directly ties bidding to revenue outcomes, not just cost. The downside: Needs high purchase volume for the algorithm to predict accurately. Not available for non-purchase optimization events. Best for: E-commerce brands with strong purchase data who want to guarantee a minimum return threshold. Cost Cap vs. Bid Cap: The Real Differences This is the comparison most people get wrong, so let me be very specific about what’s different. Cost Cap Bid Cap What it controls Average CPA across all conversions Maximum bid per individual auction Can individual conversions exceed the cap? Yes. Some will be above, some below. The average is the target. No. Meta will not bid above your cap in any single auction. What varies? Individual conversion costs fluctuate Daily spend fluctuates Daily budget usage Tends to spend your full daily budget May not spend your full budget if it can’t find enough auctions at your price Learning phase behavior May overshoot the cap initially, then stabilize May severely limit delivery initially if the cap is too tight Volume vs. profitability Leans toward volume with cost guardrails Leans toward profitability with volume trade-offs When it works best You want to scale while keeping CPA roughly predictable You want hard cost control on every conversion The fundamental mental model: With cost cap, you’re telling Meta: “Spend my budget, but try to keep the average cost around $X.” With bid cap, you’re telling Meta: “Only compete in auctions where you can win for $X or less. If you can’t find enough of those, spend less.” The trade-off is always volume vs. control. Cost cap gives you more volume with less precise control. Bid cap gives you more control with less predictable volume. When to Use Each Bidding Strategy Here’s the practical decision framework I use: Phase 1: Discovery (new campaign, new offer, new account) Use Highest Volume. You don’t know your CPA yet. You don’t know which creatives work. You need data fast. Let Meta spend freely and collect baseline numbers. Stay here for 7 to 14 days or until you have at least 50 conversions. Phase 2: Optimization (you know your numbers) Switch to Cost Cap. You now have a baseline CPA and you want to maintain it while scaling. Set your cost cap at your target CPA (not your break-even, your target). This gives Meta flexibility to find more volume while keeping your average cost in check. If your CPA starts creeping above target despite the cost cap, it usually means competition is rising or your creatives are fatiguing. See our article on detecting creative fatigue early for the specific automation rules to catch this. Phase 3: Profitability (you want to protect margins) Switch to Bid Cap. You know exactly what a customer is worth and what you can afford to pay. Set the cap at your maximum acceptable CPA with a 10 to 20% buffer for auction competition. Accept that you’ll spend less budget but at better unit economics. Phase 4: Value optimization (e-commerce with varied AOV) Layer in Minimum ROAS or Highest Value. These only make sense when you have strong purchase value data and want to optimize for revenue, not just conversion count. The transition principle: You don’t pick a strategy and stick with it forever. You graduate from one to the next as your data matures. As Stackmatix’s analysis found: “You switch strategies based on performance signals, not a calendar.” This phased approach aligns with the campaign lifecycle framework in our Meta Ads automation playbook: launch with broad settings, validate with data, then tighten controls as you scale. Can I Use Cost Cap or Bid Cap on a Brand New Account? Short answer: you can, but you probably shouldn’t. Cost cap and bid cap both rely on Meta having enough data to predict conversion probabilities. On a brand new account with zero conversion history, Meta is guessing. And guessing with a bid cap usually means one of two things: Your cap is […]

For the past two years, Meta has been stripping away manual controls. Interest targeting became a “suggestion.” Detailed targeting expanded by default. Advantage+ took over audience selection. The message was clear: trust the algorithm. Then, in June 2025, Meta did something that seemed to contradict all of that. They gave advertisers value rules. Value rules let you tell Meta’s algorithm that certain audience segments are worth more (or less) to your business, and to adjust its bids accordingly. You can bid 60% more for women aged 25 to 34. Or bid 40% less for users in a specific country. Or deprioritize a placement where your conversions have a high refund rate. All without creating separate ad sets or fragmenting your campaign structure. As Jon Loomer put it in his deep dive on the feature: value rules address a real weakness in Meta’s optimization. The algorithm is designed to get you the most results possible. But it doesn’t inherently care about the value of those results. It optimizes for volume, not quality. Value rules let you layer business intelligence on top of algorithmic efficiency. The feature launched for Sales and App Promotion campaigns in June 2025, expanded to all campaign objectives by August 2025, and received significant enhancements in 2026 including placement-specific rules and device platform adjustments. But here’s the thing Meta tells you upfront in the setup screen: “When you use value rules, you may see more conversions from your preferred audiences, but your overall cost per result may increase.” That warning isn’t decoration. Meta’s own documentation says value rules can increase your cost per result by 20 to 1,000%. Not a typo. One thousand percent. So the question isn’t whether value rules are powerful. They are. The question is whether you know enough about your business data to use them without lighting money on fire. How Value Rules Actually Work (With Examples) The logic is straightforward. Value rules are bid multipliers applied at the audience segment level within Meta’s auction system. Every time your ad is eligible to appear, Meta calculates a Total Value Score: Total Value = (Advertiser Bid x Estimated Action Rate) + Ad Quality + User Value When you set a value rule, you’re adjusting the “Advertiser Bid” component for specific audience segments. A +50% rule means Meta bids 50% higher for people matching that segment. A -40% rule means Meta bids 40% less. What you can target with value rules: Age ranges GenderLocation (countries, regions, states) Mobile operating system (iOS or Android) Device platform Ad placement (Feed, Stories, Reels, Audience Network, Marketplace) You can combine up to 2 criteria per rule (for example, “women aged 25 to 34” or “iOS users in California”). You can create up to 10 rules per rule set, and up to 6 rule sets per account. Rule priority matters. When a user qualifies for multiple rules, Meta applies only the first matching rule in the sequence. So if Rule 1 is “+20% for women in California” and Rule 2 is “+50% for iOS users,” a woman in California using an iOS device gets the +20%, not the +50%. Order your rules from most specific to least specific. Example 1: E-commerce LTV optimization Your data shows women aged 25 to 44 have an average lifetime value of $850 over 12 months. Your overall average is $530. Women in this age range are worth about 60% more to your business. Setup: Rule 1: Increase bid +60% for Women, Age 25-44 Rule 2: Decrease bid -20% for Men, Age 55-65+ Your current CPA is $45 across all demographics. With the +60% bid increase, you’re willing to pay up to $72 to acquire a woman aged 25 to 44, because her LTV justifies it. Meanwhile, you’re bidding less for a segment that your CRM data shows has a high return rate and low repeat purchase rate. Example 2: Geographic performance differences You run a B2B SaaS product. Leads from major US metro areas convert to paid customers at 3x the rate of leads from smaller markets. Setup: Rule 1: Increase bid +40% for Location: New York, San Francisco, Chicago, Los Angeles, Austin Rule 2: Decrease bid -30% for Location: [lower-converting regions] Meta’s algorithm still reaches all audiences. But it bids more aggressively for impressions in high-converting metros, steering more budget toward the leads that actually close. Example 3: Placement quality differences Your data shows that Feed conversions have 2x the average order value compared to Audience Network conversions. Setup: Rule 1: Increase bid +30% for Placement: Facebook Feed, Instagram Feed Rule 2: Decrease bid -50% for Placement: Audience Network Instead of excluding Audience Network entirely (which reduces the algorithm’s delivery options), you’re deprioritizing it through bid adjustments. Meta will still deliver there when it’s extremely cheap, but your budget concentrates on placements where the conversion quality is higher. How Value Rules Impact Campaign Pacing and Spend This is where things get practical and where most advertisers underestimate the effect. Value rules don’t change your daily budget setting. Your budget stays the same. What changes is how aggressively Meta competes in auctions for specific segments. When you increase bids for a segment, Meta enters higher-priced auctions to reach those people. That means: You win more auctions for that segment. More of your budget goes toward the people you value most. You pay more per impression for that segment. Higher bids mean higher CPMs. Your budget depletes faster if a large portion of your audience matches the high-bid rule. When you decrease bids for a segment, the opposite happens. Meta is less competitive in auctions for those people, which means fewer impressions but lower cost per impression. The net effect on pacing: If your high-value segment is a small portion of your total audience (say, 15%), the pacing impact is manageable. The budget shifts toward that 15% without dramatically changing overall delivery. But if your high-value segment is 60% of your audience and you’re bidding +50%, you’ve effectively increased your average bid across most of your delivery. Your budget will pace faster and may exhaust earlier in the day. What to watch for: Check hourly delivery after activating value rules. If your budget is exhausting by 2 PM instead of running through the full day, your bid increases are too aggressive for your budget. Monitor impression share by segment. Are you actually getting more delivery to the segments you increased bids for? If not, competition might be too high at that bid level. How Value Rules Impact Cost Per Result Let me be blunt about this. Value rules will almost always increase your average cost per result. Meta tells you this outright during setup. The question isn’t “will costs go up?” They will. The question is “will the value of the results go up by more than the cost?” Here’s the math that matters: Without value rules: 100 conversions at $45 CPA = $4,500 spend Average customer value: $530 Total customer value: $53,000 ROAS: 11.8x With value rules (+60% bid for high-LTV segment): 85 total conversions at $55 CPA = $4,675 spend 50 of those conversions are from the high-LTV segment ($850 average value) 35 conversions from other segments ($530 average value) Total customer value: (50 x $850) + (35 x $530) = $61,050 ROAS: 13.1x Fewer total conversions. Higher CPA. But higher total customer value and better ROAS. That’s the trade-off. Value rules sacrifice volume efficiency for value efficiency. If you’re optimizing for top-line conversion count, value rules will look like they’re hurting you. If you’re optimizing for revenue and LTV, they can look very different. Without Value Rules With Value Rules Conversions 100 conversions 85 conversions CPA $45 CPA $55 CPA Avg Value $530 avg value $640 avg value Total Value $53K total value $61K total value ROAS 11.8x ROAS 13.1x ROAS Fewer conversions. Higher CPA. More revenue.  Value Rules vs. Narrow Audience Targeting Before value rules existed, the standard approach for targeting high-value segments was audience fragmentation. You’d create separate ad sets for different demographics, each with its own budget. Women 25 to 34 in one ad set. Men 35 to 44 in another. Different budgets reflecting different segment values. This approach has three problems in 2026: 1. It fragments the learning phase. Each ad set needs 50 conversion events per week to exit learning phase. If you split one campaign into 4 demographic ad sets, each one needs to generate 50 events […]