Quick answer: how do you decide what to cut from Meta placement, age and region breakdowns?
Treat each breakdown as an average, not a verdict. First convert segment ROAS into contribution margin after tax, COD/RTO and returns. Then send each losing segment to the one lever that still works in late 2026: an account-level placement block, an audience control, a value rule, a creative fix, or no action at all.
Your cutting tools changed in 2026: here is what is left
You can no longer untick placements inside the ad set for Sales, Leads, Engagement or App promotion campaigns. Awareness and Traffic campaigns keep ad-set placement selection.
Around 20 August 2026, advertisers reported an in-product notice saying that placement, platform, device and OS exclusions were leaving ad sets. By 12 October 2026, Meta had published an undated help page, 'What's new in placement settings in Meta Ads Manager'. It confirmed the change and listed the alternatives, as Social Media Today reported. Meta says the change is rolling out gradually, so it may not be in every account yet.
Audience settings changed earlier. Advantage+ audience now applies automatically to new campaigns where it is available. Under it, age, gender, lookalikes and detailed targeting are only suggestions. Minimum age, location, language and custom audience exclusions are controls Meta will not go beyond.
That leaves 5 possible responses to any segment that looks weak:
| Lever | What it does | Where it lives | Strength |
|---|---|---|---|
| Account-level placement exclusion | Blocks Audience Network, Marketplace or Right Column, now also by objective | Advertising Settings | Hard off |
| Audience control | Sets minimum age, location, language and exclusions | Ad set | Hard limit |
| Value rule | Raises or lowers bids by age, gender, location, OS or placement | Advertising Settings | Dimmer, max -90% |
| Per-placement creative | Fits the ad to Reels, Stories or Feed | Ad | Fixes the cause |
| Leave alone | Lets Meta keep pricing the segment at the margin | - | Default |
Read breakdowns as averages, not verdicts
A low segment ROAS (return on ad spend) is an average, not proof that the next rupee there is wasted. Meta already pushes spend toward the segments that convert cheaply. That keeps their average high while each extra impression in them gets more expensive.
This is the breakdown effect. If you cut the 'worst' segment, spend moves into pricier auctions in the 'best' one, and your overall cost per purchase can rise.
Which breakdowns to pull
Use the Breakdown menu for Age, Gender, Age and Gender, Country, Region, DMA (Nielsen's US marketing areas), Placement, Platform and Impression Device. You can combine breakdowns from different categories, such as Age with Day. You can't combine two from the same category.
In Sales campaigns, also pull Audience Segments. It splits results into new, existing and engaged customers. A region with a high ROAS is often just the place where your repeat buyers live.
Our rule of thumb before acting
- Use at least 14 days of data, ideally 28, and the same attribution setting every time.
- Require at least 50 purchases (or 30 qualified leads) in the segment. Spend in the segment should be at least 3 times your break-even CPA (cost per acquisition). Below that, the gap is mostly noise.
- Don't compare festive windows such as Diwali, Ramadan or BFCM (Black Friday/Cyber Monday) with normal weeks. CPMs (cost per 1,000 impressions) rise unevenly by region and age during these periods.
- Skip ad sets that are still learning. Their breakdowns are unstable, so get them out of the learning phase first.
Turn segment ROAS into margin before you cut
Judge every segment on contribution margin after tax, product cost, shipping, COD/RTO and returns. Reported ROAS treats a refused cash-on-delivery parcel in one state the same as a prepaid delivery in another. If you haven't set your threshold yet, start with your break-even ROAS.
Tax sits on both sides of the sum
On the revenue side, the purchase value your Pixel or Conversions API sends may include GST in India, VAT in the UK, EU and Gulf, or sales tax in the US. Whether it does depends on your store settings. Strip tax out before comparing segments.
On the cost side, the rules differ by market:
- India: advertising carries 18% GST. If you are billed from abroad, you self-assess it under reverse charge and claim it back as input tax credit where eligible.
- UAE: UAE tax advisers report that Meta adds 5% VAT unless you submit a valid TRN. We could not confirm this from Meta.
- UK: VAT-registered businesses self-account at 20%.
- Australia and Canada: Australia charges 10% GST and some Canadian provinces charge 13% HST.
- Location fees: some industry reports say Meta adds location-based fees of up to 3% in some countries. We could not confirm the rate from Meta, so check your invoices.
If you can't reclaim the tax, your real spend is higher than Ads Manager shows. That matters when you compare countries.
COD and RTO reshape region results
In India, the Gulf and parts of South-East Asia, many orders are COD (cash on delivery). Some are refused or undelivered and come back as RTO (return to origin). You pay two-way shipping and get no revenue, yet Meta still counts the order as a purchase. Pull RTO by region from Shopify, WooCommerce or your courier. Also check where UPI payments often fail at checkout, because failed prepaid attempts push more buyers to COD.
Worked example: two regions with opposite stories
Example: a D2C brand spends ₹30 lakh a month on Meta. Over 28 days, Region A and Region B each received ₹5,00,000. The average order value is ₹1,180 including 18% GST. Product cost is ₹350, shipping is ₹100 per delivered order, and each RTO costs ₹150 in two-way shipping.
Step 1: Read what Ads Manager shows
Region A: 1,200 purchases, ₹14,16,000 revenue, ROAS 2.83. Region B: 1,600 purchases, ₹18,88,000 revenue, ROAS 3.78. On these numbers, Region B looks about 34% better.
Step 2: Rebuild contribution with RTO
Net revenue per order after GST is ₹1,000, so each delivered order leaves ₹550. Region A's RTO rate is 5%. Region B is COD-heavy, with a 30% RTO rate.
Contribution = delivered orders × ₹550 − RTO orders × ₹150 − ad spend
Region A: 1,140 × ₹550 − 60 × ₹150 − ₹5,00,000 = ₹1,18,000
Region B: 1,120 × ₹550 − 480 × ₹150 − ₹5,00,000 = ₹44,000
Step 3: Pick the lever
Region B earns about 37% of Region A's contribution, despite its higher ROAS. It is still profitable, so excluding it would throw away ₹44,000. Instead, apply a small value-rule decrease to Region B and work on the cause, for example with COD confirmation calls or a prepaid incentive.
This is why we judge ads on profit rather than platform ROAS. For more on the gap, see why a 3x ROAS campaign can still lose money.
Which lever for which finding: a fixed order of checks
Run each weak segment through these checks in order, and stop at the first one that applies. This keeps hard blocks for real constraints and stops you from reacting to averages.
- Is it a legal, policy or serviceability problem? Use a hard control: a location exclusion, minimum age, or an account-level placement exclusion.
- Is there enough data? If it is below the thresholds in our rule of thumb, leave it alone.
- Is it above break-even on margin? Leave it alone, even if it is below the account average.
- Is it a weak placement with creative that doesn't fit? Fix the creative first.
- Is it below break-even with enough data? Apply a value-rule decrease and judge it over 2-4 weeks.
- Is it still below break-even after a strong decrease and 4 weeks? Move to a hard control, or accept the segment as a cost of reach.
Placements
To switch a placement off completely, go to Advertising Settings, open Placement Controls and turn on 'My business can only advertise on specific placements'. Then untick the placements you don't want.
This covers Audience Network (native, banner, interstitial and rewarded video), Facebook Marketplace and the Facebook right column. The setting applies to every campaign in the account, and you can now set exclusions by objective. Meta intends it for legal, policy or business requirements, not for fine-tuning ROAS.
Awareness and Traffic ad sets still allow exclusions. In those, untick the 'Allow limited spending to excluded placements' box if you want a full exclusion. When ticked, Meta's in-product text reportedly says it aims to spend around 5% of budget on each excluded placement. Account-level controls have no such carve-out.
Weak Reels or Stories results usually mean a 4:5 feed ad squeezed into 9:16 space. Use per-placement creative customisation before you dim the placement.
Age and gender
In most Sales campaigns, leave age and gender alone. Intervene only for legal or product reasons. You can raise the minimum age to as high as 25 without leaving Advantage+ audience.
A minimum above 25, or a maximum age, needs original audiences. Original audiences give up the flexibility Meta credits with 7.2-14.8% lower cost per result in its own A/B tests.
Lead generation and service brands face a different problem. Meta may push budget toward people aged 65+ because they produce cheap leads, which can be low quality. Dim that age band with a value rule rather than removing it, if value rules are available for your objective. Then send qualified-lead signals from your CRM back to Meta through the Conversions API (CAPI), so it learns what a good lead looks like.
Regions
If you can't ship to or serve an area, exclude it as a location. Then make sure 'Reach more people likely to respond to your ads' is off, because with it on, Meta can deliver outside your chosen locations.
For areas you can serve but where margin is poor, use a value rule instead. In India, review RTO by state or pincode. In the Gulf, split results by country, because tax, delivery and COD habits differ. In the US, use the DMA view.
Value rules dim a segment, they never switch it off
Value rules adjust bids by age, gender, location, mobile OS and placement, but a decrease is capped at 90%. A dimmed segment can still win auctions and spend. Increases can go up to 1,000%.
- Value rules are set up under Advertising settings. PPC Land reports they work with Sales and App promotion campaigns using website or app conversions, but not with ROAS goals, bid caps or maximise value of conversions. We could not confirm this from Meta, so check your account.
- Each rule set holds up to 10 rules, with up to 2 criteria each.
- According to PPC Land, only the highest-priority matching rule adjusts the bid. Put specific rules (for example, Region B plus Audience Network) above general ones (Region B alone).
- Meta warns that cost per result may rise. You are trading cheap volume for valuable volume, so judge the result on margin, not CPA.
- Allow 2-4 weeks before you judge a change.
Start with decreases of 20-30% and make one change at a time. Log the date of every change. Compare against a holdout campaign or region, not a simple before-and-after view, because seasonality and creative changes muddy the comparison.
What most guides get wrong about cutting segments
Much of the advice that ranks today was written before the August-October 2026 changes, or it treats attributed ROAS as truth. Here are the common errors:
- 'Switch to manual placements and untick Audience Network in the ad set.' This no longer works for Sales, Leads, Engagement or App promotion. Use account-level exclusions instead.
- 'Right Column and Marketplace can't be excluded any more.' Wrong. Both, plus Audience Network, can still be blocked at account level.
- 'Meta hasn't published a help page.' Outdated. The help page went live in October 2026.
- 'Use a value rule to switch off an age band.' Value rules can't do this, because the -90% floor still lets the segment spend.
- 'Test 25-34 against 35-44 in separate ad sets.' Under Advantage+ audience, age ranges are suggestions, so the test isn't clean.
- 'Build campaigns for winning segments and pause the losers.' This ignores the breakdown effect and splits your learning across more campaigns.
Before trusting any instruction, check the date of the guide and the objective of your campaign. The same click path can work for a Traffic campaign and fail for a Sales campaign.
A 30-minute monthly breakdown audit
Once a month, work through the same 6 steps so that decisions come from a process, not from a bad day.
- Pull placement, age, gender, region and Audience Segments breakdowns for the last 28 days, using a fixed attribution setting.
- Remove tax from revenue, then add RTO, returns and shipping by region from your store data.
- Mark segments that meet the data thresholds and sit below break-even on margin.
- Classify each one as block, control, dim, creative fix or leave alone, using the order of checks above.
- Make at most 2 changes and set a holdout.
- Log each change with its date, and review it after 2-4 weeks.
For a shorter weekly pass, the 7-day breakdown method covers the same splits on recent data. ScaleWin's assistant runs this pass automatically. It pulls 7-10 days by age, gender, placement, device, region and hour, and compares landing-page views with link clicks before suggesting a change.
For how breakdowns fit with bidding, creative and tracking, see the complete 2026 guide to Meta ads optimization.
- Profit Margin Calculator - Free profit margin calculator: gross, operating and net margin plus markup in one form, with margin after Meta and Google ads and target-mar
Read next in this series
Key takeaways
- Sales, Leads, Engagement and App promotion campaigns no longer allow ad-set placement selection. Hard blocks now sit at account level, and you can set them by objective.
- A low-ROAS segment is an average. Meta has already priced it, so the next rupee there may earn more than you think.
- Rebuild region and age results with GST or VAT, COD/RTO and shipping before you cut anything. The ranking often flips.
- Value rules can lower a bid by up to 90%, but they never switch a segment off. Use hard controls only for legal, policy or serviceability reasons.
- Act only on 14+ days of data with at least 50 purchases in the segment, and judge changes over 2-4 weeks against a holdout.
Meta ads optimization: frequently asked questions
Can I still exclude Audience Network on Meta in 2026?
Why did placement options disappear from my Sales campaign ad set?
Do value rules completely stop ads in a placement or age group?
Should I exclude age groups with low ROAS in Meta ads?
How do I stop Meta from showing ads outside the regions I deliver to?
Does GST or VAT on Meta ads change how I should calculate ROAS by country?
Sources
- Meta Business Help Centre: Breakdowns in Ads Manager
- Meta Advantage+ audience
- Social Media Today: Meta offers updated ad placement control guidance
- Social Samosa: Meta removes manual placement selection for some objectives
- PPC Land: Meta's value rules might cost more than they're worth
Let ScaleWin read your ad data every day
ScaleWin checks your Meta and Google Ads daily, judges every ad on profit and tells you exactly what to change.
Start free, no card



