Meta Advantage+ budgets: the levers it ate, the margin math
Meta Advantage+ budgets run on automation by default. What each handover costs, the quiet 20% placement leak, and the two levers that still cut CAC.
Partner links appear below. Buying through one pays The Margin a commission and costs you nothing extra. How we make money.
Up to a fifth of your Meta ad budget can now land in places you explicitly blocked, and the setting that allows it ships switched on. That is the line item in February 2026's change: Meta merged manual setup and its "Advantage+" automation into one flow, the automation is now the default for every new Sales, Leads and App Promotion campaign, and the machine comes pre-approved to pick your audience, choose where ads appear, split the budget, and tweak your ads. Read the handover as a profit question, not a workflow one. Your ads and clean sales tracking still move your cost per result. Everything else is Meta's now, drift included.
What happened
In February 2026 Meta collapsed the old choice between manual setup and Advantage+ into a single campaign flow. You no longer pick a lane. You get one setup with the AI options already switched on: automated audience selection, automated ad placement (where your ads appear across Facebook, Instagram and Meta's partner apps), automatic touch-ups to your ads, and a shared budget the system splits as it sees fit. Meta confirms it on its own Advantage+ campaign help page. The manual controls had been thinning for a while. The option to exclude specific interest groups went away in early 2025, and in March 2026 Meta added a control to cap spend on existing customers, an admission that the automation was over-spending on people who already buy from you.
The money angle
For a big company with a media team, this is a workflow change. For a small business owner running their own ads, it changes where the budget leaks and where the return comes from.
of budget onto ad locations you excluded
if you exclude four locations
genuinely different angles
Targeting becomes a suggestion, not a setting. Under the automated audience system, only location and minimum age are hard limits. Your interest picks, your customer lists, even your "people similar to my buyers" audiences are treated as hints that Meta spends outside of whenever its model thinks it will do better (Jon Loomer walks through how soft those inputs are). If your product suits only a narrow buyer, that looseness raises what you pay for a lead worth having.
Ad locations get handed over too, and budget can drift onto cheap space. When you exclude a location on a Sales or Leads campaign, Meta shows a checkbox, frequently ticked by default, that lets it spend up to 5% of budget on each excluded location anyway. PPC Land documented the behavior. Exclude four and a fifth of your budget can land exactly where you tried to block it.
That leaves the ads themselves as the lever that moves your numbers. When the system owns who sees the ad, where, and for how much, you control what the ad says and what you are offering. One tired ad gives the machine nothing to choose between; five honestly different angles give it room to find you a cheaper sale.
Where it breaks
The other break is treating Advantage+ as a targeting tool and starving it of ads: give the machine one ad and you hand it your whole budget with no way to find a better outcome.
What to change this week
Feed every campaign four to six genuinely different ads, not four crops of the same image, because the system can only test what you give it. Set the exclusions that still hold: existing customers, recent purchasers, anyone you never want to pay to reach again. Block ad locations you do not trust at the account level (Advertising Settings, then Placement Controls), which overrides everything and has no 5% leak checkbox. Check the placement breakdown report weekly for drift. And keep your sales tracking clean, because if the code that reports purchases back to Meta is broken, the system optimizes toward noise and you pay for results that are not real.
Your move
If leads are the goal, the weak point is almost never Meta. It is what happens after the form: inquiries sit, nobody replies for hours, and the spend is wasted no matter how good the targeting was. A connected customer-follow-up system that fires an automatic text and email the second a lead lands, the kind GoHighLevel runs, recovers more wasted budget than any audience tweak.
What the work needs
Metathe AI owns targeting, you own the ads and exclusions
GoHighLevelinstant text and email the second a lead lands| Lever | What it costs you now | The counter-move |
|---|---|---|
| Targeting | Your audience picks are hints, so spend drifts past your ideal buyer | Hard exclusions for people you never pay to reach again |
| Ad locations | Up to 5% of budget per excluded location, ~20% across four | Block at account level, where no leak checkbox exists |
| Budget split | Leans toward cheap space and buyers you already own | Weekly breakdown check, cap existing-customer spend |
| The ads | One tired ad gives the system nothing to test | 4-6 genuinely different ads, replaced when they wear out |
| Sales tracking | Broken tracking buys results that are not real | Verify monthly that reported sales match your books |
What to watch
Watch whether the manual escape hatches survive. Each release has quietly removed another control, so if exclusions soften further, your ads and your offer become the only real levers left. Build that muscle now while it still pays. The same machine-owns-targeting shift is playing out in Google Ads and behind the rising ad prices squeezing paid budgets. More under our paid ads coverage and the full article archive.
Frequently asked questions
How much of a Meta budget can a default Advantage+ setup waste?
The one leak you can put a number on: Meta may spend up to 5% of a campaign's budget on each ad location you excluded, via a checkbox that often ships ticked. Exclude four locations and a fifth of the budget can drift onto space you blocked. Add spend on existing customers, which Meta only recently let you cap, and a hands-off campaign can quietly tax itself before a single new buyer shows up.
Does handing Meta the controls raise or lower my cost per customer?
It depends on what you feed it. Advantage+ tends to hold or improve cost per result when it gets four to six genuinely different ads and clean sales-tracking data, because the system has options to test. Feed it one tired ad or broken tracking and it optimizes toward noise, and your cost per new customer climbs without any warning in the dashboard.
Where should the budget I used to spend on targeting go now?
Into making more ads. Audience tinkering no longer moves the number because your audience choices are treated as suggestions. The inputs that still change cost per result are the ads themselves and the offer, so the money that used to fund targeting tests earns more as a steady pipeline of varied ads.
What is the fastest payback fix on a default Advantage+ campaign?
Block unwanted ad locations at the account level (Advertising Settings, then Placement Controls) instead of inside each campaign. It takes minutes, overrides the 5% leak checkbox entirely, and stops paying for space you already decided you do not want. Capping spend on existing customers is the second-fastest recovery.
About The Margin
The brief on marketing that actually pays. Funnels, CRM, paid ads and lead gen for operators who care about ROI. How we work
The Margin · Daily Brief
Know what changed before it costs you
Each weekday: one shift in marketing, what it does to your numbers, and the move that protects them.
Short, priced, weekday mornings. Out in one click if it stops paying.
Keep reading
Related briefings
Google Ads budget controls 2026: stop the quiet overspend
Google Ads budget controls are back: negative keywords, placement exclusions and a search terms report put a price on Performance Max waste.
The move. Three new controls let you stop Performance Max from spending on junk and from buying back your own brand. Turn them on and you recover budget that was leaking straight out the bottom of the funnel.
Meta CPMs are rising into H2. Protect your payback
Meta CPMs climbed double digits in 2025 and Q4 runs hottest. The levers that hold ROAS and CAC when impressions get more expensive, before you cut spend.
The move. Pricier Meta ads raise your cost per new customer even when your ads didn't change. The fix is not less spend. It is more ad variations, a stronger offer, and better customer retention so each customer is worth paying more to win.
AI recommendations are rerouting buying. The margin view
AI recommendations convert about a third better than other traffic, per Adobe. Feeds, reviews and schema get you named without spending ad budget.
The move. Buyers sent your way by an AI assistant turn into customers about a third more often than other visitors, and the way in is free: Google's product feed, Perplexity's merchant program, machine-readable page labels, and fresh reviews. If you are not in the data, you are not in the answer, and you pay for that buyer somewhere else.