Meta Cuts Credit Cards for Big Advertisers, Cites Fraud Fight

Meta projected that roughly 10 percent of its 2024 ad revenue, about $16 billion, would come from ads for fraudulent or banned goods, according to internal documents Reuters reviewed last November. That number is the lens through which the April 1 payment change should be read: a platform with 3.58 billion daily users across Facebook, Instagram, and WhatsApp is quietly pulling credit cards out of the hands of an undisclosed slice of its advertiser base, and the fraud math is the most plausible reason why.

The Numbers

The headline figure is the one Meta will not confirm: how many advertisers actually lose card access on April 1. The company's own framing, as Payments Dive reported, is "a very small percentage of advertisers." Smaller advertisers are explicitly exempt. No spending threshold has been disclosed. That leaves a bounded but wide range: somewhere between the long tail (untouched) and the global brand tier (presumably already on invoicing). The affected cohort is the mid-to-upper merchant segment, and Meta is refusing to size it.

Anchor that against the platform's scale. 3.58 billion average daily users in December 2025, per Meta's January quarterly filing, is the inventory pool every one of these advertisers is bidding into. The platform does not have a demand problem. It has, by its own internal projection, a fraud problem worth roughly $16 billion in annualized revenue exposure, which is the figure the November Reuters report attached to 10 percent of 2024 ad revenue based on four years of internal documents.

The mechanics of the new system matter for any finance team modeling cash flow against ad spend. Advertisers moved off cards get assigned a credit line covering both ad accounts and WhatsApp business accounts, with payment due 30 calendar days from invoice date. Meta is explicit in its terms that this credit line "is not credit or financing." If the limit is hit, ads and WhatsApp business activity pause until payment clears or the limit is raised. The alternative is direct debit, either an ACH-style pull from a linked bank account or a manual transfer to a Meta bank account.

Compare that to the previous regime, where card payments were triggered automatically by spending thresholds. Under cards, the float belonged to the advertiser and the rewards belonged to the advertiser. Under invoicing, the working capital terms are arguably better (30 days net), but the rewards are gone, and the credit-line ceiling is now a hard throttle on campaign velocity. Trade group Merchants Payments Coalition pegs typical U.S. interchange at 1 to 4 percent, averaging around 2 percent. On a seven-figure monthly Meta spend, that 2 percent in lost cashback or points is real money the marketing team will notice in Q2.

What's Actually New

Three things are genuinely new here, and one thing is not.

What is new: the verification step. Chris Pollard, founder of media-buying platform Ads Uploader and the person who first surfaced the policy change in a March 3 blog post, flagged that invoicing forces advertisers to verify bank or other identifying information with Meta. That is a structurally different KYC posture than handing over a Visa or Mastercard number. A stolen card can fund a scam-ad campaign in minutes; a verified bank relationship and a 30-day invoice cycle creates a much longer window for Meta's trust and safety systems to detect anomalies before the platform is on the hook for a chargeback.

Also new: the credit-line ceiling as a circuit breaker. Card-based spending throttles were essentially statement-cycle limits enforced by the issuer. The new ceiling is enforced by Meta itself, against both Ads Manager and WhatsApp business accounts simultaneously. For agencies running cross-property campaigns, that is one consolidated kill switch instead of several issuer-level ones.

And new: the explicit segmentation. Meta is openly running a two-tier billing system now. Smaller advertisers keep cards. A subset of larger ones do not. The threshold remains unpublished, which is going to make it harder for any single agency to predict whether a client lands above or below the line until the email arrives.

What is not new: invoicing itself. Google Ads, the trade desks, and most enterprise SaaS billing have offered net-30 invoicing to large accounts for years. The Meta Marketing API already supported invoiced billing for enterprise accounts. What changed is the direction of pressure: Meta is now pushing advertisers onto invoicing rather than treating it as an option requested by the buyer. The source does not disclose how many advertisers are affected, which matters because the difference between "5,000 accounts" and "500,000 accounts" is the difference between a routine policy tweak and a structural rebuild of how performance marketing buys Meta inventory. If the affected cohort exceeds 50,000 accounts, expect at least one major agency holding company to publicly comment within 60 days. If it stays quiet, the cohort is small.

What's Priced In for Performance Marketing

Performance marketers have been operating under the assumption that card rewards on ad spend are a permanent line item. That assumption is now broken on Meta and will probably break elsewhere within 18 months. AdExchanger called the rewards loss "collateral damage" from the fraud fight, which is the right frame: nobody at Meta woke up wanting to take points away from media buyers, but if the choice is $16 billion of fraud exposure versus the cashback yield on legitimate advertisers, the trade is obvious.

What is already priced in: the operational shift to invoiced billing. Any agency above the mid-market tier already has accounts payable workflows for net-30 vendor invoices. Plugging Meta into that pipeline is finance-team plumbing, not a re-architecture.

What is not priced in: the working-capital impact on small-to-mid agencies that float client spend on corporate cards and bill clients on net-30 or net-60. Those shops were earning 1 to 4 percent on the spread plus float. Removing the card rail compresses that margin to zero on the affected accounts. Some of those agencies will respond by repricing client retainers. Others will not realize the hit until Q3 reporting.

Also unpriced: the WhatsApp business-account coupling. Bundling Ads Manager and WhatsApp under one credit ceiling means a brand running a customer-service-heavy WhatsApp deployment can have its paid-ads paused by overspend on the messaging side, or vice versa. That is a new operational dependency that conversational-commerce teams have not modeled.

Contrarian View

The consensus read is that this is a fraud-prevention move dressed up as billing housekeeping. I'd argue the more interesting interpretation is treasury management. Meta on net-30 invoicing collects 30 days of float on a meaningful chunk of its ad revenue. At the company's scale, the present-value impact of shifting any non-trivial slice of advertisers from card-immediate to invoice-net-30 is not a rounding error. It also reduces interchange leakage to issuing banks, which is the same cost line that has driven the 21-year merchant lawsuit against Visa, Mastercard, and the major banks now headed for a settlement hearing in New York next month.

Put differently: fraud is the public rationale, and it is real. But the financial incentive to push large advertisers off cards exists independently of the fraud math, and the decision likely scores well on both dimensions inside Meta's finance org. We do not know the bank-fee savings figure, but the lower bound is roughly 1 percent of affected gross spend (the floor of the interchange range) and the upper bound is around 4 percent. If the affected cohort represents even 5 percent of Meta ad revenue, the annualized interchange savings clear nine figures.

Key Takeaways

• Meta is moving an undisclosed slice of advertisers off credit cards onto net-30 invoicing or direct debit starting April 1, with smaller advertisers exempt and no published spending threshold.
• The fraud context is the $16 billion figure: roughly 10 percent of 2024 ad revenue was projected as coming from fraudulent or banned-goods ads, per Reuters' November reporting on internal documents.
• Affected advertisers lose 1 to 4 percent in card rewards (around 2 percent on average per Merchants Payments Coalition data) and gain a credit-line ceiling that can pause both Ads Manager and WhatsApp business accounts.
• The unanswered question is cohort size. If more than 50,000 accounts are affected, expect public agency holding-company commentary within 60 days; silence implies a smaller cohort.
• Watch for Google and the major DSPs to follow with similar invoicing-only policies for upper-mid-market advertisers within 18 months if Meta's fraud metrics improve measurably post-April.