Meta ad accounts: limits, billing thresholds and account maturity

Fewer than you think. Every ad account fragments your signal: conversion history, audience learning and delivery reputation accumulate per account, so five accounts each spending a thousand a month are weaker than one account spending five thousand. Splitting also multiplies the administrative surface - five payment methods, five spending limits, five things to be flagged.

Split only when there is a structural reason. Separate legal entities with separate invoices. Separate currencies, which cannot be changed after the first spend. Genuinely separate time zones, which cannot be changed once billing has run. Different clients inside an agency. Those four are real. "Different campaigns" and "different brands under one company" are not - use campaigns and naming conventions for that.

Note also that the number of ad accounts a portfolio may open is capped and scales with verification and spend history. New portfolios start with very few. This is another argument for not treating accounts as disposable: you cannot simply open a sixth one because the fifth got messy.

Most advertisers assume Meta charges monthly. It does not. Automatic payments work on a billing threshold: you accrue costs and are charged whenever you reach a threshold amount, and again on your monthly billing date for anything left over. A new account typically starts with a low threshold - the equivalent of a small double-digit amount - and it rises in steps as you accumulate successful payments.

The practical consequences are worth knowing before finance asks. Early on you will see many small charges rather than one clean invoice, which alarms accountants. Any failed payment - an expired card, a bank fraud block, a monthly card limit - pauses delivery immediately and can push the threshold back down. And because thresholds rise with reliability, an account that pays cleanly for months can spend materially more per cycle before hitting a billing stop.

• Use a business card with a limit well above your monthly spend, and a backup payment method on the account.
• Warn the bank before a scaling month. A card blocked for suspected fraud on a large charge is one of the most common causes of a mysterious mid-flight pause.
• Match the ad account time zone and currency to how the business reports, at creation. Neither can be changed afterwards without a new account.

Three different ceilings can stop delivery, and confusing them wastes hours. The account spending limit is one you set: a hard cap on lifetime spend for the account until you raise or reset it. It is excellent protection against a compromised account or a fat-fingered budget - and an excellent way to halt every campaign at 2am if you forget it is there.

The campaign or ad set budget is the ordinary daily or lifetime budget. Worth remembering that daily budgets are averages, not caps: Meta may spend meaningfully more on a high-opportunity day and correct across the week.

The platform-imposed limit is the one nobody chooses. New, unverified or previously flagged accounts carry a maximum daily spend imposed by Meta. It rises with verification, payment history and clean policy record. If your budget increase simply does not spend, and there is no policy notice, this is usually why.

"Account maturity" and "account warm-up" get talked about as folklore. They describe something real, but the mechanism is unglamorous: the system performs better when it has more reliable evidence about you, and it extends more latitude to entities with a clean record. Maturity is the accumulation of both.

• Conversion history. Volume of optimised events, at a consistent definition, over a continuous period. This is the single largest factor. Fifty purchases a week teaches the system far more than five.
• Signal quality. A properly deduplicated pixel and conversions API, high event match quality, verified domain, correctly prioritised events. Poor signal makes an old account behave like a young one.
• Payment reliability. Successful charges, no disputes, no repeated declines.
• Policy record. Rejections, appeals, restrictions and repeated edits after rejection all sit on the record - at account, portfolio, page and domain level.
• Continuity. Sustained activity beats bursts. Long dormancy degrades what the system knows about your audience even if the account is old.
• Page and business standing. Page quality, feedback scores, verification status and the health of the rest of the portfolio.

Note what is absent: the calendar. An account that has existed for four years and spent nothing for three of them is not mature. An account that has spent consistently and cleanly for eight months usually is.

Every ad set re-enters a learning phase after creation or a significant edit, and stays there until it accumulates roughly fifty optimised events in a week. Underneath that number, performance is unstable and reported costs are close to meaningless. Small advertisers commonly live permanently below it - which is why their results feel random.

The escape is structural, not tactical. Consolidate: fewer ad sets, larger budgets, broader audiences so the events land in one place. Optimise for an event that actually happens often enough - for a property developer, qualified lead rather than signed reservation; for a hotel, initiate checkout rather than only completed booking - and use the rarer event to validate quality afterwards. And stop editing: every meaningful change resets the learning you just paid for.

This is also where a serious measurement plan earns its keep. If the business cannot say which single event defines success, the account will be optimised toward whichever event was easiest to install.

The fastest way to destroy a working account is to double the budget on a Monday morning. Large sudden increases push an ad set back into learning, change the auction dynamics it had solved, and frequently produce a week of worse results that gets diagnosed as creative fatigue.

Scale in steps of roughly twenty to thirty percent, spaced far enough apart that each step exits learning before the next begins - three to four days is a reasonable rhythm. If a step must be larger, duplicate rather than edit, so the proven ad set keeps running while the larger one stabilises. Add new creative inside existing structures instead of building new campaigns for every idea, and keep the optimisation event fixed for as long as you can bear.

A new account is not a smaller version of a mature one - it has less information, a lower ceiling and no benefit of the doubt. Treat the first six weeks as deliberate evidence-building rather than a performance test.

• Verify the business and the domain before the first campaign, not after the first problem.
• Install the dataset properly - pixel plus conversions API, deduplicated - and confirm events fire before spending. Learning from a broken signal is worse than no learning.
• Start with a modest, continuous daily budget rather than a large burst. Continuity is what is being demonstrated.
• Begin with a mid-funnel event that has volume; move to the scarcer, more valuable event once the volume exists.
• Launch conservative creative first. An early policy rejection on a new account is disproportionately expensive.
• Do not change currency, time zone, payment method and optimisation event in the same week. When something breaks you want one variable to look at.

Done this way, an account becomes an asset with a balance sheet value: it delivers more efficiently than a new one for the same money, and that gap is precisely what you lose every time an account is abandoned and rebuilt. Protect it accordingly - and read the restrictions piece before you find out the hard way what puts it at risk.