Why hardening matters more than vendor tier
A fresh BM5 from a top-tier vendor and a fresh BM5 from a mid-tier vendor have a survival differential of roughly fifteen percentage points. A hardened BM5 from either vendor, by contrast, sits within two percentage points of every other hardened BM5. Hardening narrows the gap that vendor selection opens. That is the empirical pattern from the editorial test fleet, and it is why we publish a hardening protocol that sits independently of which vendor you bought from.
The fourteen-day protocol below is the version we apply to every fresh Manager that enters the test fleet. Buyers under quota pressure compress it to ten days; buyers running aggressive scaling extend it to twenty-one. The principles do not change. Slow start, predictable signal, signal-watching at every step.
The first seventy-two hours
Day zero is delivery day. Credentials arrive in your dashboard, you import the cookies into your antidetect browser, you set a residential proxy that matches the BM's registration country, and you log into Facebook with the owning profile. That is the entire activity for day zero. No BM access, no ad-account creation, no payment-method addition. Let Meta record the login as a normal session.
Day one is owning-profile validation. Browse the news feed for ten to fifteen minutes from the owning profile. Like one or two pages that match the kind of content the profile would naturally engage with. Send a friend request to a Meta-suggested connection. Do not yet visit Business Suite. Do not yet open Ads Manager. The point is to look like a returning user, not a buyer rushing to deploy budget.
Day two repeats day one with slightly elevated activity. Update the owning profile's profile picture if needed. Browse Marketplace for a few minutes. Comment on one post. Send a brief Messenger message to one connection if natural to the profile's history.
Day three is the first contact with Business Suite. Open it from the owning profile, look around, do not create or modify anything. Read any policy reminders Meta surfaces. Close the tab. Visiting and leaving is normal user behaviour; visiting and creating is the automated-buyer pattern that triggers review.
Week one: owning profile and billing
Days four through seven are the billing-history seeding phase. The goal is to teach Meta's review pipeline that this BM has a real billing relationship with a real entity. We add a primary payment method on day four — typically a virtual card from a reputable issuer with billing address that matches the BM's claimed business location. The mismatch between card BIN country and Manager country is a documented review trigger; matching them is foundational.
Day five is for verifying the payment method but not spending. Some payment processors run a one-dollar verification charge that drops off automatically. That charge is a real billing event and Meta records it. Let it settle for forty-eight hours.
Day six and seven are for browsing Ads Manager from inside the BM. Open the campaign creation flow, look at the targeting options, do not actually create a campaign. Look at the audience insights tool. Look at the performance tab. Each of these visits is a small trust-building event that costs nothing and adds up.
Week two: ad-account naming and creative
Day eight to fourteen is the campaign-launch phase. Launch one campaign in one ad-account inside the BM. Spend cap is fixed at fifteen dollars per day. Targeting is broad, creative is white-vertical, naming is descriptive — 'Q2-2026 Brand Awareness Campaign' rather than 'test-1'. Ad-account names appear in Meta's review tooling, and descriptive names read as genuine business activity while generic names read as automated buying patterns.
Slot two ad-accounts on day eleven, but do not create campaigns in them. The presence of unused ad-accounts inside a Manager is normal — businesses provision capacity ahead of need. Campaigns in two slots out of five looks like an active small business; campaigns in five out of five on day eleven looks like a buyer rushing to deploy.
By day fourteen the BM has shown Meta a complete pattern: real owning profile, real billing relationship, descriptive naming, single-campaign starting volume. That pattern is what Meta's review pipeline rewards. Cost-per-click should drop noticeably from week one's baseline, and the BM is now in a hardened operational state.
Three signals that say slow down
Cost-per-click well above target during week one is the earliest warning that Meta is filtering. If CPC sits forty percent or more above market for two consecutive days, slow the protocol by three to five days. Do not advance.
Spend-cap behaviour that does not match the listing is the second signal. If the cap is meaningfully lower than what the vendor advertised, the BM is mis-described — file a replacement claim within the twenty-four-hour window.
Soft policy warnings during the protocol window are the third signal. They indicate Meta's review pipeline has flagged something. Resolve the warning, retreat to a safer creative for forty-eight hours, then return to schedule.
Common mistakes we still see
Cold-loading the BM on day one with a high-budget campaign and a creative pulled from another network. The single most common reason fresh BMs fail.
Generic ad-account naming. 'Test 1', 'Account A', 'BM Slot 2' are all read as automated buying patterns. Use descriptive names that match a plausible business.
Skipping the residential proxy. Datacenter proxies, mismatched-GEO proxies, and shared proxies all degrade the BM's trust signal during hardening.
Adding the payment method and immediately attempting a high spend. Let the payment relationship settle for forty-eight hours before any spend.
From the comments (38 total)
Descriptive ad-account naming was the missing step in our flow. Survival rate jumped after we changed.
Compressed the protocol to ten days for one BM and saw the predicted survival drop. Going back to fourteen for the next batch.