My Stripe account got frozen at 3am — what now

It is 3:07am. Your phone buzzed because the Slack webhook that monitors Stripe's account status API threw a red. The dashboard says "Your account has been temporarily suspended pending review." Checkout on the flagship brand is already dark. Customers in Australia and the UK are hitting decline walls that look like your fault.

You are going to want to log in and start clicking. Do not. The first 30 minutes of a freeze determine whether you recover in 5 days or 25 days. Here is what a multi-brand operator who has been through this four times actually does.

1. Minute 0-10: confirm the blast radius

Open three tabs: Stripe dashboard, your status page, and Google Analytics real-time. Confirm (a) which account is frozen — if you run multiple Stripe entities one freeze does not always cascade, (b) whether new charges are failing or the entire account is read-only, (c) whether payouts are also frozen or only new capture. The combination tells you whether this is a risk review, a compliance review, or a chargeback spike reaction.

2. Minute 10-30: do nothing destructive

No refunds, no mass customer emails, no bank account edits, no "update your business description" edits. Every action during a freeze window is logged and the review team treats activity as evidence. If you refund 400 customers in a panic to "clear the queue," you just handed Stripe's risk model a flag pattern that looks like a fleeing operator.

3. Minute 30-60: redirect checkout

If you have a backup processor, flip the router. If you do not, replace the checkout button with a holding page: "We are completing a payments system upgrade. Enter your email and we will follow up within 24 hours with a payment link." This recovers intent, which is the thing you cannot buy back later. A customer who hits a generic decline wall assumes you are fraud and moves on. A customer who sees a planned-maintenance page waits.

4. Hour 1-2: wake the right people

Your agent of record at the acquirer. Your bookkeeper (because you are about to stand up a new MID and the 1099-K math will matter at year end). Your head of operations (because automated rebills on the frozen account need to be paused before they queue up as failed charges that hurt you). Do not wake your lawyer at 3am unless you have a Reg Z or money transmission angle; a freeze is a commercial dispute, not a legal one, until it stops being.

5. Hour 2-8: read the freeze email like a lawyer

"Temporarily suspended pending review" is different from "permanently terminated" is different from "unable to continue processing." The first gives you 5-15 days and a good chance of partial reinstatement. The second means MATCH-list risk. The third usually means Stripe found something in underwriting that they cannot get comfortable with and you are not coming back. The language determines whether you fight the freeze or move past it.

6. Hour 8-24: apply to a backup rail

Every high-risk operator should have a backup MID pre-approved before they need it. If you do not, today is the day you learn. Apply to 2-3 acquirers in parallel — Payment Plugins for Stripe, Durango, Soar, PaymentCloud, Authorize.net through a high-risk reseller. Approval takes 3-10 business days at best. The applications you start this morning become your rescue rail by next Monday.

7. Day 2-3: respond to Stripe's questions, once

Stripe's review team will ask for documents — usually bank statements, supplier invoices, refund policy, and a business model summary. Respond once, cleanly, with everything they asked for and nothing extra. Do not chase up every 6 hours. Do not escalate to three different support reps. Reviews that get multiple touches from the merchant stretch longer than reviews that come in as single clean packages.

8. Day 3-7: recover held funds

If the freeze is "pending review," Stripe typically holds funds for 90-180 days from the last transaction date. Some reviews release early; some do not. Build a spreadsheet of outstanding captures and settlements so that when the release window opens you know exactly what you are chasing. For strategy, see reserve release playbook.

9. Day 7-14: migrate the portfolio

This is where multi-brand operators either win or learn. If you have 8 brands on one Stripe, the freeze took them all dark. If you have 8 brands on an orchestration layer routing to 2-3 acquirers, the freeze took ~33% of volume dark and the rest kept running. See Stripe vs multi-brand orchestration.

10. Day 14-30: write the postmortem

What triggered the freeze. What would have caught it 30 days earlier. Which KPIs you are now watching daily. What backup MID coverage you now have and how you are rotating volume to keep each MID \"warm\" so none of them go cold-review when you actually need them. This postmortem is what separates operators who get frozen twice a year from operators who get frozen once and never again.

11. 3am detection: monitors worth running

Stripe does not tell you the account is suspended until you log in or try to capture. Build a synthetic probe — a $0 auth every 15 minutes to a test card — that pages you on failure. Pipe account.updated webhooks into Slack. Watch charge.failed rates in 5-minute buckets. The 2-3am window is when most freezes trip because Stripe runs risk model batches overnight.

12. What to tell your team at 4am

"Stripe is in review. We are running the playbook. Checkout is on the holding page. No customer action needed until 9am Pacific. Do not refund. Do not mass-message. I will update in 2 hours." Calm language beats panic. The team follows the tone you set between 3 and 5am more than any SOP you wrote in March.

Emergency sequence

• Confirm blast radius across all Stripe entities before acting.
• Holding page up in first hour; no destructive clicks.
• 2-3 backup MID applications submitted by end of day 1.
• One clean response to Stripe review with all requested docs.
• Portfolio migration plan in writing by day 7.
• Postmortem + monitor upgrades by day 30.

The actual fix is pre-freeze

Everything above is triage. The durable fix is to not be single-rail in the first place. Running multi-brand orchestration means one freeze takes down one slice, not the whole business. See the pricing page for what orchestration costs versus what a single freeze costs in downtime, or skip to the application if you already know.