When Stripe holds your funds for the full 180-day reserve period, the cash gap is real. The math: a company doing $480k/month with payouts frozen for 180 days is looking at roughly $2.9M of frozen working capital. The four moves that bridge it: stand up a parallel processor immediately, run a customer prepayment push, draw on an AR-backed credit line treating the Stripe balance as AR, and (in cases of bad-faith hold) escalate legally. Done right, you can replace 60–80% of the frozen cash within 30 days.
Stripe's Services Agreement gives them the right to hold a merchant's balance for up to 180 days after account termination or pause. In our 2024-2025 data set of 14 AI companies whose accounts were paused, six were held for the full 180 days. The median hold was 132 days. The shortest was 47.
If you're reading this in a pause: the cash is going to come back. The question is whether you survive the gap. Most of the failures we've watched were not from the pause itself but from running out of runway 60–120 days into it because nobody had a cash plan.
Why 180 days, exactly?
180 days is the maximum chargeback window for most card networks. Visa and Mastercard allow disputes to be raised up to 120 days from the transaction date (in some cases up to 540 days for specific dispute categories). The 180-day Stripe hold is calibrated to cover the realistic chargeback exposure on the last batch of transactions you processed before the pause.
Inside Stripe's SSA, section 6.5 (or equivalent in your jurisdiction) covers the right to hold reserves up to this maximum. In practice, the hold duration depends on:
- How recent your last successful charges were (the chargeback exposure decays from those).
- Your historical chargeback rate (higher rate → longer hold).
- Whether refunds / partial returns are still being processed against existing transactions.
- Whether Stripe has any other open issues with your account (KYC, content, etc.).
What Stripe does with the funds during the hold
Stripe holds the funds in a pooled non-interest-bearing account. The funds remain legally yours but are subject to a security interest in Stripe's favour. During the hold:
- Existing refunds are honored. If a customer requests a refund and the original transaction was within Stripe's refund window (typically 120 days), it comes out of the held balance.
- Chargebacks are honored. Disputes filed within the network window come out of the held balance, plus the $15 dispute fee per case.
- At day 180 (or whatever Stripe decides), the remaining balance is released to your bank account in a single transfer.
You don't earn interest. The balance doesn't accrue dividends. It's dead money for the duration of the hold.
The cash-flow problem (with real numbers)
Let's run a concrete scenario. An AI inference company at $480k/month MRR, moderate growth, 18 employees, $4.3M raised. The Stripe pause happens on day zero.
| Component | Amount |
|---|---|
| Held balance at pause (last 14 days of receipts) | $224,000 |
| Monthly burn (payroll + infra + AI inference cost) | $295,000 |
| Cash on hand at pause | $1.8M |
| Implied runway pre-pause | ~6.1 months |
| Implied runway post-pause (no replacement revenue) | ~6.1 months — until day 180+ release of held balance |
| Implied runway if revenue stays at $0/mo through hold | 5.4 months on cash, then $224k recovered + needs new processor |
The scary line: no replacement revenue. Most pauses don't stop revenue instantly — existing subscriptions usually keep processing for 2–6 weeks before Stripe pulls them. But new signups stop on day one, and over 60–90 days the existing base churns or downgrades.
Without replacement revenue, this company runs out of cash roughly day 145, which is before the held balance is released. That's the failure mode we see most often.
Move 1: stand up a parallel revenue rail in < 72 hours
The single highest-leverage move. If new signups can route through a different processor (Macropay, Paddle, Adyen, Braintree) within 72 hours of the pause, you protect new revenue and you start replacing the cash flow.
At Macropay our migration intake is built for this case. We can stand up a working checkout in < 4 hours and run a parallel-write window so existing Stripe subscriptions continue processing in Stripe while new signups go through us.
Expected cash recovery from this move: 50–75% of normal monthly revenue within 14 days, depending on signup velocity.
Move 2: customer prepayment push (raises 30-60 days of runway in a week)
Email your top 30-50 customers. Offer a 10-15% discount on annual prepayment paid via the new processor (or ACH / wire if available).
In our 2025 data, this move raises an average of $84k per AI company for companies at $300k-$500k MRR. The high responders are usually mid-market customers ($1k-$5k/month) who use the budget cycle. Consumer customers rarely prepay annually even at 20% off.
The script that works:
Hi [name],
Quick note — we're consolidating our billing infrastructure this month and rolling out a few customer-friendly changes. As part of that, we're offering a 15% discount on annual prepayment, available through the end of [date].
For your account at [$X/month], that's [$Y] for 12 months, billed today.
Reply "yes" and I'll send the invoice via [new processor]. Happy to set up a quick call if it's helpful.
[name]
Don't mention the Stripe pause. The customer doesn't need to know, and once the new processor is in place they won't notice the change.
Move 3: AR-backed credit line, treating held balance as AR
Most venture-debt facilities and revenue-based financing arrangements allow draws of up to 70-80% of accounts receivable. The held Stripe balance is, technically, AR. The lender will discount it (the discount depends on your relationship with the lender, the hold duration, and Stripe's historical record of releasing on time), but you can usually draw 40-60% of the held amount.
For the scenario above, a $224k held balance could yield ~$110k of credit-line draw.
Lenders who'll typically engage with this:
- SVB / First Citizens venture debt (if you have an existing facility).
- Brex / Mercury credit lines (smaller draws, faster turnaround).
- Specialist revenue-based financing (Capchase, Pipe, etc.) — they understand AR-backed AI revenue.
- For larger holds, factoring firms — though factoring rates (8-15% APR) are usually worse than a credit line.
Move 4: the legal nuclear option
If Stripe holds funds in bad faith — i.e. without a documented chargeback or refund exposure justifying the full hold duration — the legal options are real but slow.
- Arbitration under Stripe's SSA — the contract requires AAA arbitration. Timeline: 4-9 months. Cost: $15k-$60k. Outcome rate (in our data): about 50%.
- Small-claims court — for amounts under your state's small-claims cap ($10k in CA, $5k in NY). Faster but limited by the cap.
- Demand letter from counsel — usually $2k-$5k. Sometimes gets a partial release. Sometimes gets ignored.
- Coordinated complaint to CFPB / state AG — works occasionally for egregious cases. Slow.
Legal escalation rarely accelerates Stripe's timeline meaningfully. In our 14 cases, only 2 had legal action that materially shortened the hold. The others were released on Stripe's own schedule. Don't spend money on legal before you've exhausted the first three moves.
What to do before this happens
If your Stripe account is currently fine but you're reading this proactively:
- Set up a second processor today. Cost: a few hours of integration. Macropay's sandbox lets you wire up a parallel checkout that's dormant until you flip a flag.
- Keep 60 days of operating cash outside Stripe. Anything in a Stripe balance is at-risk; anything in your bank is not.
- Document your customer base. A simple export of customer names, emails, and amounts. If you ever have to do the "prepayment push," you don't want to be exporting data from a paused dashboard at midnight.
- Keep your Stripe risk score clean. The six signals from our companion piece on why AI companies get flagged are how the pause starts.
We had our second processor on standby for 14 months before we needed it. When the pause came, the parallel checkout was already there. Customers never noticed. The $224k Stripe held came back six months later as a windfall, not a lifeline.— Helena Park, COO, Pact Labs
If you're reading this in the middle of a pause and need the 72-hour version of the playbook, our recovery guide is here. For the underlying signals that triggered the pause, the radar diagnostic walks each one. And if you want to model what your post-pause stack looks like with an MoR carrying the underwriting risk instead, the calculator is on /compare.