The email arrives on a Friday, because it always does. Your platform access ends on a date measured in days, or it has already ended. Somewhere between two hundred and twenty thousand traders are mid-evaluation, several are waiting on payouts, and your storefront is still happily selling challenges you may not be able to deliver.
What follows is the sequence we walk clients through. It assumes the worst version: short notice, no warning, and no alternative platform already running. If you are reading this before it happens, read deplatforming risk instead and make sure you never need this page.
Hours 0–4: Establish Scope and Preserve Everything
Two jobs, in parallel, and both are clocked. Nothing else matters yet.
- Pin down the exact scope in writing. Which accounts, which servers, which date and time, and — critically — whether you get a read-only window after trading stops. Get it from the party that can actually enforce it, which is often your broker or intermediary rather than the vendor. Ambiguity here produces every subsequent mistake.
- Export while you still can. Closed trade history, open positions with entry prices, account balances and equity, evaluation state per account (phase, day counter, profit target progress), drawdown baselines and equity high-water marks, and the full ledger of deposits, fees and payouts. Export in the most raw format available, twice, to two destinations. If your CRM is already the system of record — synced continuously from the platform — this step is a verification rather than a rescue, which is the entire argument for having one.
The irreversible mistake: letting access lapse before the data is out. Everything else on this page can be recovered from, slowly. A lost trade history cannot — it is the evidence behind every pass, every breach, every payout and every dispute you will have for the next two years, and it is the first thing a regulator, a processor or a comparison site will ask you to produce.
Hours 0–4: Stop Selling
Simultaneously, and without waiting for a plan: pause new challenge sales on the affected platform. Disable the checkout links, take the buy buttons down, switch the storefront to a holding state. It feels like the most expensive decision in the room and it is the cheapest one you will make. Selling an evaluation you cannot deliver converts a platform problem into a refunds problem, a chargeback problem, and — where the sale was made knowing delivery was doubtful — a problem with a much worse name. Decide your refund posture now too, before the first ticket arrives: full refunds on anything sold in the last 48 hours is the default that ages well. See refund handling.
Hours 4–12: Tell the Truth, Once, in Public
Silence is what kills firms in this scenario. Traders discover platform outages within minutes and fill the vacuum with the worst available explanation, which in this industry is "they are about to run off with the money". You get one chance to be the primary source.
One public statement, on your own domain, linked from every channel, containing:
- What happened, in plain terms, without blaming a party you are still negotiating with.
- What it means per cohort — traders mid-evaluation, funded traders with open positions, traders with payouts pending, and traders who bought in the last few days. Every trader must find their own situation named.
- What is protected: evaluation progress preserved, time-limited clocks paused, payouts continuing on schedule. Only promise what you have already verified you can do.
- When the next update lands — a specific time, not "shortly" — and then hit it, even if the update is "no change yet".
Behind the statement: support macros for the five questions that will arrive, a status page, and a separate note to affiliates and partners, who are about to be asked by their audiences and deserve to answer accurately. Your support desk earns its budget this week.
Hours 12–48: Stand up the Alternative and Rebuild State
Now the technical work. With a CRM holding your system of record and 1-click bridges, connecting a second platform is a configuration task, not a procurement cycle. The hard part is not the connection — it is recreating state correctly:
- Recreate accounts at their true position: current balance and equity, not starting balance. Getting this wrong in the firm's favour is the single fastest way to a public dispute.
- Carry the evaluation clock. Days traded, minimum-day requirements and time limits must resume where they stopped. Pause time-limited challenges for the outage window and say so publicly — nobody should fail a challenge because of your vendor.
- Reset drawdown baselines carefully. Trailing drawdown high-water marks and daily-loss baselines have to be transplanted, not recalculated from a fresh starting balance. This is the detail migrations get wrong most often — see drawdown rules.
- Verify the rules engine fires identically. If enforcement lives in the CRM's rules engine rather than in platform plugins, the rule set is unchanged and only the data source moves — test it on a sample before opening the doors.
- Handle open positions explicitly. Publish whether they were closed at termination, at what price, and how that is reflected. Never leave a trader to infer it.
The full non-emergency version of this move — parallel running, phased cohorts, comms — is in migrating off MetaTrader without losing your traders.
Hours 48–72: Keep the Money Moving
If you do exactly one thing right this week, make it this: payouts continue on schedule. A firm that pays through a platform crisis is remembered as well-run. A firm that pauses payouts "until things stabilise" is remembered as something else, permanently, in reviews that outlive the incident by years.
This is only possible if your payout pipeline is independent of the trading platform — eligibility computed from records the CRM holds, requests reviewed in a queue your staff control, disbursement through your PSP. Firms whose payout eligibility could only be calculated by querying the platform they just lost discovered the dependency at the worst possible moment. See payout rules and automated payouts.
Same window: refund anyone who bought into a product you could not deliver, without making them ask twice.
"Traders forgive an outage. They do not forgive silence, and they never forgive a missed payout."
— Roman Onta, Executive Director, Broker CRM & UI/UX
Day 4 Onwards: Make It Unrepeatable
Once trading is stable, the post-mortem has exactly one required output: an architecture in which this cannot happen again. Keep the second platform live rather than dismantling it. Confirm that your system of record is yours and exportable on demand. Establish who your vendors' actual counterparties are — if you are not the customer, find out who is. Negotiate notice periods and data-export rights into every contract you sign next. Then schedule a failover rehearsal twice a year, with a written runbook, and treat a failed drill as a real incident. The full audit is in deplatforming risk.
Key Takeaways
- Hours 0–4: pin the termination scope down in writing and export everything — trade history, open positions, balances, evaluation state, drawdown baselines and the ledger. Lost history is the only unrecoverable loss.
- Pause challenge sales immediately and set a refund posture. Selling an evaluation you cannot deliver turns a platform problem into a chargeback and conduct problem.
- One public statement within 12 hours, naming every trader cohort, promising only what you have verified, with a specific time for the next update — silence is what actually kills firms here.
- Rebuild account state (balance, clock, drawdown baselines) rather than recreating fresh accounts, pause time-limited challenges for the outage — and never pause payouts.
Frequently Asked Questions
What Is the Single Most Important Thing to Do First?
Export your data while access still exists — trade history, open positions, account and evaluation state, drawdown baselines and the ledger — and pause challenge sales at the same time. Everything else can be recovered from later; a lost trade history cannot, and it is the evidence behind every pass, breach, payout and dispute you will face afterwards.
Should Evaluations Be Paused, Reset or Voided?
Paused and carried over. Traders should resume at their true balance, equity, day count and drawdown baseline, with time-limited clocks stopped for the outage window. Resetting evaluations because of your vendor's decision is the fastest route to a permanent reputation problem, and voiding them is worse.
How Long Should a Platform Migration Realistically Take?
With a CRM holding the system of record and 1-click bridges available, connecting and configuring an alternative platform is days rather than months — the work that takes real time is recreating account state accurately and communicating with each trader cohort. Firms with no CRM spine and no second bridge should assume weeks.