Forecast Cash Flow from Marketplace Payment Cycles

Why marketplace sales are a poor proxy for cash

Gross sales on a marketplace dashboard are not cash. Between an order and the money in your bank account sit delivery, the return window, the settlement cycle and a set of deductions. If you plan supplier payments, salaries or inventory buying from sales figures, you will overestimate the cash you have.

Three things make marketplace cash different from a normal receivables ledger:

• Timing is event-driven. Payout usually depends on dispatch, delivery or the end of a return window, not on the order date.
• Amounts change after the sale. Returns, cancellations, fee adjustments and penalties reduce what you finally receive.
• Each marketplace has its own rules. Cycle length, hold periods and deduction heads differ, so one blended assumption rarely works.

Step 1: Map the settlement cycle for each marketplace

Start with documents, not guesses. For every marketplace, note down how the payout is triggered and when it is released. Look in your seller agreement, the seller panel's payment policy pages, and past settlement reports.

1. Trigger event: delivery date, dispatch date or return-window expiry.
2. Lag: the number of days from the trigger to the payout, and whether payouts run daily, weekly or on fixed dates.
3. Holds and reserves: any amount withheld for returns or disputes, and when it is released.
4. Bank holidays: whether the payout shifts when a payout date falls on a holiday.

Then check the map against history. Take the last 8 to 12 settlements and measure the actual gap between trigger and credit. Policies change, so confirm the latest terms with your account manager or the current seller policy before locking assumptions.

Step 2: Convert sales into expected cash

Build a simple bridge from gross sales to expected settlement. Every line should be traceable to a source document.

• Gross sales delivered in the period, from the order or sales report.
• Less returns and cancellations, using your own return rate by category, not a company-wide average.
• Less commission and fees, as per the rate card or commercial terms in your agreement. Do not use remembered percentages; pull them from the current document.
• Less taxes collected or deducted at source, as applicable. Confirm treatment and rates with your tax adviser and the settlement report.
• Less advertising and other charges that the marketplace adjusts against payouts.
• Add or subtract adjustments such as reversals, penalties and earlier short payments recovered.

The result is the expected settlement. Keep return rates and fee assumptions as separate input cells so you can stress-test them.

Step 3: Place cash in the right week

Use a weekly grid for 13 weeks. Each row is a marketplace and each column a week. Sales delivered in a week flow into the week in which the payout lands, based on your Step 1 lag.

Run three versions: base, slow (payouts delayed by a week) and high returns. The gap between base and slow shows how much working capital buffer you need.

Step 4: Compare forecast with actual settlements every week

A forecast is only useful if you learn from the variance. After each settlement, reconcile the credited amount against the expected amount and tag the difference: timing, return rate, fee mismatch, missing order or unexplained.

Fee and commission differences deserve special attention because they repeat. If a deduction does not match the agreed terms, it will recur on every payout until someone raises it. Tools such as MarketplaceMetrics by Fundle reconcile sales against settlements, check commissions and fees against agreed terms, and track returns and recovery cases, so variances become a worklist rather than a surprise. Myntra Excel reports are reviewed today, and other marketplaces are in scope discussion.

Feed the findings back into the model: update the actual lag, return rates and recurring deductions each month.

Common mistakes to avoid

• Using order date as the cash date. Always use the payout trigger plus lag.
• One return rate for everything. Fashion, footwear and accessories behave differently.
• Ignoring recoverable amounts. Short payments and disputed deductions are real cash, but only count them once a claim is accepted.
• Forgetting seasonal sale periods. Large sale events raise volume, returns and ad spend together, often stretching cash needs before payouts arrive.
• Not versioning assumptions. Record when a policy or rate card changed so older variances are explainable.