It depends on the split's charge schedule. Cashout splits charge when a cashout is detected. Earnings splits charge on a fixed cycle you pick when you create the split: weekly, bi-weekly, semi-monthly or monthly.
Every schedule boundary is UTC midnight.
The schedules
Schedule | Charges | Available on |
On platform payout | When a cashout is detected | Cashout splits |
Weekly | Every Monday | Earnings and cashout |
Bi-weekly | Every other Monday | Earnings and cashout |
Semi-monthly | The 1st and the 16th | Earnings and cashout |
Monthly | The 1st | Earnings and cashout |
Manual | Never automatically | All |
Weekly and bi-weekly always bill on Mondays. This is fixed policy, not a per-split setting — the start date decides when accrual begins, not which weekday you are billed.
Why your Monday charge appears on Sunday
Boundaries tick at midnight UTC, which is late afternoon in the US. A charge dated Monday is generated Sunday evening Pacific.
This is the single most common surprise with scheduled billing. Nothing has gone wrong — the cycle is correct, you are just seeing it in a different time zone.
Ledger invoice emails are held until 5:00 AM Pacific on the boundary date, so the email arrives at a reasonable hour even though the invoice was generated the evening before.
When the first charge lands
A schedule charges at the first boundary after the start date, not on it.
A semi-monthly split starting 1 August first charges on 16 August. A weekly split starting on a Monday first charges the following Monday. This is deliberate: it gives a full period to accrue before the creator sees a bill.
Bi-weekly and the alternating week
Bi-weekly bills every other Monday. Melon enforces a minimum gap of 13 days between charges, so a short cycle never causes two charges in consecutive weeks.
Manual schedules
A split set to manual never charges automatically. The balance still accrues and stays visible; nothing is collected until someone triggers it. This is used to pause billing during a reconciliation without losing data.
Related
