How Thai Sellers Should Reconcile PromptPay, Card, and Shopee Payouts in 2026

Run a skincare shop out of a stall in Bangkok's Chatuchak market, sell through your own site, Shopee, and a LINE OA, and by week two the bank statement stops making sense. PromptPay transfers land one by one with no order reference. Card settlements show up in batches with the fees already stripped out. Shopee, Lazada, and TikTok Shop each pay out on a different cycle, with their own deduction logic bolted on top. The accountant still needs books that tie out in FlowAccount, PEAK, or a spreadsheet, with a tax invoice trail the Revenue Department will actually accept.

The matching order is the same whether you're in Bangkok, Ho Chi Minh City, or Jakarta: order ID first, settlement batch second, payment method third. Don't start from the bank statement alone — keep one daily export per channel and match each order to a settlement line before checking the balance.

What to prepare first

• One order export covering every channel: your own site, Shopee, Lazada, TikTok Shop, and LINE chat orders. If orders live in five separate back offices, reconciliation is dead before it starts. Zort syncs Shopee, Lazada, TikTok Shop, and offline POS into one dashboard from ฿600 a month — worth it the day you cross two channels, overkill if you're Shopee-only.
• A separate settlement export per rail: PromptPay credits from the bank, card settlement from the gateway, and the payout statement from each marketplace seller centre.
• A fee map per rail. Beam takes no merchant fee on PromptPay QR. That's the cheap default for a QR-only shop. Cards start around a 1.8% discount rate, and that gap is where gross and net stop matching.
• A cut-off rule, and stick to it. Marketplace payouts lag orders by days, so decide up front whether revenue books at order confirmation or at payout — mixing the two gets flagged in a PP.30 review.

Daily workflow

1. Export yesterday's orders from every channel into one sheet, keyed by order ID.
2. Pull the PromptPay credit list from the bank and the card settlement report from the gateway.
3. Match order IDs to settlement lines first. Only then compare the matched total against the actual bank deposit.
4. Park every unmatched line in an exceptions list with a reason code. Don't let it roll forward — a ฿3,000 gap in March is a ฿30,000 mystery by December.

Fifteen minutes a day beats four hours at month-end. Yesterday's mismatch is still explainable: the customer who paid twice, the QR that expired, the test order that shipped by mistake.

PromptPay reconciliation

PromptPay is instant but anonymous. A bank statement line tells you who paid and when, not which order it covers.

• Dynamic QR per order. A QR generated per checkout carries the exact amount, so matching it against an open order is usually safe. Two identical amounts on the same day are the classic false match — flag them instead of auto-matching.

Route PromptPay through a payment gateway instead of a bare bank QR, and every credit shows up already tagged with an order reference. HitPay does exactly that — PromptPay and PayNow credits feed straight into bookkeeping, and the free tier charges per transaction instead of a flat fee. That suits a shop under a few hundred orders a month; past that, check a flat-fee gateway before assuming HitPay stays cheaper.

• Slip verification. Chat orders paid by transfer still arrive as a slip photo in LINE. Verify the slip against the statement same-day. A slip that never lands in the bank is the single most common loss in Thai chat commerce. Catching it same-day takes five minutes, not a month-end audit.

Card and e-wallet reconciliation

Cards settle in batches, net of fees, so reconcile the batch rather than the transaction. The gateway's settlement report lists which orders a deposit covers and the fee taken per order. Book that fee as an expense line. Net it silently into revenue and you understate both revenue and cost, VAT base included. TrueMoney, ShopeePay, and LINE Pay behave the same way as cards: batch payouts, per-transaction fees, their own report to chase down.

Marketplace payout reconciliation

Shopee, Lazada, and TikTok Shop each run their own economy inside the payout. Commission, transaction fee, campaign co-funding, shipping subsidies, and returns all net out before the money lands. Match it in two steps. First, the seller centre's income statement against your order list, per order. Second, the payout total against the bank deposit, per cycle. The number worth watching is the effective take rate per marketplace per month — when it drifts, a fee changed somewhere, or a campaign setting quietly started eating margin. Zort pulls the order side of this match from a single export instead of three separate seller centres, which is the entire reason sellers running three marketplaces end up paying for it.

Accounting handoff

Thai books need a Thai paper trail: tax invoices (ใบกำกับภาษี), PP.30 VAT reports, and withholding certificates. FlowAccount starts at ฿165 a month and covers most solo sellers fine. PEAK starts higher, at ฿208 a month. It earns that extra ฿43 with native e-Tax Invoice and PP.30 automation, plus bank reconciliation built for Thai banks — worth it once an accountant is actually involved, not just you and a spreadsheet. Hand the accountant matched data, not raw bank statements — the reconciliation context doesn't survive the handoff otherwise.

Month-end close checklist

• Exceptions list at zero, or every surviving line has an owner and a reason.
• Effective take rate per marketplace, computed and checked against last month.
• Card and e-wallet fees booked as expenses, never netted into revenue.
• PP.30 revenue total ties to the matched order export, not to bank deposits.
• Payouts in transit (ordered this month, paid next) listed explicitly, not guessed at.

Common mismatch cases

• Same-amount PromptPay collisions. Two ฿590 orders, one credit. Resolve it by timestamp and customer name, never by assumption — a lucky guess today is a mismatched refund next week.
• Marketplace return after payout. The clawback lands in a later cycle. Match it to the original order, not to the cycle it shows up in.
• Expired QR, paid anyway. A customer scans a stale QR, the money arrives, and no open order is waiting for it. The exceptions list catches this one.
• COD remittance shortfalls. Courier COD lands as a lump sum days later, common for sellers shipping up to Chiang Mai and other upcountry provinces. Treat it as its own rail with its own export.

Build the close before volume arrives

Reconciliation stays boring until a campaign doubles order volume overnight and the books stop tying out. Sellers who get through an audit without panic already had that daily fifteen-minute match running before the spike hit. Set the exports up this week. Keep the order fixed — order ID, then batch, then method — and start tomorrow morning with yesterday's orders. That's it. Don't wait for month-end to find out the system doesn't work.