# Shopify and Amazon Accounting Integration | Tembrant

> Custom Shopify and Amazon accounting integration: each payout split into sales, fees and refunds, matched to the bank deposit and posted to Xero or QuickBooks.

Source: https://tembrant.com/ecommerce-accounting-integration
Last updated: 2026-11-21

[E-commerce operations](https://tembrant.com/ecommerce-operations) / Shopify and Amazon accounting integration: payouts reconciled into Xero or QuickBooks

# Shopify and Amazon accounting integration: payouts reconciled into Xero or QuickBooks

Every payout broken into the orders, fees and refunds behind it, tied to the bank deposit, then posted as one balanced entry.

[Start a project](https://tembrant.com/contact)

On this page

* [When custom beats a connector app](#when-custom-beats-a-connector-app)
* [How we would build it](#how-we-would-build-it)
* [Proof from our own code](#proof-from-our-own-code)
* [What the first version includes](#what-the-first-version-includes)
* [How the work is run](#how-the-work-is-run)

A Shopify payout lands in the bank as one deposit. Behind it are hundreds of orders, card fees, refunds, chargebacks and perhaps a reserve. Amazon’s settlements add referral and fulfillment fees, advertising charges and reimbursements, on their own schedule. Then the bookkeeper has to make the books agree with the bank.

When the numbers don’t meet, the difference usually goes into a clearing account, and month after month that account holds a balance nobody can explain. Revenue is booked net of fees, fees are lumped together, and the sales tax figures are hard to defend.

## When custom beats a connector app

Connector apps that post daily or per-payout summaries into Xero or QuickBooks work for many sellers. Start there if one fits.

A custom integration is worth building when:

* You sell on several stores, marketplaces or currencies, and each settles differently.
* Your clearing accounts never return to zero.
* Your accountant wants fees, refunds and sales tax on their own accounts, by channel.
* You need to trace any posted number back to the orders that make it up.

## How we would build it

### The payout is the unit

The system would pull each channel’s settlement data through its API: Shopify’s payouts and balance transactions, and the settlement reports and financial events in Amazon’s Selling Partner API. Payment processors such as Stripe and PayPal work the same way. Each payout is stored with every line behind it: order, refund, fee, adjustment, reserve, transfer.

### Check the payout against itself first

Before anything is matched or posted, the lines must add up to the amount paid out, exactly, in cents. A payout that doesn’t add up is held with the difference shown. It is never posted with a plug.

### Match the payout to the bank deposit

Each payout is matched to a bank deposit. A shared payout reference wins first. Then amount and date are matched within a short window, closest dates first. Bank lines come from a bank feed or from statements read by a [bank statement parser](https://tembrant.com/bank-statement-parser-development). Every match records how it was made and can be approved or rejected.

### Post one balanced entry per payout

Each matched payout becomes one journal: gross sales, sales tax collected, refunds, each fee type, reserves and the net deposit, mapped to your chart of accounts. A journal that doesn’t balance to zero is refused before it is sent. The system would post to Xero or QuickBooks Online through their accounting APIs, keeping the payout ID on each journal so a second run can’t post it twice.

### A clearing account that empties

Sales are recorded when orders happen. Payouts clear them when the cash arrives. At any date, each channel’s clearing balance equals the payouts not yet received, and the system can list them. If the balance differs from that list, the review queue says by how much and why.

### Closed months stay closed

When a month is closed in your books, late adjustments for it are posted in the open month with a reference to the original, not pushed into the closed one.

The matching rules and exception design come from our [reconciliation software development](https://tembrant.com/reconciliation-software-development) work.

## Proof from our own code

We have not shipped a marketplace integration. These checks already run in production in [Lekhavo](https://tembrant.com/products/lekhavo), our bookkeeping product for Indian accounting practices:

* **Statements checked against themselves.** If a statement’s opening balance plus its transactions misses its closing balance, the import is held before any transaction is committed.
* **Two-pass matching.** Lekhavo’s transfer matcher pairs a debit and a credit that share a bank reference within two days. It then pairs the rest by equal amount, taking every same-day pair before any one-day pair, and every one-day pair before any two-day pair. Each match records its method and waits for approval.
* **Balanced exports or none.** Lekhavo’s export to Tally, an accounting package widely used in India, refuses any voucher that doesn’t balance to zero paise.
* **Locked years.** Once a financial year is locked, a change to a match touching that year is refused.

## What the first version includes

* Payout import from your main store and one marketplace or processor
* Payout self-checks, bank matching and a review queue
* Balanced journals posted to Xero or QuickBooks Online with your account mapping
* A clearing-account report per channel
* An audit trail of every match, post and override
* Source code, a written spec and runbooks

If invoices also need to come out of the same orders, see [invoicing software development](https://tembrant.com/ecommerce-invoicing-software-development).

## How the work is run

Tembrant builds with an AI software factory: many agents work in parallel, held to fixed gates. Every change starts from a written spec, tests are written first and run against your real settlement files, each change gets independent reviews, and a person signs off before it ships.

If your marketplace clearing account hasn’t been zero in months, [send us a description of your channels and ledger](https://tembrant.com/contact).

## Questions

Why not use a connector app?

If a connector app posts your payouts the way your accountant wants, use it. Custom makes sense when you sell on several channels and currencies, when your clearing accounts never get back to zero, or when you need to trace any posted figure back to the orders behind it.

Which accounting systems can it post to?

We'd post to Xero and QuickBooks Online through their accounting APIs as balanced journals, using your chart of accounts. Other ledgers can take the same entries through their own import format or API.

What happens when a payout doesn't add up?

It isn't posted. If a payout's lines don't sum to the amount paid, or no bank deposit matches it, the payout goes to a review queue with the difference and the lines involved. Nothing reaches the ledger with a gap nobody mentioned.

Have you built this for a Shopify or Amazon seller?

Not yet. Our production reconciliation engines match Indian bank statements to ledgers and bond interest schedules to their totals. A marketplace payout has the same shape: a settlement, the transactions behind it and a deposit that should equal their net. We'd build yours with the same design, scoped to your channels' settlement data.

## Related

* [Custom reconciliation software development](https://tembrant.com/reconciliation-software-development)
* [Custom invoicing software development for online sellers](https://tembrant.com/ecommerce-invoicing-software-development)
* [Custom bank statement parser development](https://tembrant.com/bank-statement-parser-development)
* [E-commerce operations overview](https://tembrant.com/ecommerce-operations)

## Have financial data that has to tie out?

Tell us what you want to build. We reply with questions, a plan and an estimate.

[Start a project](https://tembrant.com/contact)

![](https://tembrant.com/emblem.webp)
