E-commerce Reconciliation Reference Build

Reference build

What the engine does, on data we control

Northshore Supply Co. is a fictional company and the dataset is synthetic. Its standard e-commerce accounting workflow has already run. The reference build shows how MatchPass can validate that baseline, investigate the recurring reconciliation work that remains, and route only the items that still require human review.

Reference build · synthetic data

Before · the monthly reconciliation evidence

4,812 supporting commerce, accounting, processor, wholesale and bank records
6 synthetic source datasets: commerce activity, existing connector and accounting records, PayPal, Stripe, wholesale and AR activity, and bank activity
The standard accounting workflow has already processed the normal commerce activity
173 residual items remain that would otherwise require recurring investigation

Build · the MatchPass method

7 high-level reconciliation rule families configured for this synthetic workflow
Baseline validation across source activity, existing accounting output and settlement evidence
Processor and off-platform reconciliation
Bundled settlement and remittance handling
Settlement-timing and adjustment logic
Evidence-backed exception routing, with ambiguous items never forced into a match

After · synthetic engine output

4,639 baseline records validated (96.40%)
173 residual items entered the MatchPass reconciliation workflow
149 residual items mechanically explained or resolved through the configured rules
24 items remained for human review

These results are from the Northshore Supply Co. reference build using synthetic data. The workflow demonstrates the MatchPass method; it is not a client case study or a production result. Production rules and outcomes depend on the client workflow.

Reference build

The standard stack already ran. This is the work that still remains.

Northshore Supply Co. is a fictional company and the data is synthetic. It shows how a recurring month-end reconciliation gets turned into documented rules, and where accounting judgment stays with the accounting team.

Northshore Supply Co. · March close Reference build · synthetic data

Existing stack

Standard commerce activity is already posted by the existing connector, and most normal activity closes cleanly. None of that is in question here.

Remaining problem

At each close the accounting team still pulls several sources together and works through substantially the same investigation in a spreadsheet.

Commerce activity→ Existing connector→ Accounting system→ Standard activity handled→ Residual work remains→ MatchPass starts here

4,812Monthly evidence records

4,639Baseline validated · 96.40%

173Residual items investigated

149Mechanically explained by rules

24Need human review

Synthetic figures for this reference workflow only. Baseline validated means the existing accounting output agreed with its supporting source evidence, not that MatchPass posted it.

Why this was a candidate for automation

Redundant spreadsheet work The same exports were repeatedly combined into one reconciliation workbook.

Repetitive cadence The investigation ran again each close using substantially the same source types.

The same questions

Does this payout already exist in the books?
Which settlement does this belong to?
Is this difference just timing?
Which invoices make up this deposit?
Is this adjustment already represented?

Repeatable rules Most of the work relied on identifiers, amounts, settlement relationships and approved rollups.

Judgment was the minority Only a smaller residual population genuinely needed an accountant to decide.

Northshore Supply Co. · March close · synthetic reference workflow

Selected rows shown from the larger synthetic monthly evidence set. These are the records the team would otherwise pull together to investigate what still does not tie.

Commerce sourceShopify

Underlying store activity, before anything reaches the books.

Payout date	Payout	Order	Gross	Net
2026-03-14	po_9H2KD4	#1042	1,284.50	1,246.95
2026-03-14	po_9H2KD4	#1043	96.20	93.11
2026-03-14	po_9H2KD4	#1044	2,410.00	2,339.81
2026-03-30	po_9H7RM1	#1188	1,940.00	1,883.74
2026-03-31	po_9H7RM1	#1191	1,180.00	1,145.87
2026-03-31	po_9H7RM1	#1194	880.00	851.41

Existing booksConnector output

What the normal accounting workflow already produced.

Posted	Reference	Type	Amount
14-MAR-26	po_9H2KD4	Payout posting	3,679.87
18-MAR-26	po_9H2KD4	Payout posting	3,679.87
14-MAR-26	ST-K8Q2	Processor clearing	2,416.94
31-MAR-26	po_9H7RM1	Payout posting	3,881.02

Processor activityPayPal · Stripe

Settlement and balance activity held outside the store.

Created	Id	Net	Payout
2026-03-14	ch_3Nx8Ta2eZ	551.04	po_1OqPzT
2026-03-14	ch_3Nx9Cb7rK	1,006.75	po_1OqPzT
2026-03-14	ch_3NxAqD5mW	859.15	po_1OqPzT
2026-03-22	ADJ-7Q2M	118.60	adjustment

Date	Type	Net
03/14/2026	Checkout	309.21
03/14/2026	Checkout	144.38
03/17/2026	Chargeback	-212.40
03/17/2026	Dispute fee	-20.00

Wholesale / ERPOff-platform AR

Invoices and remittances that never touch the storefront.

Invoice	Customer	Due	Amount
W-2210	Harbour Outfitters	12-MAR-26	1,240.00
W-2211	Harbour Outfitters	12-MAR-26	860.50
W-2214	Harbour Outfitters	19-MAR-26	412.75
W-2219	Cedar & Co.	26-MAR-26	688.00
W-2222	Northline Retail	27-MAR-26	96.20

BankCash settlement

What actually arrived, and when.

Date	Description	Amount
14-MAR-26	ACH CREDIT SHOPIFY	3,679.87
14-MAR-26	STRIPE TRANSFER ST-K8Q2	2,416.94
16-MAR-26	PAYPAL INST XFER	453.59
20-MAR-26	EFT HARBOUR OUTFITTERS	2,513.25
28-MAR-26	ACH CREDIT	96.20
02-APR-26	ACH CREDIT SHOPIFY	3,881.02

AdjustmentsBooks sub-view

An accounting-side view of the existing books, not a separate source dataset.

Date	Type	Reference	Amount
22-MAR-26	Correction	JE-0441	118.60

Why the existing books are ingested too. MatchPass validates the existing accounting output against the supporting source evidence before applying the workflow's custom reconciliation rules. That is what makes the baseline trustworthy, and it is what lets a duplicate posting be identified rather than assumed. This is the architecture for this synthetic workflow, not a requirement for every implementation.

These are high-level rule families configured for the synthetic Northshore workflow. Production rules are tailored to each approved client process. Open any rule to see a business-level example of what it settled.

Residual items in

173

Explained by rules

Left for review

—

Run time

—

Idle · press run to start 0.00%

Northshore Supply Co. · March close Synthetic reference workflow

4,812

Monthly evidence

4,639

Baseline validated

173

Residual investigated

149

Explained by rules

Need human review

The accounting team starts with 24 evidence-backed review items instead of rebuilding a 173-item investigation in a spreadsheet.

Representative items shown from the full synthetic residual population.

MatchPass automates repeatable reconciliation logic. The accounting team retains judgment, classification and approval. Nothing here is posted to an accounting system automatically, and no rule chooses an accounting treatment on its own.

Before MatchPass

Existing connector already running and handling standard activity
The same exports pulled again each close
The same spreadsheet rebuilt from scratch
The same reconciliation questions asked again
173 residual items investigated by hand

After MatchPass

Full evidence population validated against the books
Recurring rules applied consistently every close
149 residual items mechanically explained
24 items routed for actual accounting review
Each result carries its supporting evidence and reason

Northshore Supply Co. is a fictional example. The workflow demonstrates how MatchPass can structure a recurring reconciliation process. Production rules and results vary by client, and every figure shown here comes from a synthetic dataset we generated ourselves. It is not a client case study or a production result. Designed to complement standard e-commerce accounting connectors rather than replace them.

E-commerce reconciliation reference build

What the engine does, on data we control

The standard stack already ran. This is the work that still remains.

Have a reconciliation like this one?