# Ecommerce attribution reconciliation worksheet

Looqus Editorial · Version 1 · 15 September 2026
Editable starter worksheet. All worked numbers are synthetic USD. This does not reproduce a vendor algorithm or estimate causal lift.

## Use this worksheet

Choose one completed reporting period. Reconcile store records before comparing channel credit. Keep incompatible scopes separate and give unresolved differences an owner. Use approved exports; remove personal customer information from shared copies.

## Business basis

- Period, timezone and extraction time:
- Included order states:
- Gross or retained revenue:
- Refund/cancellation treatment and observation horizon:
- Tax/shipping treatment:
- Currency and conversion policy:
- Product/fulfilment/payment/return costs included:
- Spend scope:
- New versus returning customer definition:

## Report settings register

| Report | Model | Eligible interactions | Window(s) | Date basis | Observed/modeled distinction | Filters and known gaps |
|---|---|---|---|---|---|---|
| Meta Ads | | | | | | |
| Google Ads | | | | | | |
| GA4 | | | | | | |
| Triple Whale / Northbeam, if used | | | | | | |

Record actual settings. Different products use “window” differently. Do not assume a universal default.

## Synthetic reconciliation

| Item | USD |
|---|---:|
| Merchandise revenue, after discount, excluding tax/shipping | 200.00 |
| Refund | −40.00 |
| Retained sales | 160.00 |
| Product cost | −70.00 |
| Variable fulfilment/payment/return costs | −20.00 |
| Advertising spend | −60.00 |
| Contribution after advertising, before fixed overhead | 10.00 |

Social and search might each claim the same $160 under separate configured attribution rules. Adding the claims yields $320, not another sale. Keep the business ledger at $160 retained sales. This possible overlap is different from sending a duplicate purchase event to one destination.

## Discrepancy register

| Difference | Amount/scope | Likely cause to test | Evidence required | Owner | Resolution or stated limitation |
|---|---|---|---|---|---|
| | | | | | |

Investigate order population, revenue definition, duplicates, refunds, identity, timing, model and data maturity. Do not assign the unexplained remainder to a favorite channel.

## Spending decision record

- Decision that depends on this comparison:
- Reconciled business evidence:
- Remaining uncertainty:
- Whether an experiment or qualified MMM analysis is needed:
- Contribution threshold and cash constraint:
- Approved action and limit:
- Responsible operator:
- Verification evidence and review date:

Attribution credit does not establish additional demand. Improving matching can increase reported conversions without increasing store sales.