# Desk log EAM-REPLAY-20260823

**Status:** First-party illustrative retail composite; not customer, order, accounting, campaign, benchmark, or independently audited data  
**Run date:** 2026-08-23  
**Evidence checked:** 2026-08-31  
**Operator:** InfiniSynapse Data Team  
**Attestor:** William Zhu, InfiniSynapse cofounder ([GitHub @allwefantasy](https://github.com/allwefantasy))

The operator compared two contribution definitions over an illustrative 14-day window: Monday used COGS only; Thursday added fees and last-click advertising. The displayed output contains 17,200 illustrative order lines, seven changed SKU ranks, four sign flips, and 91 lines without dated cost. Thursday's rank and SKU-level advertising allocation were held because the definition moved and no defensible join rule was available.

The [CSV](https://infinisynapse.com/blog-media/ecommerce-analytics-metrics/downloads/grain-drift-EAM-REPLAY-20260823.csv) exposes four aggregates and two held decisions. The [verification script](https://infinisynapse.com/blog-media/ecommerce-analytics-metrics/downloads/verify-EAM-REPLAY-20260823.py) checks displayed output only. Source rows, SKU identities, monetary values, SQL, cost vintages, ad-allocation records, and reconciliation totals are unavailable.

External context: [U.S. GAO Assessing Data Reliability](https://www.gao.gov/products/gao-20-283g), [UK Government AQuA Book](https://www.gov.uk/guidance/the-aqua-book), and [U.S. Census Quarterly Retail E-Commerce methodology](https://www.census.gov/retail/ecommerce/how_surveys_are_collected.html) (retrieved 2026-08-31). These publishers did not review this run.