Marketing Campaign and Sales Impact (EU)

Overview

A decade of marketing and sales at weekly grain, where the product is the lift mechanism itself. Twelve invented consumer-goods brands sell through ten invented European retailers; 472 campaigns buy television, digital, social, retailer media networks and print across 548 weeks; and every week's actual units are built as baseline plus seasonality plus per-channel lift plus residual. The decomposition ships as its own table and the build asserts it sums exactly, row by row, except one documented brand-quarter whose attribution window is double counted: the digital and social columns both claim an overlapping conversion window, the rows sum above actuals by the recorded excess, and finding the quarter is the exercise. Lift carries diminishing returns, asserted on the shipped rows: lift per dollar falls as spend rises. The channel mix moves the way the European market moved, asserted: print declines slowly and keeps a material long tail while retailer media networks multiply from a low base, and the 2020 commerce shift lifts the online share of units more mildly than in the American sibling and holds most of it. Retailer sales split the brand-week actual exactly; one retailer reports weeks starting Sunday while everyone else starts Monday, so a date join silently drops it, and one campaign's tracking failed in flight, spending money against zero measured impressions. Built for media-mix modelling, attribution teaching, incrementality analytics, retail analytics, ETL pipelines and BI logic, with ground truth included. The licence and disclosure sections state the provenance of the data.

Row Counts by Table

Counted from the files that ship, not estimated.

Table Relationships

One parent record and seven child tables, each joined back on the same key.

Schema

Every table ships with typed columns, referential integrity, and documentation.

Sample Data

A snapshot of real rows from the dataset (values are fully synthetic).

Version History