Layer 1 -- Scraping
Product universe + movement snapshots
Two scraped datasets covering the entire market at brand x product level: the master data file (1,631 SKU x price rows across 18 retailers) establishes who sells what at what price, and the weekly tracking xlsx (509 SKUs x 5 stock-count snapshots over 17 days) captures how fast products move off the shelf. Stock decrease between snapshots = implied units sold.
Layer 2 -- Field research
Geographical velocity + brand positioning
The scraping covers a short window and a limited store set. Field research expands both: the X-cite 21-branch survey records typical-month unit sales per category with top-brand attribution across Mega/Super/Express store tiers, and 40+ store visits by the field team capture verbal monthly velocity, Hasawi positioning, and channel dynamics at sanitary stores, electronics shops, and AC specialists.
Layer 3 -- Seasonality & projection
Expand to full year x national market
Each product line carries a 12-month seasonality curve (refrigerators flat, water coolers peak summer, water heaters peak winter) that corrects for where the March observation sits in the annual cycle. Field-observed per-store-tier velocities are then multiplied across the 156-store census (4 channels: X-cite 44 branches, Other Dept Store 30, Electronics 26, Construction/Sanitary 56) to produce a national volume estimate.