From 12 siloed systems to unified analytics
Industry: Retail · Size: 200 employees, 45 stores · Warehouse: BigQuery · Sources: 12 systems, 340 tables
A regional retail chain automated extract, load, and modeling — from two days of weekly manual reporting to warehouse dashboards the same afternoon.
| Setup | 4 hours |
| Tables ingested | 340 |
| Pipeline failures (8 months) | 0 |
Problem
Sales lived in Shopify, inventory in Postgres, finance in Google Sheets, marketing in HubSpot, support in Zendesk, and POS data in CSV exports. Leadership reports were two to three days stale. A consulting quote for a six-month warehouse project was $350k plus a full-time engineer afterward.
Solution
Skippr ran as a single process in their existing AWS environment. It connected to all 12 sources (CDC where available, APIs, S3, Sheets). Schema discovery mapped 340 tables. Data landed in BigQuery with exactly-once delivery. Metabase connected to BigQuery for the first cross-system dashboard. Later Shopify API schema changes were absorbed as additive evolution.
Outcome
- Manual Monday/Tuesday reporting dropped.
- First cross-system query in minutes after data landed.
- Zero pipeline failures in eight months of operation.
See ELT product and Quick start: BigQuery. More stories: Customers.