Unifying four sales channels into one monthly report
A Hong Kong design brand replaced days of manual spreadsheet work with a clean, auditable monthly report across online, in-store, distributors and wholesale.
The brand designs and sells its own products across four very different channels: an online store with a third-party warehouse, its own in-store retail, distributors who mostly take stock on consignment, and wholesale buyers. Every month, someone rebuilt the numbers by hand, retailer by retailer.
The reports did not line up. Each retailer used a different format, some had no barcodes at all, and consignment and buyout revenue followed different rules. Totals rarely matched, and the work left no time for the analysis that actually drives buying and store decisions.
Four channels, four formats
Each retailer's monthly export looked different, so consolidation was manual every time.
No reliable way to match products
Some retailers had no barcodes and used their own SKUs, so the same product appeared under many names.
Consignment and buyout differ
Revenue rules changed by channel, so a single formula gave the wrong number.
Days of manual work, fragile totals
Reporting took days and small mistakes broke the totals that leadership relied on.
We built a reporting engine that takes each retailer's raw file, cleans it into a consistent sheet, and produces category, store and channel summaries plus one management report. A separate safety sheet shows every assumption and edge case, so the numbers are checkable rather than trusted blindly.
A product master and retailer crosswalk match each retailer SKU to a canonical product. Low-confidence matches go to a review sheet for a person to confirm; nothing is silently auto-filled, and consignment rules stay configurable per channel.
- Adapters for the main retailer report formats
- Product master plus a reviewed retailer crosswalk
- Category, store and channel summaries
- A safety sheet that surfaces assumptions and edge cases
- Configurable consignment and buyout revenue rules
Weeks, not months, per retailer added
One expensive problem, proven before it scales. The A1 to POC method, run in weeks not quarters.
The win was reclaimed time and numbers people could trust.
- Reporting that took days now takes minutes, every month.
- Reclaimed hours go into store-level analysis that drives buying decisions.
- A safety sheet replaces blind trust, so errors are caught before leadership sees them.
- One product master ends the monthly argument about which number is right.
We design and deploy to the ISO/IEC 27001 (information security) and ISO/IEC 42001 (AI management system) frameworks. Data stays where it should, decisions that carry real cost keep a human in the loop, and every model call is logged for audit.
Designed and deployed to these frameworks. Not a certification claim.
Have a problem like this?
Bring your messiest operational problem. In 30 minutes we will tell you where AI would pay back fastest, and where it would not.
Discuss an operations problem