The challenge
A kitchen production report arrived weekly as a CSV with inconsistent padding and column drift. Someone opened it, reformatted it by hand, and pasted the result into a spreadsheet that management then argued with.
Objectives
- Consume the weekly file without manual cleaning
- Separate new customers from returning ones reliably
- Give management one dashboard rather than a rebuilt spreadsheet
- Make the numbers reproducible week to week
What we built
A data model in Zoho Analytics that ingests the weekly file on a schedule, normalises the formatting quirks in a query table, and feeds a set of dashboards built around specific decisions rather than a wall of widgets.
Technology
CSV automation · Query tables · Scheduled imports · Zoho Analytics
How it was delivered
- Scheduled import configured to handle the file's padding logic
- Query tables to derive new versus returning customer cohorts
- Blended model joining production data with order records
- Role-based dashboards for kitchen, sales and ownership
- Automated report delivery to inboxes on the reporting day
Outcomes
- The weekly reformatting job disappeared
- Cohort splits are calculated the same way every week
- Management reads one dashboard instead of comparing spreadsheets
- New reports can be added to the model without rebuilding it