When a business changes its operating systems, reporting usually takes the hit. Either you migrate years of history into the new systems and pay for it, or you leave it behind and lose the comparison.
It doesn’t have to work that way.
JING Tea is a premium London tea brand selling through direct e-commerce, wholesale to hotels and restaurants, and international retail. We’ve run their management reporting for three-plus years. From day one the layer unified several disconnected sources — Sage accounting, Sage inventory, manual inventory feeds, hand-built budgets — into one consolidated view. The multi-source reporting problem, solved once.
Then last year they replaced their entire operating stack — Sage out, Shopify + Xero + CIN7 in. The same layer absorbed the stack change. No history migrated. No reporting blackout. The management view stayed the same through the switch.
Because the reporting layer isn’t built on the source systems. Onetribe’s Governed Data Layer sits above them. Sage history stays on Sage. New data flows in from Shopify, Xero, and CIN7. The layer pulls both eras into one model — same chart of accounts, same channels, same drill-down from P&L to the individual sales order. The stack changed. The management view didn’t.
The same layer is picking up marketing data next — from first marketing touchpoint through to booked revenue and gross margin . With Claude AI connected, the team can ask questions in plain English against numbers that already reconcile across systems.
One layer. The same design that unified their sources keeps them stable when the systems change.
Context
- Premium single-garden tea brand founded in London in 2004
- Multi-channel operations: direct e-commerce, wholesale trade to hotels and restaurants, international retail
- Sources single-garden teas from China, India, Sri Lanka, Japan, and Taiwan
- Onetribe engaged 3+ years ago to build full financial and operational management reporting across multiple data sources
- Initial stack: Sage for accounting and inventory, with additional manual inventory inputs and manually assembled budgets
- Migrated last year to a new tech stack: Shopify (sales), Xero (accounting), CIN7 (inventory, warehousing, operations)
- Reporting layer currently being extended with a marketing data feed
- Data model is AI-ready; Claude AI integration live for ad-hoc analysis
Challenge
- Reporting spread across multiple disconnected sources — Sage accounting, Sage inventory, manual inventory inputs, manual budgets — with no unified layer to tie them together
- Cross-cut management questions required manual consolidation before an answer was possible
- Decision to migrate the operating stack to Shopify + Xero + CIN7 risked breaking reporting continuity
- Standard migration route would have required rebuilding three-plus years of historical data in the new systems
- Finance and operations needed to continue reconciling performance across old and new during and after the transition
- Forward roadmap required adding marketing data and AI-assisted analysis without further system sprawl
What Was Implemented
- Direct integrations with all source systems, old and new: Sage (accounting + inventory), manual inventory inputs, manual budgets, then Shopify, Xero, and CIN7
- Proprietary Governed Data Layer as the single reporting substrate unifying legacy and modern systems
- Standardised chart of accounts, product and SKU master, customer hierarchies, and channel taxonomy across both stacks
- Automated daily data extraction replacing manual inventory and budget consolidation
- Plan versus actuals reconciled in one model — variance analysis from group P&L down to individual sales order
- Multi-channel view combining direct e-commerce, wholesale trade, and retail in a single consolidated layer
- Phased migration support: old and new systems ran in parallel inside the reporting layer during cutover
- Historical data preserved on source systems — no historical data migration required
- Marketing data layer currently being added — extending the model from marketing touchpoint through to revenue and margin
- Claude AI integrated for ad-hoc natural-language analysis against governed, reconciled data
Outcomes
- Full management reporting maintained across three-plus years of operations and a complete tech stack change
- Historical data kept intact on source systems — no historical data migration required
- Sage-era history and Shopify/Xero/CIN7-era current performance queryable side by side in one model
- Plan variance analysis and drill-down from financial headline to sales order available across both old and new data
- Multi-channel performance (e-commerce, wholesale, retail) unified into one consolidated view
- Reporting layer absorbed the tech stack change without additional rebuild cost or added complexity
- Marketing layer being added on the same substrate — future end-to-end view from marketing spend to financial impact
- AI-ready data model supporting Claude AI integration for natural-language analysis
- Infrastructure proven as a stable analytical layer across underlying system changes — Onetribe positioned as the standard multi-system reporting layer for the business