In Q4 2025, we completed one of the largest Dynamics 365 Commerce rollouts in the UK: 120 stores across 3 brands, migrated from a legacy POS system to unified commerce in 90 days. No store closures. No lost transactions. Here is how we did it.
The Challenge
Our client — a multi-brand retailer with 120 locations across the UK and Ireland — was running three separate systems:
- A 15-year-old POS system that the vendor had stopped supporting
- A separate e-commerce platform with no inventory visibility
- Manual order management through spreadsheets and email
The result: customers could not buy online and pick up in store. Inventory was inaccurate across channels. Returns were a nightmare. And the IT team was spending 60% of their time keeping the legacy POS alive.
Why 90 Days?
The client had a hard deadline: their busiest trading period starts in November. We kicked off in July and had to be live across all 120 stores by early October, with a month of buffer for stabilisation.
Traditional rollouts of this scale take 6-12 months. We committed to 90 days by applying three principles:
- Standardise first, customise later — We deployed a standard Dynamics 365 Commerce template across all stores and deferred brand-specific customisations to Phase 2
- Wave-based rollout — 10 stores per wave, 2 waves per week, starting with the lowest-risk locations
- AI-assisted data migration — Our proprietary migration toolkit handled product catalogue, customer, and transaction history conversion automatically
The Architecture
The technical stack we deployed:
- Dynamics 365 Commerce as the unified commerce platform
- Commerce Scale Unit (CSU) deployed in each store for offline resilience
- Dynamics 365 Finance as the back-office ERP
- Power BI dashboards for real-time sales and inventory visibility
- Copilot for store managers (natural language inventory queries)
All stores connect to a central Commerce headquarters, but each CSU can operate independently if the internet connection drops — critical for retail environments.
Week-by-Week Breakdown
Weeks 1-2: Foundation
- Configured Commerce headquarters with the client’s product hierarchy (12,000 SKUs across 3 brands)
- Set up the standard POS layout with brand-specific theming
- Built data migration scripts for customer records (450,000 loyalty members)
- Configured pricing engine with the client’s complex promotion rules
Weeks 3-4: Pilot
- Deployed to 10 pilot stores across all 3 brands
- Trained store managers with a 2-hour hands-on session
- Ran parallel operations (old POS + new POS) for 5 days
- Identified and fixed 23 configuration issues before broader rollout
Weeks 5-10: Wave Rollout
- 2 waves per week, 10 stores per wave
- Each wave followed a repeatable playbook: hardware setup on Monday, data migration Tuesday, training Wednesday, go-live Thursday, support Friday
- A dedicated “war room” monitored every go-live in real time
- Cumulative success rate: 98.3% of stores went live on schedule
Weeks 11-12: Stabilisation
- Resolved remaining edge cases (3 stores with non-standard hardware)
- Enabled omnichannel features: buy online pick up in store (BOPIS), ship from store, endless aisle
- Deployed Power BI dashboards to regional managers
- Handed over to the client’s internal team with documentation and runbooks
The Results
After 90 days and 120 go-lives:
- 50% faster order processing — transactions that took 45 seconds now take 22 seconds
- Real-time inventory accuracy across all channels (was 72%, now 99.1%)
- 4x inventory turns through AI-driven replenishment
- 3-month ROI — the project paid for itself through reduced stock-outs and improved sell-through
- Zero unplanned downtime during the entire rollout
Lessons Learned
1. Hardware matters more than you think. We specified new receipt printers and barcode scanners for all stores. Trying to reuse legacy hardware would have added weeks of compatibility testing.
2. Store manager buy-in is everything. The pilot stores became our advocates. By wave 3, store managers were requesting to be in the next wave rather than dreading it.
3. Offline resilience is non-negotiable. Two stores lost internet during their first week on the new system. The Commerce Scale Unit handled it seamlessly — transactions synced when connectivity returned.
4. Data migration is the hidden risk. Product data was in worse shape than anyone expected. Our AI migration toolkit saved us weeks by auto-mapping and cleansing product attributes.
Could This Work for Your Retail Operation?
If you are running a legacy POS, managing multiple brands, or struggling with omnichannel — this approach scales from 10 stores to 500. The key is standardisation, wave-based rollout, and a team that has done it before.
Explore our Retail & Commerce solutions or talk to our retail team.