Northwind Traders — Retail

TL;DR

Northwind Traders: 65 VMs, 12 SQL databases, 8 .NET apps. ERP and loyalty follow Stabilize (months 1–3); e-commerce and customer analytics follow Transform (months 2–7). Fabric unifies the Customer 360 view across all channels.

Northwind Traders is a specialty food retailer with 80 stores and a fast-growing e-commerce platform. Their online business now accounts for 40% of revenue — but the technology behind it was designed for 10%.

The Challenge

Northwind’s growth is creating urgent problems:

• Seasonal scale failures — Black Friday, holiday seasons, and flash sales regularly overwhelm the e-commerce platform. Last year’s holiday sale crashed for 3 hours, costing an estimated €2M in lost revenue.
• Siloed customer data — In-store purchases, online orders, loyalty program activity, and customer service interactions are stored in separate systems. There is no unified Customer 360 view.
• Slow time-to-market — Deploying new features to the e-commerce platform takes 4-6 weeks due to manual testing and deployment processes. Competitors ship weekly.
• Limited analytics — Marketing decisions are based on monthly sales reports. There is no real-time visibility into customer behavior, inventory levels, or promotion effectiveness.

Why now?

Northwind’s CEO has announced a “digital-first” strategy. E-commerce is expected to reach 60% of revenue within 2 years. The current infrastructure cannot scale to meet this target — and every failed peak-season event erodes customer trust.

The Assessment

Azure Migrate reveals the estate:

Category	Count	Finding
Windows Server VMs	65	50 migration-ready, 10 need OS upgrade, 5 can retire
.NET applications	8	5 are .NET Framework (including e-commerce), 3 are .NET 6+
SQL Server databases	12	10 compatible with SQL MI, 2 need feature review

The Path Decision

Workload	Path	Rationale
ERP and inventory management	Stabilize	Stable, back-office, moderate change frequency
Loyalty program backend	Stabilize	Works well, just needs better infrastructure
E-commerce platform	Transform	Must scale elastically, needs rapid deployment
Customer analytics service	Transform	New build — designed cloud-native from scratch

Execution

Stabilize (Months 1-3):

• Migrate ERP and loyalty VMs to Azure
• Migrate databases to SQL Managed Instance
• Enable SQL MI Mirroring to Fabric for inventory and loyalty data

Transform (Months 2-7):

• Modernize e-commerce: .NET 8, containerize, deploy to Container Apps with autoscaling rules for traffic spikes
• Build customer analytics service as a new cloud-native app with Azure SQL Database
• CI/CD pipeline with automated testing — deploy in hours, not weeks
• Both databases mirrored to Fabric

The Payoff

%%{init: {'theme':'neutral'}}%%
graph TB
  classDef azure   fill:#0078d4,stroke:#005a9e,color:#fff
  classDef onelake fill:#742774,stroke:#5a1e5a,color:#fff
  classDef bi      fill:#fde8f9,stroke:#742774,color:#3a003a
  subgraph h1["Stabilize"]
    ERP[("ERP → SQL MI")]:::azure
    LOYALTY[("Loyalty → SQL MI")]:::azure
  end
  subgraph h2["Transform"]
    ECOM[("E-commerce → Azure SQL DB")]:::azure
    CUST[("Customer Analytics → Azure SQL DB")]:::azure
  end
  FAB(["Fabric OneLake"]):::onelake
  PBI["Power BI<br/>Customer 360 Dashboard"]:::bi
  ML["Data Science<br/>Recommendation Engine"]:::bi
  RTI["Real-Time Intelligence<br/>Inventory Alerts"]:::bi
  ERP -->|"Mirror"| FAB
  LOYALTY -->|"Mirror"| FAB
  ECOM -->|"Mirror"| FAB
  CUST -->|"Mirror"| FAB
  FAB --> PBI
  FAB --> ML
  FAB --> RTI
  style h1 fill:#e6f3ff,stroke:#0078d4
  style h2 fill:#e6f3ff,stroke:#0078d4

Business outcomes:

• Peak-season resilience — Azure Container Apps autoscaling is tested against holiday traffic scenarios with rollback and performance thresholds agreed in advance
• Customer 360 view for the first time — in-store, online, and loyalty data unified in Fabric, powering personalized marketing and a recommendation engine
• Deployment cycle reduced from 6 weeks to same-day — the e-commerce team ships features weekly, matching competitor pace
• Near-real-time inventory visibility across stores and warehouse — enabling stockout reduction targets and more accurate online availability display