Case Study RetailCloud – Multi-Region Google Cloud Auto-Scaling for Black Friday (UK)

Project Overview

RetailCloud is the cloud modernization initiative for an omnichannel high-street fashion brand in Manchester, UK, preparing their e-commerce storefront for Black Friday and holiday peak traffic spikes.

Client & Location

HighStreet Fashion Group plc
Location: Manchester, UK
Industry: Fashion & Retail E-commerce

Challenges

• Past holiday shopping seasons resulted in catastrophic server crashes and estimated revenue losses over £800k.
• Traffic bursts surging from 2,000 to over 120,000 concurrent online shoppers within minutes.
• Monolithic database bottlenecks causing slow checkout processing times.

Solution

SM Softwares designed a resilient Google Cloud GKE architecture with horizontal pod autoscaling (HPA), Cloud Spanner database read-replicas, and Cloudflare Enterprise edge caching for static media.

Technologies Used

• Google Cloud Platform (GKE & Cloud Spanner)
• Cloudflare Enterprise CDN & DDoS Shield
• Terraform, Helm Charts & ArgoCD
• K6 Load Testing (Simulated 200,000 concurrent users)

Results & Business Impact

• 100% uptime maintained during the 96-hour Black Friday through Cyber Monday peak.
• Processed over 18,500 orders per minute with average page load under 780 milliseconds.
• Auto-scaling scaled down infrastructure post-event, eliminating wasteful idle server costs.