Caso de Sucesso

How a leading retailer stays fashion forward and mainframe powered 

mainframe powered floating globe in a server room

Founded more than 150 years ago, this leading retailer has evolved from a traditional bricks-and-mortar retailer into a global fashion powerhouse. Today, the company has more than 800 stores, 100 international websites, and an 11-million-square-foot warehouse network capable of shipping 600,000 units per day. 

ícone do pico da montanha

Desafios

  • Supporting 800+ stores with real-time inventory and order processing 
  • Scaling operations to handle 600,000 shipments per day  
  • Maintaining high system availability while managing billions of database calls per week  
  • Ensuring uptime during major upgrades and avoiding costly outages  
  • Addressing workforce retirement and knowledge retention 
ícone de lista de verificação

Resultados

  • Achieved up to 96% offload from the GP central processor  
  • Improved batch job performance from 3,000 to 9,000 updates per second 
  • Recovered an estimated £380,000 per week in lost revenue through improved uptime 
  • Lowered cost per MIP from GBP 0.19 to GBP 0.05, while quadrupling usage 
  • Seamless upgrades without website downtime 
  • Positioned AI as a strategic enabler for development, knowledge retention, and productivity 
ícone de lâmpada

Solução

  • Adabas & Natural on the IBM Z16 mainframe 
  • Adabas Cluster Services 
  • Adabas Fastpath 
  • Adabas for zIIP 

“The mainframe isn’t just part of our business—it is our business. Without it, we couldn’t deliver the scale, speed, or reliability our customers expect.”  

Head of Technology Services and Operations, Leading retailer 

From high street to high tech 

For this leading fashion retailer, technology has always been more than a support function. It is a strategic enabler, a competitive differentiator, and increasingly, a service platform for other brands. At the heart of the platform is the IBM Z mainframe, powered by Software AG’s Adabas & Natural. Far from being a legacy system, the mainframe is a modern, scalable, and cost-effective foundation for this retail empire. 

This retail company is known for embracing digital innovation at scale. And its technology strategy is simple: build where it adds value, buy where it makes sense, and always prioritize customer experience. This philosophy has led to a hybrid model where the company not only sells its own products, but also acts as a fashion aggregator, offering warehousing, contact center, and e-commerce services to leading fashion brands. Its ability to scale, integrate, and deliver reliably across this complex ecosystem depends on the mainframe. “No mainframe means no business,” says the company’s Head of Technology Services and Operations. 

Modernizing the mainframe without disruption 

From retail stores and online ordering to warehousing and customer service, the IBM Z mainframe underpins every critical operation. It processes billions of database calls weekly, supports real-time inventory management across channels, and ensures uptime even during major upgrades. 

The Domain Architect for Mainframe Systems adds, “Our busiest database handles around nine billion calls a week. Adabas Fastpath saves us eight billion of those. That’s the kind of efficiency we rely on. But we need to constantly modernize to ensure performance.” 

Recently, the retailer implemented a Parallel Sysplex architecture with two production LPARs on separate physical processors. This setup ensures high availability and workload balancing across systems. Adabas Cluster Services was added to support the online ordering platform, enabling continuous uptime even during planned maintenance. The journey wasn’t without challenges: initial clustering introduced significant CPU overhead – up to 60 percent on some databases.  

Working closely with Software AG support, the team optimized buffer pools and cache configurations, reducing overhead to acceptable levels. “The support team was fantastic,” says the Domain Architect. “They helped us benchmark, tune, and iterate quickly.” 

Business impact: speed, scale, and savings 

The technical improvements translated directly into business benefits. The retailer can now perform major upgrades – like moving from Z14 to Z16 or z/OS 2.4 to 2.5 – without taking its websites offline. This continuity is crucial, especially when serving international partners who trade around the clock. 

During its upgrade from an IBM Z14 to an IBM Z16 processor, the retailer also implemented Adabas for zIIP. The retailer achieved up to 96 percent offload of workload from the general-purpose central processor to the IBM Z Integrated Information Processor (zIIP), improving processing performance. “Adabas loves Z16,” the Domain Architect notes. “We expected a 33 percent uplift in the amount of processing time per second, but saw closer to 40 percent. That’s a significant performance enhancement.” 

By ensuring uptime and optimizing performance, the retailer has recovered an estimated £380,000 per week in lost revenue. Batch jobs that previously ran at 3,000 updates per second now hit 9,000, thanks to strategic placement and hardware upgrades. 

The retail company’s investment in the mainframe is about value as well as performance. Since 2013, the company has driven costs down from 19 pence per MIP to just five, while quadrupling usage. “No other hyperscaler gives us this kind of value,” the Head of Technology Services and Operations says. “That’s why we continue to invest.” 

“We get discounts on our software bill if we optimize usage,” he explains, “which makes every gain a win-win. We’re efficient, and we’re rewarded for it.” 

Looking ahead: AI and the talent horizon 

The retailer is actively preparing for the future. The company is exploring how AI can enhance productivity across its technology landscape. “There’s a lot of exciting innovation happening around DevOps, microservices, and AI code assistants,” says the Head of Technology Services and Operations. “We’re eager to see how those capabilities extend to the mainframe.” 

The retail company is exploring how the Natural AI code assistant can streamline development and support talent onboarding. The assistant helps developers generate code, explain logic, and explore documentation, all within an IDE and web browser. It also compiles and validates code behind the scenes, saving time and improving consistency. 

With a significant portion of its workforce approaching retirement, the retailer sees these new capabilities as a key enabler for sustaining and evolving mainframe operations. “We’re looking at how AI can help us retain critical knowledge, accelerate development, and support our teams more effectively,” says the Head of Technology Services and Operations. The Natural AI code assistant, tightly integrated with Natural and supported by long-term memory, improves over time and will help the retailer build a scalable, sustainable development model that supports both modernization and continuity. 

A partnership that delivers 

The retailer’s long-standing relationship with Software AG has been instrumental in its modernization journey. From building its first Adabas & Natural applications in the 1980s to implementing high-performance clustering and zIIP today, Software AG has provided the tools, support, and innovation needed to keep it ahead of the competition. 

“Software AG has always been there for us,” says the Domain Architect. “Whether it’s tuning cluster services or helping us benchmark performance, they’ve been a true partner.” 

As the retailer continues to evolve, its Adabas & Natural applications and the IBM Z mainframe remain a cornerstone of its strategy: modern, reliable, and ready for the future. 

Recursos recomendados

integração de dados

Mainframe Data Integration

Aproveite os dados do seu mainframe para a melhor tomada de decisões e serviços aos clientes. A abordagem certa para uma arquitetura híbrida permite operações fluidas entre aplicações on-premises e plataformas de nuvem.

hands on a keyboard on a laptop

Mainframe Data Access

Um caminho claro para acesso não invasivo a dados do mainframe e uma visão única dos sistemas. Assim os times podem atuar de forma rápida com dados confiáveis para analytics, IA e nuvem híbrida. [...]

cyber security

Cyber Security and Compliance

The AAFMAA built an in-house anomaly detection engine that delivers real-time insights, smarter threat modeling, and seamless alerting across the org.

Conheça a história do cliente