The Co-op (Australia)

Magento eCommerce System

SSTech helped Client – Australia’s largest Member owned retailer – build the largest online bookstore in Australia with a sound infrastructure and many sophisticated features.

Phase 1: Magento Migration

The legacy system was built on Java with very slow performance, hard to update content, and high maintenance cost. We were asked to migrate everything including existing customers, orders, invoices and integrate with the ERP system.

Solutions

Magento, AWS

  • Develop a new eCommerce website based on Magento
  • Carefully implement the adapter to integrate with existing website to pull in all customers, orders, invoices, products
  • Implement a new adapter to integrate with Pronto (the ERP system)
  • Deploy the new ecommerce website to Amazon using EC2, ELB, RDS, Redis Cache, Varnish Cache, CloudFront

Achievements

  • Completely new cloud infrastructure on AWS
  • CCU: 5000
  • Checkout is a lot faster (maximum 5s)
  • Products / home pages are very fast (Varnish Cache)
  • Admin can manage content (CMS pages) without dev intervention
  • Product information / price are updated automatically, synced with the ERP

Services

  • Backend development (Magento)
  • Frontend development (Magento)
  • AWS implementation

Phase 2: Magento Customization

The large catalog (2M products) makes category filtering / search pages very slow compared to home / product details page. Due to the catalog size, Varnish cache is not efficient enough.

Solutions

AngularJS, ElasticSearch

  • Re-build the frontend in AngularJS to speed up
  • Build search engine with Elastic Search to improve search / filtering experience

Achievements

  • Product details, category, search page were replaced with AngularJS
  • Page load is significantly faster (less than 0.5s)
  • Filter / search is significantly faster (less than 1s)

Services

  • Backend development (Magento, ElasticSearch)
  • Frontend development (Angular)
  • AWS implementation (ElasticSearch)

Phase 3: PIM system

Pronto (ERP system) does not really excel at editing product information. The large catalog (2M records) makes updates from Pronto (ERP system) to Magento very slow and heavy.

Solutions

Magento, Akeneo

  • Build a PIM system based on Akeneo PIM
  • Analyze all Pronto updates, and send only necessary updates to Magento periodically
  • Allow purchasing team to edit all products information (description, images)
  • Allow stock per store instead of global stock

Achievements

  • Reduce Magento catalog from 2 million products to only 500,000 active products
  • Overall performance improve by 50%
  • Reduce the number of servers by 50%
  • Product information are significantly better with more images, detailed descriptions
  • Allow customer to find out which store has stock of a specific product

Services

  • Backend Developer (Magento, Akeneo PIM)
  • AWS implementation