Erskine Attachments — Ecommerce Platform & Salesforce Integration

Containerized Magento Enterprise for Erskine Attachments with real-time Salesforce CRM + SysPro ERP — full business transition from lead gen to transactional ecommerce.

Updated May 5, 2026

The Transition

Erskine Attachments — a manufacturer of skid steer and tractor attachments — was running their entire digital presence on a static lead-generation website. No ecommerce capability. No integration with their Salesforce CRM or SysPro ERP. Leads came in through a contact form, went somewhere, and got managed manually from there.

The business had outgrown that model. The platform needed to handle real customer data, support transactional sales, surface customer and order information directly into Salesforce, and operate with the reliability and maintainability expected of an enterprise-grade system.

My role was solutions architect and technical lead at Consolidated Equipment Group — architecting the full platform transition and leading a four-person integration team across both the Erskine and QuickAttach engagements simultaneously.

The Work

Platform Architecture

The centerpiece of the engagement was a fully containerized Magento Enterprise 2.x platform — designed for parity across development, staging, and production environments. Running three identical environments meant deployment risk dropped substantially and onboarding new developers didn’t require environment-specific debugging sessions.

Stack:

  • Magento Enterprise 2.x with custom theme and module development
  • Docker-based containerized microservice environment (Ubuntu / Alpine)
  • NGINX / PHP / Percona SQL
  • SOLR (search) + REDIS (caching — full page, op-code, session)
  • Service-level scalability and full monitoring

Salesforce Integration

The Salesforce integration was real-time, bidirectional, and covered the full commercial data model:

  • CRM Quotes: quote requests from the platform created and updated Quotes objects in Salesforce automatically
  • Orders: purchase orders synced to Salesforce Sales Cloud in real time, giving sales visibility without manual data entry
  • Customers: customer account creation and updates propagated to Salesforce, keeping the CRM authoritative on customer records

This was the integration that changed the business operationally. Before, a sales rep’s view of what a customer had ordered, requested, or asked about required hunting across systems. After, it was a single record in Salesforce.

ERP Integration (SysPro)

The SysPro ERP integration handled inventory, pricing, and order fulfillment data — ensuring that what Magento showed customers reflected what Erskine could actually deliver. Stock levels, pricing adjustments, and order status updates flowed between the two systems without manual reconciliation.

CI/CD and DevOps

Continuous integration via Bitbucket Pipelines — automated testing and deployment on every commit to the relevant branches. Identical environments across dev, staging, and production meant that a build that passed in staging deployed to production without surprises.

This level of DevOps discipline was uncommon for a manufacturer in this space in 2015–2018. It made the platform maintainable, auditable, and far less dependent on tribal knowledge about environment-specific quirks.

The Business Transition

The strategic outcome of this engagement was a full business model transition: from lead generation to direct ecommerce.

Before: a prospective customer found Erskine online, submitted a contact form, and waited for a rep to call back. The conversion path was manual from the first touch.

After: customers could browse the product catalog, request quotes, complete purchases, and have those actions immediately visible to the Erskine sales team in Salesforce — with ERP-backed inventory accuracy and automated order flow through SysPro.

The platform didn’t just add a cart to the website. It changed how the business captured and processed revenue.

What I’d Do Differently

The Salesforce integration scope expanded during the engagement as the team discovered additional data relationships that needed to be covered — Quote line items, customer address handling, order status visibility for reps. That scope expansion was manageable but added timeline friction. A more exhaustive Salesforce data model audit during discovery — mapping every object and relationship before the first API call — would have reduced mid-engagement surprises.

The platform architecture held up well, but in retrospect the monitoring setup could have been more proactive. Service-level monitoring was in place, but application-level telemetry (slow queries, cache hit rates, job queue depth) was added reactively rather than built in from the start.


See the companion engagement: QuickAttach Attachments — a sister company built on the same architecture, run concurrently.