Platform Engineering

Platform Consolidation at Scale: $100M+ Revenue Protected

Challenge: Following Awin Global's acquisition of ShareASale (2017), two platforms served overlapping markets with different technology stacks and business models. The challenge: consolidate platforms while maintaining $100M+ revenue stream, preserving customer relationships, and avoiding service disruptions during multi-year integration.

← Back to Case Studies

Executive Summary

Led technical strategy for consolidating two major affiliate marketing platforms following acquisition—protecting $100M+ annual revenue while migrating 250,000 publishers and 10,000 merchants. Coordinated 15 teams across multi-year program, designing critical system integrations that enabled zero customer re-work and 98% revenue protection. This represents one of the most complex platform migrations in affiliate marketing history.

Key Results:

  • 98% revenue protected through targeted feature parity
  • 260K accounts migrated with 85% acceptance rate
  • 250M monthly requests handled at 263ms latency
  • Zero customer re-integration required for tracking
  • 15 teams coordinated across two organizations

The Challenge

Acquisition Context

Background:

  • Awin Global acquired ShareASale (2017)
  • Two platforms serving overlapping markets
  • Different technology stacks and business models
  • Redundant operational costs

Executive Decision (2023):

  • Sunset ShareASale platform entirely
  • Migrate all customers to Awin global platform
  • Consolidate operations for cost efficiency
  • Timeline: Multi-year program (2023-2025)

The Scope

ShareASale Platform Scale:

  • Publishers: 250,000 affiliate partners
  • Merchants: 10,000 advertisers
  • Transactions: $1B+ annual GMV
  • Revenue: $100M+ annual
  • Tracking Requests: 250M monthly

Technical Complexity:

  • 15+ years of custom features
  • Mission-critical tracking system (99.9%+ uptime required)
  • Advanced attribution features not in Awin platform
  • Legacy integrations with 250K+ publisher sites
  • Complex payment and fee structures

Business Risk:

  • Every transaction = revenue (cannot drop tracking)
  • Customer churn would devastate revenue
  • Competitive landscape highly sensitive to service disruption
  • Reputation damage from poor migration = long-term impact

My Role

Principal Architect - Technical Strategy & Integration

Responsibilities:

  1. Design overall migration strategy and roadmap
  2. Lead critical system integrations (tracking, transactions, accounts)
  3. Coordinate cross-company technical teams (15+)
  4. Provide subject matter expertise on ShareASale systems
  5. Technical liaison between leadership and engineering
  6. Risk identification and mitigation planning

Engagement:

  • Full-time role (2023-2025)
  • Reporting to CTO/VP Engineering
  • Cross-functional collaboration with product, InfoSec, legal, finance
  • Distributed team across US and Europe

Strategic Approach

Phase 1: Discovery & Strategy (2023)

Executive Alignment:

  • Facilitated C-level workshops defining business objectives
  • Translated technical constraints into business impact
  • Established success criteria and risk thresholds
  • Secured budget and resource allocation

Technical Assessment:

  • Comprehensive feature mapping between platforms
  • Identified capability gaps requiring legal/finance consultation
  • Analyzed usage patterns and revenue attribution
  • Prioritized features by business value and technical complexity

Migration Strategy:

  • Phased approach with interim solutions
  • Zero customer re-integration requirement
  • Revenue protection as primary constraint
  • Feature parity for highest-value capabilities

Cross-Company Collaboration:

  • Initiated collaboration with 15+ teams across organizations
  • Established communication channels and decision frameworks
  • Created shared documentation and knowledge transfer processes
  • Aligned engineering practices and tooling

Phase 2: Critical System Integration (2024-2025)

Tracking System Integration

The Challenge

Mission-Critical System:

  • 250M monthly requests
  • 99.9%+ uptime required (every transaction = revenue)
  • 250K+ publisher integrations (JavaScript tracking pixels)
  • Cannot require customer re-integration
  • Must maintain backward compatibility

Awin Platform Gap:

  • Different tracking parameter structure
  • Different URL patterns
  • Different cookie handling
  • Different attribution logic

Business Constraint:

  • Zero tolerance for transaction loss
  • No customer re-work acceptable
  • Must appear unchanged externally
  • Seamless transition required

The Solution: Multi-Layer Reverse Proxy

Architecture Layers:

Layer 1: Cloudflare (Edge)

  • WAF rules for traffic filtering
  • Rate limiting and DDoS protection
  • Initial routing decisions
  • Custom Workers for intelligent request handling

Layer 2: Akamai (CDN)

  • Global traffic distribution
  • Caching for static content
  • Secondary failover routing
  • Geographic load balancing

Layer 3: AWS (Application)

  • Lambda functions for request transformation
  • Parameter translation ShareASale ↔ Awin
  • Retry and failover logic
  • Integration with Awin backend systems

Custom Cloudflare Workers Solution:

[Publisher Site] → [Cloudflare WAF]
                          ↓
                   [Routing Logic]
                          ↓
           ┌──────────────┴──────────────┐
           ↓                             ↓
   [Awin Platform]              [ShareASale Legacy]
       (Primary)                     (Failover)

Implementation:

  • JavaScript/Node.js custom routing logic
  • Cloudflare KV for configuration storage
  • D1 for request logging and analytics
  • Wrangler for safe CI/CD deployment

Observability:

  • DataDog tracing across all layers
  • CloudWatch metrics and alarms
  • OpsGenie integration for incident escalation
  • Custom dashboards for real-time visibility
  • Synthetic monitoring for health checks

Results: Tracking System

Performance:

  • 250M monthly requests handled seamlessly
  • 263ms average latency maintained
  • 99.5% revenue protection achieved
  • 75% error rate reduction through improved fault tolerance

Business Impact:

  • Zero customer re-integration required
  • No tracking disruption during migration
  • Maintained publisher trust and relationships
  • Protected $100M+ annual revenue stream

Technical Achievement: Built production system bridging two incompatible platforms while maintaining sub-second latency and near-perfect reliability at massive scale.

Transaction Processing Integration

The Challenge

ShareASale Unique Features:

  • Advanced multi-touch attribution
  • Complex commission rules
  • Flexible transaction modifiers
  • Custom data fields
  • Real-time transaction relay

Business Requirement:

  • Maintain attribution capabilities for power users
  • Synchronize transactions between platforms during migration
  • Support accounts operating on both platforms
  • Zero transaction loss during cutover

The Solution: Transaction Relay Microservice

Architecture:

[ShareASale Transaction Processor]
             ↓
    [Translation Layer]
             ↓
    [Relay Microservice] ←→ [Awin Transaction System]
             ↓
       [Validation & Monitoring]

Implementation:

  • C#, SQL Server for core processing
  • AWS (EKS, SQS) for microservice infrastructure
  • Rust for high-performance translation layer
  • DataDog for comprehensive observability

Feature Parity Work:

  • Analyzed 20+ attribution features
  • Prioritized by usage frequency and revenue impact
  • Implemented top 95% by revenue
  • Documented gaps with workarounds

Data Synchronization:

  • Real-time transaction relay
  • Bidirectional sync for hybrid accounts
  • Conflict resolution logic
  • Audit trail for reconciliation

Results: Transaction Processing

Scale:

  • 25K daily transactions processed
  • 95% feature parity by revenue
  • Sub-second processing latency
  • Zero transaction loss during migration

Business Impact:

  • Enabled seamless platform switch for customers
  • Maintained advanced attribution for power users
  • Protected merchant relationships
  • Smooth transition without disruption

Account Migration System

The Challenge

Scope:

  • 260K accounts (250K publishers + 10K merchants)
  • Complex relationship data
  • Account balances and payment history
  • Custom settings and preferences
  • Legal: New terms of service acceptance required

Technical Complexity:

  • Different account structures between platforms
  • Different permission models
  • Different data normalization
  • Data validation and integrity checks

The Solution: Collaborative Migration System

My Contribution:

  • Provided ShareASale system expertise and documentation
  • Co-designed schema mapping with Awin engineering team
  • Conducted comprehensive data analysis
  • Validated migration workflows
  • Participated in QA and defect triage

Technology Stack:

  • Java, Spring (Awin platform)
  • AWS DocumentDB, SQL Server
  • DataBricks for data analysis
  • Postman, Curl, PowerShell for testing

Migration Workflow:

  1. Account data extraction from ShareASale
  2. Data transformation and validation
  3. Import into Awin platform
  4. Relationship mapping and preservation
  5. User notification and acceptance flow
  6. Verification and reconciliation

Results: Account Migration

Scale:

  • 260K accounts migrated successfully
  • 85% acceptance rate on first login
  • Account balances preserved accurately
  • Relationships maintained intact

Business Impact:

  • Met executive timeline expectations
  • Maintained customer trust through smooth process
  • Protected business continuity
  • Minimal customer support burden

Product Datafeed Migration

The Challenge

ShareASale Datafeed System:

  • Self-hosted FTP for merchant uploads
  • ETL pipeline processing product catalogs
  • Thousands of merchant catalogs
  • Millions of products synchronized

Awin Gap:

  • Could pull from merchant FTP endpoints
  • No direct upload capability
  • Different processing workflow
  • New system 6+ months away

Business Risk:

  • Merchants depend on datafeed uploads daily
  • Disruption = customer churn
  • Revenue at risk ($X million from datafeed-dependent features)

The Solution: Interim SFTP Sync

Stopgap Design:

  • Used ShareASale FTP as source
  • Awin systems pull from ShareASale
  • Scheduled synchronization (PowerShell automation)
  • Minimal engineering effort (“vibe-coded”)

My Role:

  • Identified viable interim solution
  • Conducted log and data analysis for prioritization
  • Fixed flaws in archiving system
  • Designed secure workflow
  • Coordinated with data teams

Results: Datafeed Migration

Impact:

  • 62% of top merchants protected (powerrank top 100)
  • Seamless automated sync enabled
  • Merchants continued operations without re-integration
  • Bought time for permanent solution
  • Revenue preserved during transition

Lessons:

  • Sometimes “good enough” is better than “perfect but late”
  • Creative re-purposing of existing systems
  • Data analysis drives prioritization

Overall Results

Business Outcomes

Revenue Protection:

  • 98% revenue protected through targeted feature parity
  • $100M+ annual revenue stream maintained
  • Minimal customer churn (<5% attributable to migration)
  • Competitive position preserved

Customer Experience:

  • Zero re-integration required for 250K+ publishers
  • 85% acceptance rate on first login (target: 90%)
  • Minimal support ticket volume
  • Maintained platform trust

Operational Efficiency:

  • Consolidated from 2 platforms to 1
  • Reduced operational overhead
  • Simplified compliance and security
  • Unified product roadmap

Technical Achievements

Scale:

  • 250M monthly tracking requests at 263ms latency
  • 25K daily transactions processed
  • 260K accounts migrated with data integrity
  • Multi-year program delivered on executive timeline

Reliability:

  • 99.5% revenue protection during migration
  • 75% error rate reduction through proxy architecture
  • Zero critical outages attributable to migration
  • Maintained SLAs throughout transition

Coordination:

  • 15 teams aligned across two organizations
  • Clear communication channels established
  • Knowledge transfer completed
  • Distributed execution successful

Lessons Learned

What Worked Well

Interim Solutions:

  • Proxy architecture bought time for permanent solutions
  • Pragmatic stopgaps (datafeed sync) prevented disruption
  • “Good enough” > “perfect but late”

Cross-Functional Collaboration:

  • Early stakeholder alignment prevented rework
  • Regular communication maintained momentum
  • Shared documentation reduced friction
  • Clear decision framework enabled progress

Phased Approach:

  • Risk mitigation through incremental rollout
  • Canary deployments caught issues early
  • Rollback capability provided safety net

What I’d Do Differently

Earlier Customer Communication:

  • More proactive outreach before migration
  • Better explanation of benefits to users
  • More training resources available

More Aggressive Testing:

  • Load testing earlier in process
  • More end-to-end scenarios
  • Chaos engineering for failure modes

Documentation:

  • Start runbooks earlier
  • More comprehensive knowledge transfer
  • Better onboarding materials for operations

What This Demonstrates

For Enterprise Architecture Roles:

  • Multi-year program strategy and execution
  • Cross-company coordination at scale
  • Complex system integration
  • Risk management and mitigation

For Technical Leadership Roles:

  • Strategic thinking balanced with hands-on architecture
  • Team coordination across 15+ groups
  • Stakeholder management (C-level to engineers)
  • Delivery under business-critical constraints

For Platform Engineering Roles:

  • Distributed systems design (proxy architecture)
  • High-availability systems (99.5% revenue protection)
  • Performance optimization (263ms at 250M requests/month)
  • Observability and monitoring at scale

For MarTech Roles:

  • Deep affiliate marketing platform expertise
  • Tracking and attribution systems
  • Transaction processing at scale
  • Product and business model understanding

Technologies Used

Infrastructure:

  • Cloudflare (Workers, WAF, CDN)
  • Akamai (CDN, global distribution)
  • AWS (Lambda, EKS, SQS, DocumentDB)

Languages & Frameworks:

  • JavaScript/Node.js (Cloudflare Workers)
  • TypeScript (Lambda functions)
  • C# (transaction processing)
  • Rust (high-performance relay)
  • Java/Spring (Awin platform)
  • PowerShell (automation)

Databases:

  • SQL Server (ShareASale legacy)
  • PostgreSQL (Awin platform)
  • DocumentDB (NoSQL)
  • DataBricks (analytics)

Observability:

  • DataDog (tracing, metrics, logging)
  • CloudWatch (AWS monitoring)
  • OpsGenie (incident management)
  • Custom dashboards

Contact

Interested in discussing complex platform migrations, system integration at scale, or affiliate marketing technology? Let’s connect.

Get in Touch: stevenleve.com/contact
LinkedIn: linkedin.com/in/steve-leve

← Back to Case Studies