Turn product, analytics, and commerce data into a revenue operating system.
I build the data plumbing behind Google Merchant Center, GA4 ecommerce reporting, product feeds, and operational dashboards so marketing spend is trained on revenue truth.
For commerce teams where product-feed quality, attribution accuracy, and the back-office source of truth directly determine paid-acquisition ROI. Built around the Conti Digital open-source GA4 and Google Merchant Center plugins for Payload CMS, both running in production at a $5M+ ARR client.
Who this engagement is for
Who this is for
- High-AOV commerce teams running paid acquisition (Google Shopping, Meta, Pinterest) where product feed quality and attribution correctness directly determine ROAS.
- Brands on Payload CMS, Next.js, or a custom commerce platform where the open-source GA4 + GMC plugins fit the existing architecture.
- Teams whose marketing tools (Klaviyo, GA4, GMC, Pinterest, CallRail) currently work in isolation and disagree with each other in aggregate.
- Owners who want to own the data engineering surface and stop renting attribution truth from marketing-platform-defined dashboards.
- Catalogs of 1,000+ products where GMC category resolution, local-inventory feeds, and conversion event accuracy compound revenue.
- Companies preparing to scale paid acquisition where the data layer needs to be correct before the budget grows.
Who this is NOT for
- Buyers who want a marketing agency to manage their ads. This engagement builds the data engineering layer underneath; it does not run media.
- Teams that have not yet picked a commerce platform. The engagement assumes Payload CMS, Next.js, or a custom platform already exists.
- Buyers who want a one-off conversion-tracking install. The engagement scope is feed quality + attribution correctness + automation, not pixel placement.
Where this engagement starts
The problem
Commerce data engineering is the difference between paid acquisition that compounds and paid acquisition that quietly burns budget. Product feed quality, attribution correctness, conversion event accuracy, and the back-office source of truth all have to compose. When they don't, the symptoms are familiar: GMC product disapprovals, attribution gaps between checkout and ads, ROAS that disagrees with the bank account, marketing tools competing for the same conversion event, and data the owner can't actually trust to make budget decisions on.
Most commerce teams hire an agency to manage the symptom (the GMC feed, the GA4 dashboard, the Klaviyo segmentation) without owning the data engineering underneath. The result is a stack of marketing tools that work in isolation and disagree with each other in aggregate.
The fix is data engineering inside the commerce platform itself. Real feed sync with category resolution. Real conversion events with idempotency. Real attribution with the order document as the source of truth. Real audience segmentation tied to actual customer behavior, not marketing-tool-defined cohorts.
The approach
How this engagement runs
Build the data engineering layer commerce teams need: feed quality, attribution correctness, conversion event accuracy, and source-of-truth discipline tied to the commerce platform itself.
Data audit
Map the actual data flow: where the order document lives, where product data lives, where conversion events fire, where attribution is computed. Find the gaps, the duplicates, the silent failures.
Feed + attribution architecture
Design the GMC feed (with category resolution + local-inventory feed), the GA4 conversion event surface (with eCommerce schema + cross-domain dispatch), the Klaviyo segmentation logic, and the order-document source-of-truth model that ties them together.
Implementation
Build first-party adapters with typed contracts (Zod-validated), retry semantics, idempotency keys, and dedicated tests. Use the open-source Payload plugins (payload-plugin-ga4-ecommerce + payload-plugin-gmc-ecommerce) as the foundation where they fit.
Validation + reconciliation
Cross-check the attribution math against the actual order ledger. Audit the GMC feed against the catalog. Validate conversion events fired against orders confirmed. Find the silent failures and fix the root cause, not the dashboard.
Operate
CloudWatch alarms on feed sync errors, conversion event delivery failures, attribution drift. Monthly reporting against the budget model the owner actually uses to decide where to spend.
Outcomes
What you walk away with
- +40% user acquisition post-cutoverAnchor client paid acquisition liftSource: Fine's Gallery post-platform-cutover analytics (June 2025 baseline through February 2026)
- 5,400+ SKU catalogCatalog operating in production via GMC pluginSource: Fine's Gallery Google Merchant Center sync (open-source payload-plugin-gmc-ecommerce)
- 2 production-grade Payload plugins publishedOpen source maintenanceSource: payload-plugin-ga4-ecommerce + payload-plugin-gmc-ecommerce (npm + GitHub)
- Daily product feed + local-inventory feed exportFeed sync cadenceSource: Fine's Gallery GMC sync architecture (nightly Lambda + SQS / Lambda backbone)
- Idempotent dispatch with database-level dedupConversion event correctnessSource: Fine's Gallery GA4 conversion event pipeline
- Order document, not marketing-tool dashboardsSource-of-truth modelSource: Fine's Gallery commerce platform data model
Concrete deliverables
- Google Merchant Center integration: daily product feed sync, GMC category resolution, local-inventory feed auto-export, disapproval handling, and feed-quality monitoring.
- Google Analytics 4 integration: eCommerce schema, conversion event dispatch with idempotency, structured data (JSON-LD), cross-domain event tracking, and the open-source GA4 Payload plugin as the foundation.
- Klaviyo integration: catalog sync (nightly Lambda), customer segmentation tied to actual order behavior, conversion event dispatch.
- Pinterest integration: catalog automation, audience segmentation, pin publishing workflow, feed builder.
- CallRail integration: lead attribution tracking, form submission capture, phone-call correlation to orders.
- Source-of-truth discipline: the order document is the truth; marketing tools consume the truth, they don't define it. Reconciliation reports against actual ledger state.
- First-party adapter architecture: every integration lives behind a typed boundary with Zod validation, retry semantics, idempotency keys, and dedicated tests. No off-the-shelf plugins.
- Observability: CloudWatch alarms on feed sync errors, conversion event delivery failures, attribution drift. Monthly reporting against the budget model the owner uses to make decisions.
FAQ
Frequently asked questions
The full engagement assumes a custom commerce platform (Next.js + Payload CMS + Postgres + AWS) where the data engineering layer can be owned end to end. The open-source GA4 + GMC Payload plugins are the foundation.
If the underlying platform is Shopify Plus, the data engineering work runs against Shopify's APIs but the engagement is more constrained because Shopify owns the order document. For high-AOV commerce, the right move is usually a custom platform first, then this data engagement on top.
Two production-grade Payload plugins are published under the MIT license: payload-plugin-ga4-ecommerce (GA4 reporting + tiered caching + bounded concurrency + in-flight dedup + dual-layer rate limiting) and payload-plugin-gmc-ecommerce (Google Merchant Center sync engine).
Both run in production at the Fine's Gallery $5M+ ARR commerce platform. Both are linked from the open-source page on this site with full build-story articles.
Sometimes. The diagnostic is whether the attribution problem lives in the data layer (which can be fixed in place) or in the platform layer (which usually means the order document doesn't carry the truth marketing tools need). The Architecture Sprint engagement ($15K, 3 weeks) is the right way to find out without committing to a longer scope.
No. This engagement builds the data engineering layer underneath the campaigns. Media buying, creative, and bid management belong with a marketing agency. The data layer needs to be correct before the agency's work compounds; that's what this engagement delivers.
Implementation engagements from $25K. Final scope depends on integration count, catalog size, and the depth of the source-of-truth work.
An Architecture Sprint ($15K, 3 weeks) is the right way to scope a larger engagement: it produces a written architecture decision document with a specific implementation plan you can run with anyone, including yourself.
Two production-grade open-source Payload plugins (GA4 + GMC) running on a $5M+ ARR commerce platform. A 5,400+ SKU GMC catalog operating with daily feed sync and local-inventory export. +40% user acquisition lift post-cutover at the Fine's Gallery anchor client. Full build-story articles for both plugins on the site.
Related work
See this in production
Fine's Gallery Platform Modernization
Solo-built end-to-end commerce platform for a luxury marble and stonework gallery. $5M+ annual revenue running on a sovereign, client-owned AWS Organization. Zero-downtime production cutover. 20+ years of legacy operational data migrated into the new platform without loss. Marketing dominance on the new platform supports $500K+/month in commerce volume. Systematic removal of vendor lock-in across SaaS commerce, payment, and warehouse tooling. Now operates on a Fractional CTO retainer.
Peter T Conti Consulting Platform
The consulting practice's own site, built as the reference architecture for what the practice sells. Same Next.js + Payload + Postgres + ECS Fargate + Terraform stack as the anchor engagement. STS-only IAM, no long-lived credentials anywhere. Cost-tuned to run for a small fraction of the naive default AWS footprint, with full editorial workflow operated as code.
Read more
In-depth on this topic
Building a Google Merchant Center Sync Engine Plugin for Payload CMS
Merchant Center sync engine that powers $2-400k+/month in Google Shopping revenue for a 5,400-product luxury ecommerce catalog.
Building a Production-Grade GA4 Analytics Plugin for Payload CMS
How I built a production-grade GA4 analytics plugin for Payload CMS, and the architectural decisions behind making it reliable at scale.
Google Shopping for High-Ticket Ecommerce: The Fine's Gallery Playbook
How Fine's Gallery uses Google Shopping as a serious demand engine, not a checkbox marketing channel.
Credentials and Fit
Directly relevant experience for high-trust delivery
I lead architecture and implementation personally, with technical depth that spans product, cloud, and execution operations.
AWS Certified Solutions Architect - Associate
Cloud decisions are grounded in secure, cost-aware architecture with practical production tradeoff management.
Dual technical foundation
B.S. Computer Science plus B.S. Chemistry with undergraduate research and scientific presentation discipline.
AWS Certified Solutions Architect - Professional
My ability to use AWS to solve complex business requirements has been demonstrated at the highest enterprise level.
Data & Automation Inquiry
Share your current ad spend efficiency, catalog feed constraints, and conversion-tracking gaps so I can scope the highest-leverage automation plan.
Pressure-test the platform decision before you commit budget.
Schedule a complimentary 30-minute consultation to align on objectives, stress-test your architecture, and leave with a concrete set of recommendations. No obligation, no sales pitch. Just actionable technical guidance.