Email Campaign & Attribution Engine

The problem

Email platforms report opens and clicks, and a large share of both come from security scanners and mail privacy features, not from people. Campaign results lived in one tool, website visits in another, and revenue in the CRM, so nobody could say which message produced which conversation.

What I built

• Event ingestion. Every send, open, click, bounce, and unsubscribe flows into the database in near real time.
• Verified clicks. Scanner and bot clicks are filtered out, so click rates describe people.
• Campaign families and A/B tests, measured by step, by industry, and by segment.
• Sender health. Bounce and complaint rates per sender, with a hard stop when a campaign crosses the line.
• A vanity-link registry that keeps short links on the website in sync with their tracking tags.
• Content performance. Reach, reading depth, CTA clicks, and assisted conversions for every article and page.
• Audits and alerts. A scheduled audit reviews live campaigns and posts its findings to Discord.

How it works

Everything lands in Postgres next to the CRM, so a click can be traced to a website visit, a booked call, and a deal. The metrics views exclude internal and test traffic, so the dashboards stay honest.

Stack

Brevo, Supabase (Postgres views and scheduled jobs), Cloudflare Pages Functions, Discord webhooks.

Where it stands

Live, with per-page content reporting added in September 2026.