I'm Abir Hossain, a performance marketing architect with six-plus years spent turning paid media into a measurable, repeatable growth engine. Since 2019 I've managed more than $10M in ad spend and built over 50,000 campaigns and ad sets across Meta, Google and TikTok.

The results I'm proudest of, though, sit beneath the campaigns. My discipline is the connective tissue between spend and revenue — the tracking, attribution, reporting and automation that let a business know exactly what its marketing is doing, and why. I don't just run ads; I build the system that makes them accountable.

• Measurement first. Creative and budget decisions are only as reliable as the data behind them, so I establish clean GA4, GTM and server-side / CAPI foundations before scaling anything.
• Accountable to profit. Impressions and clicks are inputs — I manage to the outcome: return on ad spend, cost per acquisition, revenue.
• Systems over effort. Reporting, alerting and QA belong in scripts and pipelines, not in someone's morning. I automate the repeatable so strategy gets the hours.
• Evidence over opinion. Funnels and landing pages improve through structured testing, and every claim I make is one I can show in the data.

Meta Ads, Google Ads and TikTok Ads for paid media; Google Tag Manager, GA4, Meta CAPI and server-side tracking for measurement; and Google Apps Script, Zapier, Make and n8n for automation — and agentic AI tools (Hermes, OpenClaw) for autonomous operations — plus whatever a specific problem genuinely calls for.

I'm currently open to senior performance marketing roles and a select number of consulting engagements — particularly where measurement is broken, spend is scaling faster than the reporting behind it, or manual work has become the bottleneck to growth.

Managing a large portfolio of ad accounts meant every morning started the same way — logging into each platform, pulling numbers by hand, and rebuilding the same report. Errors slipped in, cost-per-acquisition spikes were caught too late, and hours vanished before any real optimization work could begin.

I built a scheduled pipeline that unifies the whole portfolio into a single source of truth and alerts the team before anyone opens a dashboard:

• A Google Ads MCC script pulls daily performance from every account into one central sheet — sorted latest-first, deduplicated automatically.
• An Apps Script layer formats the data and sends a card-style report (spend, clicks, CTR, conversions, CPA, avg CPC, ROAS) with a grand-total block.
• A CPA-threshold alert fires the moment any account crosses its limit — delivered to both email and a Telegram channel at 10 AM daily.
• An agentic AI layer on top (Hermes + OpenClaw) reads the raw numbers every morning — analyzing anomalies, cross-account patterns and root causes that no threshold alert would catch — and pushes the interpretation, not just the data, to the team before the workday starts.

Browser privacy changes and ad-blockers were quietly eating the brand's conversion signal. The ad platforms were optimizing on incomplete data — under-reporting sales, misattributing spend, and steadily making worse decisions. The numbers in the dashboard no longer matched the numbers in the bank.

I rebuilt the measurement layer from the ground up so conversions are captured server-side, not left to the browser:

• A clean DataLayer specification for the full e-commerce funnel — view, add-to-cart, checkout, purchase — with consistent parameters.
• GTM web + server containers wired to GA4 and Meta, with enhanced conversions and the Conversions API sending events straight from the server.
• Event deduplication between pixel and CAPI so nothing is double-counted, plus QA against real orders to confirm accuracy.

A low-ticket paid masterclass needed volume — a lot of registrations at a cost that still left room for profit. A single channel couldn't deliver both the reach and the efficiency, and a broad, unsegmented push would have burned budget on people who were never going to convert.

I ran a full-funnel campaign across two platforms, with the closing push built around intent rather than reach:

• A high-volume Meta structure (CBO for scale, ABO for control) paired with Google Demand Gen to widen the top of the funnel.
• Clean event tracking via Pixel + CAPI so both platforms optimized on real registrations, not proxy clicks.
• A segmented final-push retargeting layer — website visitors, video viewers, and cart abandoners each got a different message and offer.