Lead Gen Funnel | Paid Media & Growth Diagnostics
Category
Case Study
/
/

The Problem
My agency experience was concentrated in campaign setup, not the optimization and decision-making layer that sits above it. To close that gap, I built and launched my own funnel end to end, on my own budget, so I would be the one making the strategic calls, not just executing someone else’s brief.
What I Built
A lead magnet ("Paid Media Diagnostic Framework") and a MailerLite landing funnel
Meta Pixel and GA4 tracking, wired through Google Tag Manager, including Advanced Matching and a Conversion Linker tag
A Meta Leads campaign: single ad set, dynamic creative testing across a voiceover video and multiple static hooks, "Highest Volume" bid strategy (the only option available on a new ad account)


The Bug I Found and Fixed
My real conversion event only fired on one of two entry paths into checkout. MailerLite’s embedded checkout box lived in an iframe that blocked standard click triggers, so I’d already built a parent-domain Click Text trigger to catch it. But a second button on my bio page linked out to a full MailerLite redirect URL — a completely different technical path the original trigger never covered. I diagnosed the gap by comparing what fired in Meta’s Test Events against what each button actually did, then added a second, URL-based trigger and attached it to the same Lead event tag — and applied the identical fix to my GA4 conversion, which had the same blind spot. Caught and corrected before any ad spend, not after.
The Budget Decision — and the Math Behind It
Before spending, I ran the numbers on whether my available budget could actually produce a usable signal.
Meta’s ad sets typically need ~50 conversion events within a 7-day window to exit the learning phase and deliver reliable, algorithmically-optimized results.
My total available budget was 500 AED. At an estimated $3–5 cost-per-lead (a reasonable range for a Dubai B2B lead magnet), that’s roughly 27–45 leads total across the full spend — short of the 50-event threshold even before splitting it across multiple ad variants.
At a 14-day runway (35–40 AED/day to stay inside 500 AED), the ad set would likely still be mid-learning-phase when the budget ran out, meaning any cost-per-lead or winning-ad reading I took from it would be built on statistically thin, noisy data.
Conclusion: spending the 500 AED would demonstrate that I can launch a campaign — which the build above already proves — but would not produce a dataset large enough to demonstrate the analytical judgment I was actually trying to build (interpreting results, recommending what to scale or kill).
I redirected that budget toward practicing the analysis and recommendation layer on existing, adequately-powered datasets instead, where the constraint is my judgment, not my sample size.

What I’d Do at Scale
Run the campaign to full learning-phase exit with a budget sized to the ~50-conversion/week threshold, rather than a fixed calendar window
Add Conversions API alongside the Pixel to recover signal lost to browser-side tracking limits
Layer in TikTok and Snapchat, platforms I’ve run campaign setup on previously, to compare cross-platform cost-per-lead
Build a Looker Studio dashboard on top of the GA4 + Meta data for stakeholder-level reporting, not just raw metrics

