Live Facebook-ads Campaign Diary for Launching a Pay-per-lead Niche
Originally published Aug 22, 2026 — Billy Robinson
Billy Robinson · I Spent $635 on Facebook Ads in a Brand New Niche. Here's What Happened. ↗
A real-time, week-one walkthrough of launching a Facebook lead-generation ad campaign in a brand new niche (life insurance leads in Australia) for a pay-per-lead agency: reviewing ad-account data after $635 in spend, identifying which ad formats and hooks converted (image ads outperformed video ads; specific text hooks and images won), pausing underperformers, and using an AI ad-creative tool (Poppy) to generate new variations of the winning ads along with matching ad copy.
The presenter walks through his actual decision-making process on real data: waiting for enough days of data before making changes, comparing cost-per-lead across ad sets, and iterating both by tweaking what's already working and by testing a completely new ad style (text-overlay lifestyle images) in parallel, aiming to bring cost-per-lead down toward the low $20s.
Someone who already has, or is building (per the presenter's companion roadmap video also in this database), a pay-per-lead agency and wants a concrete look at the actual week-to-week media-buying and optimization process rather than just the high-level strategy.
This is one of the more transparent, blow-by-blow entries in this database, real ad-account numbers, real failures shown alongside wins (the video ads flopped, one image ad had too high a cost-per-lead and got paused), and the presenter is candid that a failed early test is a normal, expected part of media buying, not something to be discouraged by. The video still promotes the presenter's own paid community and ad-prompt templates, and this is a single week's data from a single niche and market, not a guaranteed repeatable result elsewhere.