
John Rambrandt
I’m a computer engineer who spends my time building for the Web and contributing to open source, but I’ve always been obsessed with how things break. I treat cybersecurity as a puzzle that involves both code and people, blending technical defense with a deep dive into the psychology of social engineering. My goal is simple: build tools that are open, scalable, and actually resilient against real-world threats.
Written by John Rambrandt (39 articles)

Meta Ads records being blocked as an onsite conversion. 1,125 blocks across 17 accounts
Meta Ads files being blocked as an onsite conversion. 1,125 blocks across 17 of 23 accounts, 32% of first replies pooled, three accounts at one block per reply.

Google Ads Target CPA held spend flat 1pm–7pm while CPL ranged $286–$1,003 by hour
Spend was flat from 1pm to 7pm. Cost per lead swung from $286 to $1,003 across the same hours. The algorithm that was supposed to handle this didn't.

Why Open Web Content is Now Just Free Training Data

Meta Ads video: 10.5M plays, 4.8% reached the end, 5.9% counted as a ThruPlay
10,523,670 Meta video plays in 23 accounts: 22.5% reached a quarter, 11.7% half, 4.8% the end. ThruPlay counted 5.9%. Completion per account ran 0.3% to 14.7%.

Google Ads broad match: 12 keywords alongside phrase match, zero spend in 10 days
Twelve broad keywords ran alongside phrase and exact for ten days and recorded no meaningful spend. Phrase already does what broad claims to. Exact does it cheaper.

Google Ads lead form conversions are attributed to click date, not submission date
Google Ads reports lead-form conversions on the click date, not the submission date. After repointing every campaign to a new form, the old one kept accruing leads for...

The GA4 conversion in Google Ads fired on a /thank-you/ page that returned 200 to curl
If your conversion fires on a URL that returns 200 to anyone, your conversion count is inflated and smart bidding is optimising to it. A thirty-second test, what we...

Google Ads counted three conversion actions on one lead. Target CPA optimised to the sum
One good lead registered three times. Smart bidding optimised to the inflated count. And the API field that says a conversion action isn't counted turned out to be the...

Google Ads Ad Strength 'Excellent' converted at 1.3% vs 3.8% for 'Good' across $75k
In one account, ads rated Excellent converted at a third the rate of ads rated Good, and Poor ads had the best click-through rate. Ad Strength is a diversity score,...


