Is anyone's cloaker actually holding this year or are we just getting lucky

Question for people still doing this rather than people who read a blog about it.

My setup hasn't materially changed since 2024 and it was still passing as of last week. Same fingerprint checks, same ASN filtering, same timing gate. I have not touched it in fourteen months and it has not tripped once, and honestly that makes me more nervous than if it broke.

So which is it. Is detection genuinely worse than everyone assumes, or am I on a list somewhere with a flag next to my name waiting for someone to press the button? What's actually failing for you this year and how did you find out it failed?

Both, and they're not in tension.

The reason your 2024 setup still passes is that most of what's called detection is still a request-time classifier, and a request-time classifier can only see what you show it. That hasn't changed and it isn't going to, because the alternative is expensive.

What has changed is that the second pass is no longer request-time. Landing pages get pulled later, from residential addresses, by things that aren't crawling on your ad click. Your gate never sees that request because it doesn't look like a click, it looks like a person. So you pass every check you can see and fail one you never observe, and the first you hear about it is the ban notice with no correlating traffic in your logs.

That's my read. I have not proven the second half of it and I'd like to be wrong.

I don't think that's right, or at least it isn't what I've seen.

Caveat: I run four accounts, not forty, so my failure rate is a small number of events. But every account I've lost in the last eighteen months has died with a review timestamp inside a few hours of a creative change, not a landing page fetch. If there were a delayed out-of-band page pull I'd expect the deaths to be uncorrelated with what I did that morning, and they aren't. They're extremely correlated with what I did that morning.

I think the creative is doing more of the work than the page, and cloaking is getting blamed for something upstream of it.

Fair, and that's a better dataset than an argument from architecture.

Though a creative change is also the thing most likely to trigger a re-review of the whole ad object, page included. Correlation with your morning is consistent with both stories.

it's detectable, it always was, the question is whether it's worth their time to look at you

so the answer is nobody knows and we're all just hoping. cool cool cool