What is your longest-lived ad account actually sitting on?

Not asking which provider is best. Asking a narrower thing: the ad account you've had alive the longest, what is it actually sitting on right now?

Mine is coming up on fourteen months, which is by far my record. It's on a residential line that has never rotated, same city as the billing address, and the machine touching it does nothing else. I don't know which of those three is load-bearing and I'm not going to experiment on it.

Give me the setup and the age. Not the theory.

Nineteen months. Static residential, one account, one browser profile, one physical box in the corner of the room that I don't otherwise touch.

The thing I'd flag is that the fragile part was never the IP, it was DNS. Two of the accounts I lost early on were on setups where the browser was going through the proxy and the resolver wasn't:

# what killed two accounts
proxy: residential, city A
dns:   1.1.1.1 via the host's own uplink, city B

Everything looked fine in every fingerprint checker I tried. Fix that before you spend money on better proxies.

Eleven month, mobile proxy, never rotate. Rotation is what kill accounts, not proxy type.

mine died at nine days so I have nothing to contribute here but I'm reading every word

Twenty-two months, and I want to be careful about what I attribute that to because I think most of it is survivorship.

Setup is unremarkable: residential ISP line, static, dedicated machine, no VM, same timezone as everything else on the account. Nothing exotic.

What I think actually matters, in rough order of confidence:

1. Nothing about the connection has ever changed. Not the address, not the resolver, not the user agent, not the screen resolution. The account has never once seen a discontinuity. I'm confident about this one.
2. The account does one thing. It has never logged into anything else, never had a second user, never been opened on a phone. Fairly confident.
3. Low, boring spend. It has never scaled hard. I suspect this matters more than the infrastructure and it's the part nobody wants to hear, because it's the part that costs money.

I have not tested any of this properly and I never will, because the test destroys the account. That's the honest state of everyone's knowledge here including mine — we all have an n of one and a story about it.

Eleven accounts, tracked since 2023, median lifetime by connection type:

Type	n	Median days alive
Residential static	4	312
Mobile, no rotation	3	188
Mobile, rotating	2	41
Datacentre	2	16

Small n on every row and the datacentre pair were also my worst creatives, so that row is confounded. The rotating vs non-rotating gap is the one I'd actually stand behind.

Had this exact argument with a partner last year - he was convinced mobile beat residential and I couldn't shift him.

Then he lost three accounts in a fortnight after his provider silently started rotating on a schedule and didn't announce it. Nothing on his end changed. He'd have sworn under oath that nothing changed, and he'd have been right.

That's the bit that gets people. Your setup doesn't have to change for your setup to change. Do you actually check yours, or do you assume?

That last question is the useful one in the thread. I check monthly now. I didn't before.

Bumping - the nineteen month one is still up, twenty-two now. Nothing changed. That's the whole update.