3 to 5x on a cold audience. 5-10x on retargeting campaigns and upsell/cross sells. It really depends on the client though.

Is there reasoning to not drive this into the ground, I.e. where profitability for the campaign hits 0?

At a practical level I wonder if you / they are cognizant of an exhaustion measure, if new campaigns are introduced rapidly enough to continue to maintain a 5-10x return, whether the spend is managed to a predefined ROI, etc.

For what it’s worth, I’ve always seemed to drive it into the ground, but those are for short term campaigns. The long term campaigns I’ve inherited have usually already been driven into the ground before me (though the PE firm I was working for bought that distressed startup on the cheap).

It depends on what you are doing. Many are just middleman, so they sell conversions they drive from FB, the difference of which is your profit. You have an incentive to keep conversion cost as low as possible and quality as high as possible to maintain your conversion payoff.

Driving it into the ground might be profitable for a product owner but not to an ad agency which pockets the middle.

I’m finding that extremely hard to believe. What’s the product category?

Find it as hard to believe as you would like but I have no incentive to lie. The most profitable campaigns tend to be high end consumer goods with a monthly recurring element built into it. The worst performers tend to be anything targeting small business.

Yes, I'm targetting small business ($25/month typical), and I've never been able to make google ads pay, even after optimizing the f*ck out of it.