creative performs — field notes · 4 min read
kill rules: when to shut an ad off
most budget isn't lost on bad ads. it's lost on bad ads nobody was willing to turn off.
the short version
- decisions get written before launch, not argued after.
- every test needs a spend threshold before it's allowed an opinion.
- kill on evidence, not on a bad morning.
- clean exits are what make room for the next winner.
01why accounts leak
the usual damage isn't dramatic. it's a handful of underperformers left on for another week because someone believed in them, and a promising test switched off after two days because the first numbers looked rough.
both are the same mistake: the decision was made by mood instead of by a rule agreed on in advance.
02thresholds before opinions
nothing gets judged before it has spent enough to say something. we set that floor per account based on your average order value or cost per lead, so early noise doesn't get promoted to a conclusion.
below the threshold, an ad is still learning. above it, it either clears the bar or it goes.
03what the rules look like
written plainly, before launch: the spend floor, the target it must hit, how long it gets after clearing that floor, and what happens on the way out — pause, iterate on the opening, or retire the concept entirely.
the rules cover winners too. a proven asset showing fatigue has a scheduled successor, so scaling down never means going dark.
04what discipline buys you
killing fast isn't cruelty to creative — it's how volume stays affordable. every dollar pulled off a proven loser funds another real test.
and after a few cycles you have something better than a lucky ad: a written record of what your market rejects, which makes the next batch sharper before it ever launches.
an ad you keep out of hope is a budget line with no job.
want this run on your account?
20 minutes. we walk your ads, spend and offer, and show you where the growth actually is.