Two decisions repeat forever in paid acquisition: this is working, spend more; this is not, stop. Both are done badly by most people — too fast in one direction, too slow in the other.
Scale gradually
Large budget jumps disrupt performance and restart learning. Increasing by roughly 20-30% every few days lets the system adapt, and it means a mistake costs you a little rather than a lot.
Two ways to grow:
- Vertical: more budget on what already works. Simplest, and limited by audience size.
- Horizontal: new creative, new audiences, new offers. Slower, but this is where real headroom comes from.
Expect efficiency to fall as you scale
Your cost per acquisition at $500 a day will usually be worse than at $50, because you are reaching beyond the easiest buyers. That is normal, not failure. The question is whether it is still profitable at the new level, not whether it matches the old number.
Kill decisively
- No conversions after spending roughly two to three times your target CPA: stop the ad.
- Cost per acquisition consistently above your ceiling for a week or more, with adequate volume: stop.
- Frequency climbing while results decline: the creative is spent, replace it.
Kill ads, not necessarily channels. One losing creative is a creative problem; a whole channel losing consistently after several honest tests is a different conclusion, and worth accepting rather than fighting.
Keep a log
Write down what you changed, when, and what happened. Without it you will re-run the same failed test in six months, and you will not remember which creative was the one that worked.
Try it
Write your kill rules as explicit numbers — CPA ceiling, spend threshold, frequency limit — and apply them to your current account today, without negotiating with yourself.