Suppose you test an ad and learn that it falls slightly short of your objective. Were you an all-or-nothing thinker, you might kill the ad, period. But suppose you had created a variation of the ad—say, an alternate incentive offer, headline or photo. Suppose both ads fall short of the objective, but one significantly outperforms the other. Now you have learned that subtle variations make a difference. Perhaps if you test additional ones, you will be able to eke your results up until you arrive at something profitable.
If an ad utterly bombs, let it go. But if you’re close, all-or-nothing thinking might just lead you to quit too soon.