How Long to Test?
Now that I’m testing one ad against another with carefully selected keywords in the titles, I’m asking myself — how long do I wait? When do I know if one ad is really better than another? Here’s what I’ve picked up from message board comments on the topic.
1. The length of time depends on search volume. I have low volume on my sites, so my samples are going to be small and I’ll have to wait longer.
2. Sometimes it’s obvious. Even in my unscientific experiments with ads, I’ve seen some ads to much better than others immediately. If it’s obvious, make the switch.
3. One person said they change after 30 clicks, another says 60 and a third said 100 clicks, just to be sure.
OK, I’m heading back to my May campaign, to sort by clicks, start pruning the losers and creating new ones to test against the winner.