PPC'ers: How long do wait before trying to beat your Adwords ads?

[Nick Brighton]

I was wondering, do you continually try and beat your winning ad, or do you run both for a certain length of time without trying to tweak them?

I'm thinking it's better to run both for a longer period of time before you start to tweak ads, even if you've got one ad that is currently way outperforming the other.

What are your thoughts?

[svtdbird]

I generally wait about a week before changing the poorer ad. Only if it's obviously performing terrible do I change it sooner.

[KirkMcD]

Originally Posted by Nick Brighton

What are your thoughts?

Dump the poor performing ad and move on to the next campaign.

[tommyp]

Yes continually try to beat it, although there does come a point where you may want to stop and finally leave the best running ad all the time. It's all a personal choice but maybe stop when you can't get better results.

Yea you do need to let them run long enough but to say it needs to be a week, or a month or whatever isn't exactly true because what any particular campaign needs will vary from campaign to campaign.

It also depends on the gumption of the one running the campaign.

If you get lots of data to work with quickly then you don't need to wait as long. Sometimes you will know within a couple of days and other times you may be testing for six months.

Sometimes people will get enough data quickly to know something but may still choose to let things run because they want to see if there is a connection with performance and day of the week or week of the month or month of the year.

Some say after 30 clicks get rid of the worst ad and write another, some say scrap the worst performing one after 100 to 300 impressions. Different people say different things.

I do think you should have enough numbers to make a determination from but what I say the bottomline is: make changes when the differences between what you are measuring (CTR for example) are statistically significant. That's what matters.

I can't remember off the top of my head but there are free tools online where you can put in your impressions, CTR, etc., and it will tell you which ads are the winners and losers and if the difference is statistically insignificant it will tell you that too and then you can choose to let them run longer.

Okay I'll be nice and find one:

Chi-Squared Split Test Calculator

It's not the only one out there and I don't know how accurate it is but it should be okay. Only tests two at a time though.

PS. Your highest CTR ad may not necessarily equal your most profitable ad. You may be in a situation where your worst CTR ad is actually bottomline the most profitable. You need to test and track. If all that stuff is over your head then start learning about it and in the meantime go ahead and do the basics like improving your ads based on only CTR if you want but eventually you want to track, test and dig deep to know what's really going on.

Originally Posted by KirkMcD

Dump the poor performing ad and move on to the next campaign.

Huh???