My AdWords click-through rate continues to climb
by Sapphire (May 18, 2005)
The title pretty much says it all. If you’ve been reading the PPC Category, you know that I’ve been experimenting with AdWords to find out just what it can do for me, with very poor results at first. But the ideas I had recently on improving my AdWords click-through rate are panning out. By eliminating general keywords, I’ve brought my CTR up to a respectable level on both the original, underperforming site and the new site I just started running. But I just realized something else today.
I took a long look at my keywords for the original site, and how they’ve performed over time. I realized some keywords had very decent CTR’s just on their own. Those are definite keepers. Others had an awful lot of impressions with very few (even zero) clicks - and mathematically, I had to get rid of those. But I was surprised to realize what some of them were.
I had believed they were specific keywords. I don’t want to get into the specifics of what my site offers, but let’s use an equivalent example. Let’s say my site sells beach towels. Only beach towels, not hand towels or washcloths or kitchen towels. Some of the keyphrases I had to delete today were as specific as “green beach towels” or “Egyptian Cotton beach towels”. Why weren’t these guys performing?
It’s anyone’s guess - even Google’s. If you work with AdWords, over time you’ll see some very odd trends. Like, very few searches for “green beach towel” but lots for “green beach towels”. It’s just a quirk of how people think when they create search strings.
What’s confusing to me, however, is when “green beach towels” gets very low traffic, but a higher CTR than “green beach towel”, which gets more traffic. That means the sort of people who would put the singular keyphrase into their search bar don’t find my ad appealing, but the sort of people who put the plural version of the exact same keyphrase find my ad considerably more appealing. I can’t imagine why, and I’m not sure what to do with that info. Except delete the underperforming keywords.
I guess this gives me some insight into why Google disables certain keywords before you can even implement them. You really have to see this stuff in action for yourself to be convinced of how random it all is.

