Ok, here we go, stay with me, “KeyGroups”!
One problem I/ and I think most advertisers have, is covering all the base’s with regards to there keywords. So you have a football league and you see t shirts for each of your teams at the moment you have to generate all your keywords with prefixes and suffixes for all keywords, so I have the teams,
Widget United
Widget Lane United
Widget Road United
And I want to add
T-Shirts
Tshirts
Football Tshirts
Football T-Shirts
And your keyword lists get very big very quickly. So this is my idea, being able to list your prefixes & suffixes to a list of “WordBaces” in your account, so my KeyGroup would contain, As “WordBaces” or “BaceWords”
Widget United, Widget Lane United, Widget Road United.
And then I would list T-Shirts, Tshirts ect.. as either a prefix of suffix or both of the KeyGroups “BaceWords”. You would also specify (“link”) an advert to a “BaceWord” inside a “KeyGroup”
KeyGroups would work in place of AdGroups, so you decide to either create an AdGroups or KeyGroup.
So what do people think? Firstly who understands? Secondly who likes it?
They could use an easy-to-understand grammer specifying alternatation, optional parts, etc:
(red¦white¦green)? widgets?
equivalent to:
widget
widgets
red widget
red widgets
white widget
white widgets
green widget
green widgets
(This is a course a Unix "regular expression". If you think that is too ugly - and I do - a different grammar could be easily constructed.)
Google either doesn't get it - or doesn't want to get it.
Google, unfortunately, is not alone in this. None of the other keyword ad networks have implemented this either.
I've been writing software since 1970. (In high school). Sometimes I get awfully disappointed seeing how little progress we are making in some areas. I simply cannot fathom why advertisers need to exhaustively list key phrases (they are NOT key "words", but "key phrases"!), when we've had efficient pattern-matching and grammars for same for decades.
I just about lost it the other day when an article showed up in the local paper about an interactive conversation with a "computer psychologist" on some website devoted to same.
Of course, I immediately recognized this as a derivitive of Eliza. Which was pretty ho-hum already in 1972 (when I was first introduced to it) once you realized what it was doing. But it was a fun introduction to Snobol to play with it and twiddle with the vocabulary... This was, of course, running on a computer that served hundreds of users on my college campus, and had a small fraction of the computing power on my desktop today - and 8 - count em' - 8 megabytes of RAM. (Which was H U G E!)
And then this 1966 technology shows up in the newspaper as the latest thing in 2006... 40 years later!
Anyway, the technology to do this has been around for 3-4 decades, but either Google doesn't get it, or they think advertisers are too stupid to get it.