None of my client sites have been hit thus far. Now, I realize as well as the next person that that could change tomorrow; maybe they just haven't gotten to me. But none of my client sites have EVER been seriously hit, and we've been in this pretty much from the beginning of AdWords. I've never had more than five or six keywords go inactive at a time. Therefore, I think I might be doing something (totally inadvertently, cause I don't know for sure what it is) right.
So here are a few items from my checklist; if they can be used on your site, you might try. If your campaigns have gone completely dark, it could hardly hurt at this point. Various theories have been posted about what's affecting all this. I can only tell you what I do and don't do.
I use a combination of broad and phrase matches; and I make *extensive* use of the negative keyword filter. I really believe this must be an important step that makes Google think I'm serious about relevancy and keeping track of how my campaigns are going. I also use Dynamic Keywords in the titles when it's appropriate, and for Dynamics I use broad match, but I don't have more than a couple thousand running that way (so they don't think I'm a keyword junkie) and again, I'm using the negative filter to aim for relevancy.
I also don't think it's necessarily all based on low CPC. I have a handful of words going up to $27 or so, but I also have a boatload that maintain at 4 and 5 cents (I don't think I have any less than 4 cents at the moment) By a boatload, I'd say at least a third of them. But my campaigns also hold higher CPC keywords, and maybe that makes a difference, not having ALL low CPC. If it does, you could maybe try adding a few higher CPC keywords, ones that maybe aren't as likely to get so many clicks (like a four word exact match phrase, for example) just to have it there - if you get a couple of clicks, it won't be a big deal, and meanwhile, you don't look like you're trying to bottom feed. (I mean no disrespect by the term bottom feed - I'd love to be paying 4 or 5 cents for everything if I could get away with it)
My product keywords point directly to the page on which that particular product is offered, or the category page for the product plus similar products. I make sure my client has added text describing the products and categories to the page before I will point a keyword to that page; I tell them 50 words minimum describing what the widget is and what the benefit of owning it will do for the customer. This is for every product and category page, even on the site with 1500+ products. This is never going to be finished, it's ongoing, because there are so many products and they are always adding new ones. But all the time put into it pays off big time - our CTR on many of these products is well above 50% and their sales (conversions) have quadrupled since October of last year.
I also have to go through periodically and spot check my landing urls, to make sure they're still valid. This is important on my dynamic sites, because my customers enter products themselves, and sometimes they change something around so that the landing url changes, and before I started checking, I would find out that I was running ads going to non-existent product pages, or product ads being redirected to the home page instead of directly to the product page.
95% of my keywords point to a page where that particular word or phrase is included in the page, usually more than once, but not so much that to the naked eyeball it would look like keyword stuffing. If it's an informational page and not a direct product/order page, I try to make sure that that page contains descriptions of widgets, information about the use or abuse of the widgets, and/or latebreaking news about the widget.
I make sure my ad text is unique and looks as different from my competitors ads as possible. If a competitor copies my ad (which often happens) I may watch to see if I'm still maintaining position and if my CTR is holding up, but usually I just go ahead and change my ad. Yea, I can P&M about someone being slimy enough to steal my copy, but while I'm doing that, he's eating my lunch. So I change it. Yea, sometimes I lose the history. Again, I can spend time tearing out my hair, or I can keep moving. I have found that if I keep my ad copy very focused and my ad --> landing page relationship tight, the history comes back very close to what I had originally, very quickly.
If I have a time or date sensitive ad, I make sure it doesn't run past the point of its relevancy. Do you have any idea how many 4th of July ads are STILL running right now, on the 13th of July? I do, because I have a personal site with a related theme with AdSense on it - heck, there's one ad shows up for a musical event that was held in March of 2006. That's not only not going to get any clicks, but it also may look to Google like you're not paying attention to the quality of your ads.
There must be lots of other things that other people are doing that I haven't thought of - would love to hear about them, if you'd care to post.
Worst case (and the one I suspect): the new minimum bids are all about boosting CPC on certain high ROI KWs, and the 'quality score' is a phantom.
Best case: until Google defines what it means by 'quality', any changes to landing pages will just amount to shots in the dark. IMO, energies are better spent where we can actually expect results.
I hope that we can all recoup by finding other ad venues that will be more appreciative of our business. Google has been completely consumed by its own hubris.
Quality is 100% subjective.
The problem is - many of us have spent several years trying to meet Google's "quality" measure, just to get screwed with everyone else.
Alot of wasted work. Don't repeat our pasts.
