Keywords:
Specific widget- (QS, OK .30)
Specific brown widget- (QS, OK .30)
Specific brown top widget- (QS, Poor $1.00)
All keywords were in the same ad group pointing to the same landing page.
We started by creating new landing pages for each keyword:
www.example.com/widgets/specific_widget.htm
www.example.com/widgets/specific-brown-widget.htm
www.example.com/widgets/specific-brown-top-widget.htm
Landing pages were updated to be keyword specific. Keyword only in title meta tag, keyword twice in description meta tag, keyword only in keyword meta tag. Moderate to heavy keyword use in body with use of h1, h2, and h3 tags. All occurrences of keyword in body BOLDED.
Next we deleted “specific brown widget” and “specific brown top widget” from the ad group. Then we updated the AD for “specific widget” to include keyword in title and description line 1 and pointed ad to keyword specific landing page. NO CHANGE IN QS.
Next we created NEW ad group for “specific brown widget” and NEW ad including keyword in title and description line 1 pointing to this keyword’s landing page. QS CHANGED FROM “OK .30” TO “GREAT .04”
Next we created NEW ad group for “specific brown top widget” and NEW ad pointing to this keyword’s landing page. The keyword was too long to be included in the title of the ad so we used “specific widget” in the title and the keyword in the description line 1. QS CHANGED FROM “OK .30” TO “OK .10”
We then deleted the keyword, ad, and ad group for “specific brown top widget” and created a NEW ad group and NEW ad. Everything was the same except in the title of the ad we used “{KeyWord: specific widget}” and the QS CHANGED FROM “OK .10” TO “GREAT .04”
NEXT DAY: QS for “specific widget” changed from “OK .30” TO “OK .20”. We then deleted the keyword, ad, and ad group for “specific widget”. We changed the name of the landing page from “specific_widget.htm” to “specific-widget.htm” and recreated the ad group and ad pointing to the renamed landing page. QS CHANGED FROM “OK .20” TO “GREAT .03”
Observations / Suggestions:
When naming landing page files use the keyword.
Multi word keywords should be separated by hyphens, not underscore, in the landing page name.
Keyword use on landing pages including use of h1, h2, h3 tags.
Bolding keywords on landing pages.
Keyword in landing page title meta, description meta, and keyword meta tags.
Keyword in ad title and description line.
If the keyword is too long for title line use the {keyword insertion}.
Keyword specific ad groups.
It’s all pretty basic stuff but it seemed to work for us, hope it helps.
[edited by: eWhisper at 3:45 pm (utc) on Mar. 18, 2007]
[edit reason] Please use example.com for sameple links. [/edit]
We noticed a keyword with a strong QS and the ONLY reference to the keyword on the landing page was in an "alt text" tag for a graphic.
It’s like I said to some of these Google people. You’ve totally “worn me out”. To spend a couple thousand dollars you want me to expense five thousand dollars in work, which you may penalize the next week. I’m not here to learn all the nuances of John Madden Football and pour over the rulebook on a daily basis. I’m here to spend money on advertising to make money not learn an algo.
Great intentions by the original poster, but I don't think any of these things help in a way we'll ever understand. We don't do any of those things, and all of our major keywords - over 500+ - for a VERY comptetive industry are GREAT with minimum bids in the 3-5 cent range.
Thats great limitup, we to have many, many keywords in the "great" QS area that we did nothing to or for; however we also had a small number that registered in at "OK" or worse. I simply wanted to post about what WE did and how it affected OUR QS.
I tried your above method, and it made my quality score worse.
venrooy, It wasn’t really meant to be a "method", just documenting the results WE got by doing specific things.
Personally I think it's purely a money alogrithm. It doesn't matter what your page looks like, or how well it lines up with your keywords. The site that makes them the most money for any particular keyword, is going to get the highest quality score.
I agree 100%
For the simple reason that no algorithm would be able to establish the "quality" of a page for a visitor, by just analyzing the web page, out of context. PageRank shows that the quality of a page has to be established in the context of the web, but landing pages are one-off, often without any referring links to it.
IMHO QS is all about CPC, CTR and historic data.
For instance in our campaign we have:
- simple widgets [QS=Great]
- black simple widgets [QS=OK]
?
all the keywords are on our land page!
Systematically following the steps taken by limoshawn will go a long way to achiving this, and for limoshawn it is all valid since the specific keyword is actually being offered for sale on each of the landing pages.
However, blindly following these steps without actually ending up with valuable content on the landing page (Content pumped full of keywords will be relevant but may still be worthless - e.g. "brown widgets are widgets that are brown etc.") will not provide a rubust fix.
OTOH - providing lots of valuable, relevant and specific content without necessarily following all the steps suggested will, IMHO, still provide a Great QS
However, blindly following these steps without actually ending up with valuable content on the landing page (Content pumped full of keywords will be relevant but may still be worthless - e.g. "brown widgets are widgets that are brown etc.") will not provide a rubust fix.
That’s a great point Pengi. We didn’t do any sort of keyword stuffing on the landing pages, we simply made sure that the specific keyword was on the page at least once with a "headline" tag and in the meta tags for the page.
Mimmo, I can't speak for you, however when we reviewed the negative QS we received on some keywords the immediate response was "we sell (specific brown top widgets) how is this keyword not relevant?" but when we looked at our site in absolute terms as an algorithm would we realized that while we sold "specific brown top widgets" there would be no way for an algorithm to know this.
So in absolute terms is the specific keyword "black simple widgets" represented on your landing page, in your ad title, in your ad text? Is the keyword represented in that order, "black simple widgets" and not "simple black widgets" or "black simple widget"?
Just some food for thought.
[edited by: eWhisper at 3:46 pm (utc) on Mar. 18, 2007]
[edit reason] Please use example.com for sample links. [/edit]
Does anyone have any information on how Google views..
/specific-widget
vs
/specific-widget.htm
The first method doesn't include the file extension, but lands them on the same page. Do keywords in the url carry the same weight as keywords in the file name?
I would also ad we did see an improvement on one site where we only added the privacy policy. This was in the last update but it did make an improvement.
Remember some of those major companies whose ads rank well put new meaning to the word cheap. If they even smell they can get ads cheaper for lack of advertisers and abundance of clicks they’ll bid lower. Plus the newbies with no bid history aren’t going to start bidding up for quite a while. They can't replace you immediately.
"widget support"
Page name, title, image alts, h1, h2 intro paragraph and sprinkled throughout all contain "widget support".
QS and min bid did NOT change from OK, 20cents. And, widget support is basically my whole business but there are many more appropriate phrases that i concentrate on. Also, ad text was:
{Keyword:Widget Support}
Widget Support
Quick Response and Great Rates.
...provided this observation is valid, I find it rather comical. I always thought of PPC as being the quick fix for people who either couldn't or didn't want to compete in the SEO game. But now you have to do SEO techniques to compete on even PPC? Hmm. I wonder what the secondary implications of this move are? And I wonder if Google has fully thought this one out.
I started a new ad group and some of the keywords were new too.
Any fix to having the search term in the ad? This is a brutal requirement to always meet.
"I refused to touch my account for many days and 75 of my keywords came back last night without doing a thing."
I too experienced an increase in KWs being activated yesterday. Do you think Google went too far and is trying to roll some of the changes back? Or do you think you're benefiting from other advertiser's KWs being diabled?
Both.
Tiki It’s not atypical in my areas to have a decrease in deactivated keywords over the weekend because Adword clients are pausing accounts. Since the words didn’t return to inactive today there has does seem to be been some rollback. I caution against reading anything into that. My personal opinion is they stalled the full QS implementation. In other words a great many accounts are still unaffected. If there is going to be a rollback I would speculate it to take place in 4-12 weeks. All smart businesses know at some point they won’t be able to get back a percentage of the business they drive off. Personally with them I think they’ll leave many accounts in a permanent-testing phase to gain more data and see how far they can push bids. Plus a roll back would indicate a mistake on there part and people would know to gang up on them once the same thing occured again.
After 3 days at QS "great .05" the keyword "specific brown top widget" changed to "OK .10" at the same time the average position changed from 4.3 to 1.8.
I wonder if the keyword's position has anything to do with the QS algo OR if everyone advertising with this keyword was affected by the QS which changed the positioning.
keyword's position has anything to do with the QS algo
I believe position has everything to do with QS. From my observations, if the KW averages a 1.X position, G will not touch it.
It follows that I also believe the QS has very little (if anything) to do with relevance, and ad & landing page copy don't explain most QS moves. This is based on observations across several accounts.
If someone has position #1 KWs affected, and does not have glaring quality flags such as search arbitrage or affiliate landing pages, I'll reconsider. But to date, it looks like low position+high CTR=high opportunity cost is the main determinant of QS damage.
