| 11:51 am on Dec 13, 2005 (gmt 0)|
"To set expectations that I can actually meet, however, I should say now that I will not be able to go into detail regarding the algo itself."
So we have confirmation that it is an algo but not properly researched or tested if one of my clients campaign (msg #:108) is anything to go by unless of course it is a regional problem by ISP and beyond their control.
I have a question, is the problem of adword results disappearing UK only or also happening in other countries?
I am in direct contact with Adwords support in the UK at the moment and might/hope/may/may not be able to throw some light on the matter at a later date!
| 12:20 pm on Dec 13, 2005 (gmt 0)|
I'm assuming pagerank is going to be a heavily influencing factor.
| 1:48 pm on Dec 13, 2005 (gmt 0)|
I've got a new site with PR=0. And I'm running a large AdWords program. So far the effect of the change on us has been entirely positive... more traffic at less cost. So I doubt PR has any effect.
| 3:32 pm on Dec 13, 2005 (gmt 0)|
What is killing me is they are jacking up the prices on variations of keywords that I have a high CTR on. Including these variations would not read clearly in the ad text or on the landing page.
So now to get a higher quality score, I have to include these variations which will make the ad text and landing page text more awkward to read?
And some of my best keywords are those including "com" for example, "widgets com"
What about misspellings?
Do I now have to include misspellings in ad text and on my landing pages?
| 3:44 pm on Dec 13, 2005 (gmt 0)|
So what Google is basically saying is, we have to rebuild and start over with all campaigns and landing pages. I don't mind the work and the challenges but it would have been great if there was a warning about the algo change.
| 3:51 pm on Dec 13, 2005 (gmt 0)|
Ok, so...What is the process if I find a $5 inactive keyword?
Create a new ad optimized for that keyword then update my landing page and make plenty mention of that keyword?
Then will my still-inactive $5 keyword eventually drop in price after this robot crawls the page again?
| 4:19 pm on Dec 13, 2005 (gmt 0)|
When they first implemented the minimum CPC thing, I had a number of keywords that had to be set higher - including a couple dozen that were in the $5 to $10 range. I went ahead and set them to those rates (I have basic instructions from my clients to keep the ads running no matter what) I found that I *never* paid anywhere near $5 or $10 per click, and that they settled back down to a reasonable price after a few days. Of course, it's a risk. I think they were just trying to find out if I were *willing* to pay that much. (In one case, I only was charged $.04 for a word where the minimum was set to $8.00) I realize it's a risky thing to set the word that high, but if you are paying attention, you should be able to shut it down quickly if it looks like you are really being charged what it's set to.
With this last business of rating the landing page, I haven't had a single keyword go inactive, and none of these sites have a PR higher than 5, most lower. (In my own experience, PR means absolutely nothing, since I have sites with PR0 which rank number one or two for every keyword that matters) But my CPCs ARE going down and my CTRs are going up. So it's plus for me.
Of course all that could change overnight. I know that, and accept it. I can shut down all my accounts and move somewhere else overnight as well. I don't expect any guarantees from Google, and I don't guarantee they're going to keep my business either.
| 4:55 pm on Dec 13, 2005 (gmt 0)|
Hmmmm...I wonder how this will affect competition between merchants?
I have made some good $$$ bidding on competitor kewywords...Now, since my ad text doesn't mention the competitor, or the landing page, I guess my quality is lower.
But, what about comparison shopping? Wouldn't google be eliminating options by discouraging competitors from bidding on other competitor keywords and therefore limiting relevant options of maybe getting a product at a cheaper price with a competitor?
For example someone shopping for shoes...
| 5:27 pm on Dec 13, 2005 (gmt 0)|
Got off the phone with the Google rep. A few thing were told to me that makes me understand things better...
If you optimize your landing page, you may see the minimum CPC reduce once the page gets crawled. Your page will be crawled on a regular basis, so optimizing a page to the way google likes will give a quick turnaround in prices.
I raised the issue about variations and misspellings and she will talk to the tech team about that and give me feedback.
I specifically spoke about one phrase that had a max CPC of .29, had an average of CPC of .14, and a CTR of 4.7% which now is a $5 keyword. The keyword is a variation of a merchants URL "widgets com" So I wanted to know how I could optimize for "widgets com" because I already get the person to where they want to go.
And she said with my example that I have to pay the $5/click and if I achieve the same CTR, the price will decrease over time...That is rediculous...lol..Pay for a $5 keyword to get it back down to $.12 eventually?
| 1:38 pm on Dec 14, 2005 (gmt 0)|
Interesting way to solve trademark issues.
| 3:29 pm on Dec 14, 2005 (gmt 0)|
I am fairly certain that PR of a page will not matter at all with the way G views the quality of the landing page.
I had a lengthy chat with my rep yesterday and it seems like the main factors are that things are related.
your page is related to the keyword you are buying, that is related to the ad copy, that is related to the ad group, which fits into the campaign.
discount blue widgets(keyword)
Buy Blue Widgets
Shop now and save
at the Widget shop.
blue widgets (adgroup)
I am guessing that if you look at their keyword suggestion tool, for the page you want to send someone to, you will be able to pull the keywords that they think your page is about, and should pass "quality score" checks if you purchase those terms, as long as the ad copy fits what you are trying to do.
| 3:55 pm on Dec 14, 2005 (gmt 0)|
You think how the campaigns are organized weighs into this?
Suppose I am running 200 different ads? I can't do 200 campaigns...You can't do that many campaigns.
| 5:42 pm on Dec 14, 2005 (gmt 0)|
The idea of utilising how adgroups and campaigns are structured as part of the algo seems strange as it does not have any bearing on the end user (searcher) experience.
| 7:22 pm on Dec 14, 2005 (gmt 0)|
I am thinking it would be more of Adgroup to adtext to keyword relationship rather then campaign.
Think of why they may do this?
If you just put your max of 2000 keywords that are not related together in an adgroup, and write a broad ad that targets them all, that is a pretty poor way of organizing them, from both a management standpoint as well as theming standpoint. I think
Back in the day they said to use adgroups to organize and help ads show up in the content network. I am guessing this is along the same path.
Don't get me wrong, I dislike the system and am seeing 100's of terms switched to inactive, but I am not seeing as much on some of my newer better organized accounts as I am on the older / clunkier ones. I would love to see adwords roll back about 2 years to the exact match days (c:
| 7:25 pm on Dec 14, 2005 (gmt 0)|
A couple of months ago I tested creating 2 new ad groups, one with the keyword in it, one without. Bizarely, the one without had a lower min bid. But regardless, I guessed it was being factored in in some way. If ad group name is factored in, why not campaign name.
Of course it could just be the min bids were totally random. That would also explain it :(
| 3:47 am on Dec 15, 2005 (gmt 0)|
What is your opinion on using a tracking URL for a landing page and this new "spider"
Let me explain. For 2 years now I have used the same landing page for my campains... index.3.htm
A month ago I implemented some php tracking scripts so I turned my index3.htm page into a "meta refresh" page that now forwards to 3.htm
My question now is.... is this spider only spidering my "meta refresh" page or is it following through to my 3.htm
I'm trying to explain the decline in my account lately but it started way before this new spider......
Just curious if anyone has an opinion...
As a side note all these changes from the last few months are KILLING us. Untill October we were spending 15K to 30K a month on adwords for the last two years... Looks like December will be around 4K
| 4:16 am on Dec 15, 2005 (gmt 0)|
I tried very hard to discern a pattern as to why perhaps a dozen keywords in many AdGroups went to $5.00 - $10.00 and the rest stayed at 10 to 20 cents.
It does appear that the spider is looking for the keyword/phrase on the landing page.
Problem there is you have a computer doing this, not a human. One product line I advertise has $50-$100 items which I really care about selling. Those bids remained in the 20 cent range where I am happy.
There is also a similar $15.00 line of goods that I felt obliged to include on the site. I had a different
AdGroup for the cheaper items and every one of those keywords is "Inactive" asking $5.00 & up.
It appears to be the phrasing on the page that is "throwing" the spider. Any 8 year old could read the text and interpret it properly, but a computer is lost:
Finally, we offer the green widgets
Click Below for the:
Now small "small size green widgets" or "large size green widgets" ask outrageous bids. The other expensive products had a complete enough description on a single line to allow them to continue to show uninterrupted.
Noticed the same problems with numbers and synonyms
I might list a "one propeller widget" on my site and/or in my ad.
"one propeller widget" will ask a low bid, but "1 propeller widget" or "single propeller widget" would go inactive.
The spider clearly appears to travel without a Thesaurus. An obvious synonym for one of my products, again known to any 8 year old, was rendered inactive in every AdGroup that it appeared in.
So what is the answer? Especially with synonyms which is the essence of success in AdWords, thinking of the different variants searchers might use to find your product.
Clutter the page with all possible synonyms repeated?
Stuff your ALT tags?
I thought the more recent trend in SEO was that if one created a page that visitors found attractive, it's likely that search engine spiders would find it similarly attractive. Moving in the other direction will only encourage spamming tactics.
| 4:45 am on Dec 15, 2005 (gmt 0)|
|It does appear that the spider is looking for the keyword/phrase on the landing page. |
It doesn't rely on that.
Outside of the meta description and some alt tags on the bottom navigational menu (e.g. "about us") the homepage my ads link to has no text at all on it.
None of my keywords have had their minimum bid requirements changed. Avg CPC, impressions, position, and CTR have stayed in their normal ranges.
(The page also places highly on a number of keywords for our space. Google's algorithms have obviously gone quite a bit beyond text.)
FWIW - it is commercial ecommerce rather than informational content.
| 7:32 am on Dec 15, 2005 (gmt 0)|
I certainly don't wish this on you Cerny, but is there any chance that the spider simply hasn't gotten to your page yet?
It must be a mammoth task to go around to each ever-changing landing URL and give it a quality score that can be used in conjunction with determining keyword value. In fact, out of hundreds of AdGroups in my campaign, each day I find AdGroups that had been "clean" and now suddenly have a dozen keywords flagged.
They can't be doing it real time. For the heck of it, I set up an AdGroup using my usual keywords with a destination of www.eeexaaaammmppllee.com and nothing was flagged.
The way I envision your landing page based on your description of it, a banner or Flash is doing most of the "talking". I don't expect you to give out any "secrets", but think to yourself, what about your page warrants a high quality score?
I'm glad you didn't suffer from this surprise. For that matter except for the first day when I found some words I cared about were missing the mark for lack of a nickel, I didn't really lose much either.
FWIW, perhaps part of all the secret algos is that only a certain percentage of ads/pages are checked rather than each and every one as we've been led to believe. Much to my chagrin, I had a new ad running for about 2 months with a typo in the URL. It always went to "page cannot be displayed". I found that myself.
I've read of people losing most all of their keywords, folks like you losing none and me kind of in-between. Hopefully, it will normalize soon. Someone referred to this landing page quality score as a program "running in the dark" (and at a terrible time of year for many). I believe there is some truth to that and that is not meant in any way to disparage your good fortune.
The question of the day is that if, over time we figure out how to improve page quality scores, how soon can we expect to be re-spidered? Much of AdWord's appeal has been the immediate results.
| 11:45 am on Dec 15, 2005 (gmt 0)|
Is there anyway to tell what your quality score is?
| 1:51 pm on Dec 15, 2005 (gmt 0)|
|I certainly don't wish this on you Cerny, but is there any chance that the spider simply hasn't gotten to your page yet? |
Anything is possible. But, since the page ranks well on organic listings for many related keywords I'd be surprised if we have any problems. Google is able to classify and rank the page in their organic listings for relevant keywords, so why would they have any problems ranking it for paid ads?
|The way I envision your landing page based on your description of it, a banner or Flash is doing most of the "talking". I don't expect you to give out any "secrets", but think to yourself, what about your page warrants a high quality score? |
No, there's no Flash. Its just jpg/gif. Its ecommerce - think homepages like Victoria's Secret, JCrew or the Gap.
As far as secrets, I don't have any. We've made no attempts to optimize the site for search engine rankings.
As far as what warrants a high quality store - its the content of the site itself. If someone searches on our keywords and clicks on our ad, they find exactly what they set out to. There's no ambiguity.
| 2:30 pm on Dec 15, 2005 (gmt 0)|
I think as soon as some larger customers get burned by the arbitrary nature of this quality score mechanism, Google is going to have to reverse direction.
Big spenders don't want to get into details. They just want to lay down the big bux and expect results.
If you're senior management and you suddenly find your company held hostage by some lowly content developer who misused some meta tags which wreaked havoc with your AdWords listings - don't you think you're going to be pretty pissed off with your AdWords Rep?
| 6:42 pm on Dec 15, 2005 (gmt 0)|
|We've made no attempts to optimize the site for search engine rankings. |
Well that's good to hear. With SEO trends changing all the time (more inbound links - no, less inbound links, etc), that's what I'm hoping for too. My sites have been attracting people and Google should like that too. Like to keep it simple, straightforward and honest.
There is still something "wacky" going on with the page quality score as many can attest to. On the same page, I'll have 2 or more cells with different products of the same nature, but the same layout. Google will "pick on" a few keywords for one ad and demand a ridiculous amount while the other ads are cool.
They're still testing the darn thing as far as I'm concerned and they picked a rotten time to test.
I am looking into the possibility of HTML errors that might be throwing off the spider.
| 11:54 pm on Dec 15, 2005 (gmt 0)|
What about the lead gen businesses for the highly competitive mainstream areas that basically use IMAGE ONLY forms as a landing page?
If Goog relied on looking at the content of this type of page then they wouldn't find anything but images.
| 1:36 am on Dec 16, 2005 (gmt 0)|
maherphil - i agree. also those putting up just a sales page to sell an ebook or dvd, surely they must have been hit. they have one landing page and thousands of keywords, surely these will become inactive for search very quickly .
| 3:02 am on Dec 16, 2005 (gmt 0)|
They'll be doing changes today, I hope they are reversing course and abandon this idea.
| 7:59 pm on Dec 16, 2005 (gmt 0)|
thegreatpretender. How do you know they are making changes today?
| 9:21 pm on Dec 16, 2005 (gmt 0)|
When you login to AdWords you'll see a message stating AdWords won't be available today from 20.00 to 22.00 PST as they will be doing maintenance work.
| 9:53 pm on Dec 16, 2005 (gmt 0)|
Yeah, what are they doing introducing this right at the peak of the christmas season?
| 10:27 pm on Dec 16, 2005 (gmt 0)|
lets hope when the maintenance work is over we can rejoice at the number of keywords we have turned back on - it will be interesting to see if what happens
| 10:32 pm on Dec 16, 2005 (gmt 0)|
Well, I can announce positive news...
Google took a keyword with a high CTR, that I was paying as low as .12/click and made me change the max CPC from .29 to 1.00.
Over the first few days, I paid the full dollar, but had an even higher CTR rate than before (probably because most competitors were flushed with the keyword going inactive for them, meaning less ads to choose from, and more chance of my ad getting a click)
After about a week, that $1 per click has gone down to .55/click. Still far from .12/click, but at least the cost has quickly gone down. I also hope that since I added more content that targets my keywords, I can get a higher quality score.
I'm also happy to report with another keyowrd, that I had to move my .29 max CTR to a $1, but I was not charged a $1...I have been charged about .16...(I have a 12.5% CTR since I reset to $1) So I can confirm that even though you may be asked to pay a high dollar amount to bring from inactive, you MAY NOT be charged that amount.
The only problem is YOU MAY NOT, so you never know if it is worth paying for those $1, $5, and $10 keywords...You may pay max, you may not. I wish google would at least put in some kind of prediction as to how much you will actually pay...Because why should I take a risk of setting my max to $5 to pay anywhere between $5 and .01? I might activate more $1 keywords and $5 keywords if I get the success like I listed above.
| This 208 message thread spans 7 pages: < < 208 ( 1 2 3 4  6 7 ) > > |