homepage Welcome to WebmasterWorld Guest from 54.166.148.189
register, free tools, login, search, pro membership, help, library, announcements, recent posts, open posts,
Become a Pro Member
Home / Forums Index / Google / Google AdWords
Forum Library, Charter, Moderators: buckworks & eWhisper & skibum

Google AdWords Forum

This 107 message thread spans 4 pages: 107 ( [1] 2 3 4 > >     
AdWords: Quality Score Discussion
tonynoriega




msg:3155298
 6:03 pm on Nov 13, 2006 (gmt 0)

I hear all the talk about this, but am i missing something?

Is this "Quality Score" viewable? Can i figure out what my score is somewhere...?

Or is this some mytholocial algorithm, that everyone claims to know, but really doesnt?

thanks

 

luke175




msg:3155303
 6:07 pm on Nov 13, 2006 (gmt 0)

I don't think many people claim to know it, but yes it exists and chances are you'll never hear about it unless it smacks you.

Why do you think Google's stock price is up?

aeiouy




msg:3155464
 8:21 pm on Nov 13, 2006 (gmt 0)

They have a feature in testing right now that will show you in a general sense what your QS is for each keyword.

In other words you won't know exactly what it is but you will know if it is low, or good or high or whatever terminology they use for the different levels.

Kobayashi




msg:3155465
 8:23 pm on Nov 13, 2006 (gmt 0)

That would be a nice feature, how did you hear about that?

trannack




msg:3155472
 8:25 pm on Nov 13, 2006 (gmt 0)

Yeah - do tell. That would be brilliant - better still if they would tell us at least what parameters go into the QS - as opposed to the spiel they give us about landing page quality, ad text blah, blah "& other factors" - WHAT other factors?

AdWordsAdvisor




msg:3155556
 9:32 pm on Nov 13, 2006 (gmt 0)

They have a feature in testing right now that will show you in a general sense what your QS is for each keyword.

Aeiouy is in fact correct - although that's just about the extent of what I can say about it at this moment. ;)

Yeah - do tell. That would be brilliant - better still if they would tell us at least what parameters go into the QS - as opposed to the spiel they give us about landing page quality, ad text blah, blah "& other factors" - WHAT other factors?

There has been a fair amount of new information added in the AdWords Help Center since mid-October that should be of use in understanding Quality Score. Starting at the top level, these links are worth exploring:

Ad Quality and Performance
[adwords.google.com...]

Quality and Performance Overview
[adwords.google.com...]

Quality and Performance Factors
[adwords.google.com...]

Troubleshooting
[adwords.google.com...]

And here is one page from the 'Quality and Performance Factors' section that I hope will be useful:

What is a 'Quality Score' and how is it calculated?
[adwords.google.com...]

AWA

sailorjwd




msg:3155626
 10:41 pm on Nov 13, 2006 (gmt 0)

I guess I'm caught in a catch 22.

I am no longer able to advertise on adwords due to the high min bids.

Therefore no matter what i change on my site/landing pages my adwords quality score will never be improved since it won't be evaluated again.

Or, will you still evaluate inactive accounts? And at sometime in the future I could wake up to a functioning account again?

dylan212




msg:3155739
 12:36 am on Nov 14, 2006 (gmt 0)

Therefore no matter what i change on my site/landing pages my adwords quality score will never be improved since it won't be evaluated again.

i don't think you need to be pessimistic, your landing pages will be evaluated again:
[adwords.google.com...]

The AdWords system visits and evaluates advertiser landing pages on a regular basis. Due to the evolving nature of the quality evaluation system, the frequency may change based on improvements to the system or the level of user traffic to AdWords ads and landing pages. We plan to visit pages at least once a month but may do so more often.

ronmcd




msg:3155826
 2:54 am on Nov 14, 2006 (gmt 0)

From that link provided by AdWordsAdvisor:
Currently, the formula is best represented as: Quality Score = (keyword's CTR, ad text relevance, keyword relevance, landing page relevance)* *Where the interactions between the Quality Score variables change as we continue to refine how to measure and define quality in AdWords

Im a programmer and I only see one componant in that "formula" which is measurable. The rest is like a memo from a Dilbert manager - woolly marketing nonsense, it means absolutely nothing. And wheres the componant for "we dont like your sort anymore"? The formula suggests as long as my page is relevant I will be fine, we all know thats not true, adwords is now dictating what SORT of page I build, what type of content I need, how to structure my pages. That "formula" shows Adwords performance is NOT MEASURABLE any more, google can change it any way they want and we wont know. Without stability why should I invest any more time with adwords? How can one really split test when you dont know what the rules are? How can anyone be expected to make changes to their site knowing it could take weeks or months before the quality score is re-evaluated? Do google WANT advertisers to keep buying throwaway domains rather than improving their site for their visitors?

The uncertainty and belief rightly or wrongly that the system is unfair (arbitrage=good, sales page=bad) is a huge disincentive, if google didnt have the ppc market share it has people would go elsewhere in even greater numbers. There may be good reasons for introducing "quality" score, but without concrete and transparent explanations people have no option but to suspect the changes are made for reasons other than "quality".

It saddens me to see AdWordsAdvisor offering help pages like the one above as explanation for how the system now works. I CAN work around the changes from the last few months, but I now choose not to, even though it's still profitable. Lifes too short for me to be bothered with adwords anymore. It's frustrating and frankly unprofessional.

I once questioned whether google saw advertisers as "customers", and AdWordsAdvisor seemed genuinely upset by the suggestion. How things have changed. You dont treat your customers this way ..... or they become ex customers.

sailorjwd




msg:3155839
 3:21 am on Nov 14, 2006 (gmt 0)

Funny thing happened on the way to the poor house tonight.

I had been complaining to my poor old mother who is visiting for the week, about the adwords problems and how my income had vanished.

I was trying to explain LPQ to her.

I let her read AWA's suggesting reading links...

After a 1/2 hr I came back to ask what she thought...

She said, "those pages tell you nothing"

She was a technical writer before she retired.

rbacal




msg:3155861
 3:52 am on Nov 14, 2006 (gmt 0)

She said, "those pages tell you nothing"

I think they say a LOT, but you need to understand a lot of things before you can make sense of it. I used to try to explain some of this stuff here, but it's largely a waste of keystrokes.

If you want to improve your QS, and lower your bids, you need to understand things from google's point of view which they have explained quite clearly, at least so I can understand it. And that point of view isn't just about short term money.

But some people simply aren't going to be able to make this stuff work anymore, at least without starting over from scratch with a completely different approach to their websites, and a different understanding of what valuable websites consist of. If you're stuck thinking that thin affiliate sites are "valuable" to google, or even to the average searcher/clicker you probably can't get out of the box.

There's certainly enough information out there about how to create valuable sites, and to demonstrate value to google. I guess people have a choice. They can continue to be upset because their business approach (which was flawed in the first place) doesn't work anymore, and tweak a little or trick google.

Or, if they want to remain in the game, then the answer is to create content rich, original sites (ie. not affiliate feeds), limit dupe content, and so on. Be reputable, and an expert site.

Google certainly isn't perfect. Almost on a weekly basis they do things that are perplexing. But they aren't THAT perplexing on the QS score, now that it's been around since Summer.

Green_Grass




msg:3155862
 3:54 am on Nov 14, 2006 (gmt 0)

"Or, will you still evaluate inactive accounts? And at sometime in the future I could wake up to a functioning account again? "

Sailor.. Let the campaign stay active with a small daily budget.

It will take a while, but if you have made some changes as per their requirement Viz.. Detailed Contact details/Privacy Policy/ Samples/ Outward links etc , the quality bot will come and re evaluate the pages.

It may take upto a couple of months though.

Speaking from practical experience.

I just had an upward revision on one of my URL's in this update and the keywords sudenly became active again.. A nice surprise...

Green_Grass




msg:3155867
 3:59 am on Nov 14, 2006 (gmt 0)

"But some people simply aren't going to be able to make this stuff work anymore, at least without starting over from scratch with a completely different approach to their websites, and a different understanding of what valuable websites consist of."

rbacal not really right here..

For many webmasters, with lot of brand equity, starting from scratch is Not an option. Just make the changes Google recommends and wait a while.. It worked for me.. I really had no choice, except to wait and keep looking for new sources of traffic..

rbacal




msg:3155875
 4:15 am on Nov 14, 2006 (gmt 0)

For many webmasters, with lot of brand equity, starting from scratch is Not an option. Just make the changes Google recommends and wait a while.. It worked for me.. I really had no choice, except to wait and keep looking for new sources of traffic..

Good previous post btw. What I was referring to was, let's say a person who has good tech skills, sets up an affiliate site using feeds, etc, and has been making good money this way by bidding low on adwords, and doing well in the affiliate area.

THAT person sees his site as "quality", but that's NOT what google thinks is quality (and many surfers would agree). So google shuts off the cheap traffic, and the site itself is so weak that it can't generate "natural" traffic.

For someone who really has no content of value seperate from all other sites using the same content, all the adding privacy notices isn't going to be enough.

I 100% AGREE that if you follow the google advice, make an effort to understand the POINT of all this, and stop trying to make a buck with no other expertise except that you can buy a domain and set up a site, that the chances for success are good.

But for those who have 10,000 pages of affiliate feed dumps and nothing else, they don't have anything much to work with. They didn't create any permanent value before, and they don't have any now.

PS. There are a LOT of people (and I'll bet you are or will be one) that are making this all work. Some have offered suggestions (as you have) about how to make it work, but those suggestions often get dumped on by people who are simply not willing to recognize their clone sites were not so great before, and aren't adding value.

I think it would be just great if there could be a thread here that takes off from your post with suggestions, but every time the topic gets discussed, the threads get hijacked. A lot of people who ARE doing ok with QS simply aren't bothering to share what they've learned .

tonynoriega




msg:3156302
 4:00 pm on Nov 14, 2006 (gmt 0)

Here are my thoughts after reading all of this. Our site has been online since 96. Pretty old compared to most sites on the net....we are a solid company, multi million a year.

Google is now forcing us to design our site that will not only fit their AdWords standards, but at the same time that is forcing us to design to their "searching" standards as well. Its double fold for them....Design to their searcing algorithm, and the AdWords QS algorithm also....

Why?...well yeah, they are the most used SE on the net, so why not jump on their bandwagon and reap some benefits. But creating guidelines that only benefit their interests does not help everyone involved. The internet was based on sharing from all avenues of interest, this is not the GWW Google Wide Web....

So now, since my "Content Network" advertising has been disable, i have to re-create, or enhance my landing pages becuase if i dont, i get slapped on the wrist, and my managers are not happy.

Now it looks like it is my fault that the content my documentation dept. provided for the landing pages is not sufficient....

Screwed from both ends. Thanks Big "G"....OH, and my MAC CPC's...have all been forced to about $3.00 just to keep up.

i wish there was one week that everyone would stop their AdCampaigns and see how much money Google would lose....maybe then they would appreciate what we are doing for them, not vice versa....

luke175




msg:3156389
 5:05 pm on Nov 14, 2006 (gmt 0)

If Google really cared about quality and not just extorting its advertisers they would simply publish your quality score and tell you what was needed.

Why would this be bad?

Some people say that it's because the system could be gamed. Not so, I say.

If the requirements are as Google says (i.e. quality outbound links, relevant non-dup content, etc.) then even if they are out in the open it would STILL benefit visitors.

Would it be wrong to say, "hey affiliates, add a bunch of unique content to your site, some links to authorative sites, and be more transparent- you need a real privacy policy, etc." To "game" this system you still need to give visitors what Google wants you to give them.

Google won't ever give you transparency because they are extorting you. If they were to reveal exactly how "quality" decisions were made you would be more than a bit ticked off.

Google's stock price isn't exploding because advertisers are paying less folks.

aeiouy




msg:3156460
 6:26 pm on Nov 14, 2006 (gmt 0)

My personal opinion on the landing page part of the equation is relevancy is king, which is exactly what it says. I don't think they mean general relevancy or broad relevancy, but on point relevancy carried from the keyword directly to the ad text directly to the landing page. I have not necessarily seen a need for things like Privacy Policies, Links to Authority sites or anything of the like. Just very strong relevancy from my keyword, to my ad, to my landing page.

That means if you have a keyword grey widget, and you have an ad for widgets, and a landing page for awesome widgets, you are going to score poorly.

If you have a keyword for awesome grey widgets, and an ad for awesome grey widgets, and a landing page that is centrally focused on awesome grey widgets then you are much more likely to be okay.

Again this is my own experience with all of this, so that take that for what it is worth.

I just know I don't have any $5.00 or $10.00 keyword bids.

rbacal




msg:3156492
 6:48 pm on Nov 14, 2006 (gmt 0)

My personal opinion on the landing page part of the equation is relevancy is king, which is exactly what it says. I don't think they mean general relevancy or broad relevancy, but on point relevancy carried from the keyword directly to the ad text directly to the landing page. I have not necessarily seen a need for things like Privacy Policies, Links to Authority sites or anything of the like. Just very strong relevancy from my keyword, to my ad, to my landing page.

No, that's not how it works. Consider HOW google developed their algos for this, starting with the use of manual reviews from real people to profile relevant sites and not relevant sites. When you "get" this part, the rest makes perfect and obvious sense.

If you look at what they wrote recently, it's clear that they both are interested in and developed the algos to increase RELEVANCE TO VISITORS. That's what "relevance" means.

They started with real people, developed profiles based on their judgments, then algorized the process. That's WHY things like privacy policies, contacts, non-anonymous registrations, etc are involved, because those things ARE relevant to visitors, and contribute to the sense of reputability of a site.

That's not to say they didn't add in the variables you talk about -- they obviously had. So, the upshot is that NO SINGLE thing you do will protect you, since the whole thing is a result of a bunch of interacting variables. You change one it MAY work. Change a bunch ad it WILL work.

The easiest guide is to think about real regular folk visitors.

netmeg




msg:3156510
 7:16 pm on Nov 14, 2006 (gmt 0)

I dunno about that; only one of my client sites would fall under that (only one has a privacy policy, or anything other than general contact forms and information) and none of my sites have been hit more than a couple of cents, either this past summer or last week. I think my opinion is closer to aeiouy's on this.

rbacal




msg:3156514
 7:28 pm on Nov 14, 2006 (gmt 0)

I dunno about that; only one of my client sites would fall under that (only one has a privacy policy, or anything other than general contact forms and information) and none of my sites have been hit more than a couple of cents, either this past summer or last week. I think my opinion is closer to aeiouy's on this.

Multi interacting differentially weighted variables. Weights for each variable calculated dynamically depending on a number of other variables.

It makes algos exceedingly hard to game or manipulate, because you will NEVER see consistent outcomes, unless you have access to a huge amount of private google data.

One site will do x and it works. Another site will do x and it will do nothing. That's how multivariate stuff works.

fischermx




msg:3158213
 6:51 am on Nov 16, 2006 (gmt 0)


That means if you have a keyword grey widget, and you have an ad for widgets, and a landing page for awesome widgets, you are going to score poorly.

If you have a keyword for awesome grey widgets, and an ad for awesome grey widgets, and a landing page that is centrally focused on awesome grey widgets then you are much more likely to be okay.


Amen.
I'm one of those hit for the 5.00/10.00 clicks, that's for my main site.
However, I also ran a small affiliate campaign, which is, IMHO, perfectly disguised.
For this campaign, however, I have a min bid of $0.01 on the very main keyword! as well, I have many other keywords under $0.05.
And for this campaign I notice, I have exactly what you mentioned, my ad is very specific, as well as the keyword and the landing page as well, all fits perfectly.
My campaigns being hit by the crazy bids, however, are vague and broader.

aeiouy




msg:3159324
 1:35 am on Nov 17, 2006 (gmt 0)

Rbacal,

I have my own experience to go on, and I know what works and does not work for me. You may be over-crediting the QS landing page algorithm. It is not really in Google's interest to discourage people from improving their sites so they are not going to say one way or another.

Feel free to try a landing page that is not relevant with a privacy policy and see how that floats it. Personally I think there was very little scientific determination it what does and does not impact the QS of a landing page. People threw 25 things at the wall, and when they hit on a combination that worked, they assigned credit to all the various components.

Who actually tested a single change at a time to actually determine what made the difference?

I just know I see posts from people who have done the generic privacy policy tack on and the other things, and say they are still getting hit... and I want to know about relevancy, because I suspect that is a much bigger part of the equation. Especially since it is a huge part of the equation when judging the actual ad copy. It just reasons that it would also be of primary importance going to the landing page.

rbacal




msg:3159345
 2:04 am on Nov 17, 2006 (gmt 0)

Who actually tested a single change at a time to actually determine what made the difference?

Slamming head against monitor. It doesn't work that way and you can't intelligently "test" that way.

Multi interacting differentially weighted variables. Weights for each variable calculated dynamically depending on a number of other variables.

It makes algos exceedingly hard to game or manipulate, because you will NEVER see consistent outcomes, unless you have access to a huge amount of private google data.

One site will do x and it works. Another site will do x and it will do nothing. That's how multivariate stuff works.

If there's something I haven't been clear about above, and you don't understand, you could ask. But focusing on any ONE variable is pointless. As I've written elsewhere tweaking around the edges is a loser's game for anything but the very short haul.

heyday




msg:3159379
 3:06 am on Nov 17, 2006 (gmt 0)

Someone posted about that we should "Pause our campaigns for a day and see how Google likes it."

Wouldn't that be something.... unfortunatly thats like saying "Don't buy gas on Friday cause the prices are too high"

I would LOVE to see that happen. These guys are just too darn big and need a dose of humble pie in my opinion.

They should be bowing down to their advertisers and personaly helping those of us who have been decimated by this update.

A call to my rep and I'll I get is "Sorry to hear that" response.....

heyday

aeiouy




msg:3159898
 3:10 pm on Nov 17, 2006 (gmt 0)


Slamming head against monitor. It doesn't work that way and you can't intelligently "test" that way.

In fact that is the only intelligent way to test anything Rbacal.

Multi interacting differentially weighted variables. Weights for each variable calculated dynamically depending on a number of other variables.

Blah blah blah blah. For some reason you think Google's algo is unable to be tested. So everything is just a made up list of things. Sorry, but that is bad advice. When I have sites that don't have half the things people claim fix quality score, I don't buy that it is all necessary. Surely there can be a lot of factors, but some are obviously more important than others. I am saying that relevancy is more important. Relevancy is something that has been the core of Google's business for a while. Relevant ads to relevant keywords very important. Showing relevant ads to site pages on the content network etc. Relevancy has always been the core of their business.

Regardless, if you tell me there are 20 components to getting a good quality score, and I do one of those things and get the same result as someone who did all 20, who are you to say that I am wrong? It is silly to think you can't properly test individual components, even in a complex algorithm. Google has not said exactly what impacts quality score, so people are just doing a lot of guessing. Throwing 20 things at the wall and then claiming that is what impacts quality score, is not the least bit scientific.

It makes algos exceedingly hard to game or manipulate, because you will NEVER see consistent outcomes, unless you have access to a huge amount of private google data.

And I am telling you, you are way over blowing the complexity of the QS algorithm or ignoring the fact that some components are significantly more important in it then the others. I prove this to myself every single day.

One site will do x and it works. Another site will do x and it will do nothing. That's how multivariate stuff works.

Yeah but if I do x every time and it always works, I feel pretty comfortable in my position. I don't really care if anyone else gets their bids lower or not. Less competition for me. Yippee. I just think people are way over-analyzing and over-working the situation.

If there's something I haven't been clear about above, and you don't understand, you could ask. But focusing on any ONE variable is pointless. As I've written elsewhere tweaking around the edges is a loser's game for anything but the very short haul.

I don't have any jacked up bids on ANY of my campaigns. Different sites, different campaigns different industries etc etc etc.

Don't tell me what I do and don't know. You can claim that it is a waste of time, or the algo is some kind of evenly balanced equation, but that is not likely true.

What is funny to me is that people rarely even mention relevancy when they talk about this stuff. All I see is outgoing links, privacy policies and all these other things.... yet relevency does play an important role in it, and people choose to ignore it.

Do so at your own peril. I am not going to argue with you about this anymore. Like I said, I know what works for me. Other stuff works for other people. I don't care what anyone else does, to be quite honest.

Your advice, however, would be akin to telling people not to do split ad testing because human beings are too complex to detect any particular pattern in what ad might do better to you.

exmoorbeast




msg:3159994
 4:40 pm on Nov 17, 2006 (gmt 0)

Well I agree with what you are saying, but what exactly do you want to talk about with regards to relevancy. What if your ad copy is 100% relevant to the content of the landing page...what then? Is that the only measurement of relevancy, or are there other factors?

Surely there is a list of factors, and that each factor is weighted. We discussed one last night about the algo rewarding sites that held the user for longer than just a couple of seconds. Basically the hypothesis was that if people kept landing and coming 'back' to Goog, then the listing was deemed to be poor quality and thus irrelevant.

What other relevancy metrics can you think of?

rbacal




msg:3159998
 4:47 pm on Nov 17, 2006 (gmt 0)

Well I agree with what you are saying, but what exactly do you want to talk about with regards to relevancy. What if your ad copy is 100% relevant to the content of the landing page...what then? Is that the only measurement of relevancy, or are there other factors?

Good question. Google's use of the words quality and relevancy are a little unfortunate, since I don't think THEY mean what most people would think they mean.

If you look at what they've written, they talk about relevancy for the VISITOR, which is different than what we might think when we talk about the relevancy of the ad to the landing page.

...which makes figuring out possible metrics very hard, since they are using information about what is relevant and irrelevant to visitors that we don't have, even if we knew what they were.

I think part of relevancy to the visitor has to do with the apparent reputability and perceived extra value to the visitor. The word relevancy is a bad choice to describe this, but I think that's what google is saying, in part.

If you pages/site add value for the visitor (ie. something the person can't get elsewhere), then google calls it more relevant than a site that is just a feed dupe.

tonynoriega




msg:3160007
 5:01 pm on Nov 17, 2006 (gmt 0)

Relevancy is such a clouded word when it comes to AdWords...
Would it mean just plain ol' text information that is relevant for the user to read?

Relevant images for the user to view? Relevant links for the user to find more relevant information? Relevant H1 tags, bold tags...etc.?

so like i mentioned in another thread, im revamping my pages, but is it going to be worth it?...will my QS even get a raise if i do what is right?

and how long will that take?...i hear months before you QS will even get re-evaluated....

im telling you...im going to start something similar to the "walk outs" we used to do in high school.... walk out on Google for just one day...

fischermx




msg:3160018
 5:12 pm on Nov 17, 2006 (gmt 0)

and how long will that take?...i hear months before you QS will even get re-evaluated....


I've been tweaking my pages, one a time, and I can tell I see they re-evaluate the QS approximatively each week or less.

exmoorbeast




msg:3160032
 5:28 pm on Nov 17, 2006 (gmt 0)

If you look at what they've written, they talk about relevancy for the VISITOR, which is different than what we might think when we talk about the relevancy of the ad to the landing page.

Hang on a minute...this is a robot we are talking about here. Adbot comes from the original googlebot family that, when first born many moons ago, was only able to read <Title> and this makes me think that Goog simply wants webmasters to have relevant sites as far as ad copy, landing page copy is concerned. Seriously, how can a bot determine if a user has had a good experience when on that site. How can presenting duplicate content automatically mean the site is awful?

Your idea of a good experience may be different than mine, Larry's and the bot itself. It's a very subjective score you are talking about here. I wouldn't like to argue with you about what site adds more value to the web - Ebay, CNN, or Mom & Pops' Party Store. IMO, every site that can consistently advertise day in day out must have some value, and it would be extremely hard for Google to automatically determine otherwise.

Maybe the old adage of 'keeping things simple' comes into play here. Ie using the naming of images, text, page titles consistent with the search query. Seems to be working for me anyway!

This 107 message thread spans 4 pages: 107 ( [1] 2 3 4 > >
Global Options:
 top home search open messages active posts  
 

Home / Forums Index / Google / Google AdWords
rss feed

All trademarks and copyrights held by respective owners. Member comments are owned by the poster.
Home ¦ Free Tools ¦ Terms of Service ¦ Privacy Policy ¦ Report Problem ¦ About ¦ Library ¦ Newsletter
WebmasterWorld is a Developer Shed Community owned by Jim Boykin.
© Webmaster World 1996-2014 all rights reserved