Part of the problem with understanding the Google Adwords Quality Score updates is most people come from the point of view that Google is firmly interested in full disclosure and telling you the truth. While I’m not calling anyone at Google a liar, I will say they are premeditatedly selective about what they chose to reveal, and what misinformation they choose to correct, and what they are happy to let spread like wildfire.
Would you like an example, say I own a gardening store and you come in to buy a plant. I tell you it’s important to make sure you give the plant lots of water. You go home plant it in your backyard and make sure it’s getting plenty of water. Unless you’re very careful though chances are you’ll over water it and kill it. Why, because I didn’t tell you the reason it needs lots of water because it really needs to be planted in direct sunlight. So from the information I told you, you naturally assumed that the moisture level was the dominant factor, when in fact it’s a secondary dependant factor, based on the higher level factor of the amount of direct sunlight. Did I lie to you, technically no, as water is clearly an important part of the equation that lies along the critical path, however you assumed that I was interested in making you aware of all the factors involved. Why would I omit this information? Maybe I want you to keep coming back and buying more plants. Maybe I want to deal with professional landscapers who have bigger budgets, don’t come in bargain hunting, and already know what plants need lots of sunlight from those who don’t. Maybe I don’t really want you as customer anymore and am going after a different market.
OK Gray I didn’t choose Botany for $500 what does this have to do with Adwords. Well IMHO the recommendations Google gives are the equivalent of “make sure your plant has enough water”. While they aren’t lie they are all dependent factors, not primary ones.
Another interesting thing to note is that this third quality score update follows a lot of Google’s similar patterns. Third, WTF G this is the second have you started drinking for pubcon early? Well see the first update was quiet since nobody had their prices raised for “poor landing page quality” back in December of 2005. The real chater came in July of 2006 when prices got raised and some people saw a dramatic drop in profits. Which brings us to November 2006 when another quality score update occurred.
Hopefully in presenting things in a chronological order like that I got a few of you hunting down the right path. Can you think of any other things that get updated with several month intervals? If you said toolbar pagerank go to the back of class, don’t you know real page rank and external toolbar pagerank are completely different. I’m thinking more about the sandbox/trustbox update that happens every few months. Now c’mon Gray that’s just crazy talk are you proposing that Adwords takes into account organic factors? I mean it’s not like Adwords shares data with Adsense or anything, right?
So what do you do to get around being shafted by the Adwords Quality Score Update? To borrow a line from a great movie “the Matrix is just like a computer system: some of its rules can be bent; others can be broken”. Think of what an engineer would really consider a signal of quality, not what they actually tell you they are looking for. Find a way to give it to them, be creative, follow that crazy idea, test it to see if it works.
Update: For anyone who think I’ve got my tin foil hat firmly placed on my conspiracy theorist head, I’ll remind you it took Google over 2 years before they admitted that there wasn’t a sandbox, but there was something that could possibly be interpreted as having sandbox like qualities.
Related posts:
- Google Adwords Crossing The Streams So I was l
- Using Google Checkout Buttons Doesn’t Increase Adwords CTR If you do










{ 11 comments }
LOL at the last paragraph. Google should be a politician.
> Now c’mon Gray that’s just crazy talk are you proposing that Adwords takes into account organic factors?
Interesting point, Michael. I had never thought of that in conjunction with my landing pages. I always assumed that Google limited itself to judging based on the copy on the page, etc. I hadn’t thought about the possibility that trustrank, link data etc. might all be used in conjunction with the quality score.
Good Post!
You might have a point. However…
Many people, especially large advertisers use dozens of specialised landing pages that are not indexed. Shopping.com spends a small fortune on PPC to drive people to a page that is dynamically generated on the fly based on the referring query.
The main GoogleBot would never “see” these pages. You might test using your home page versus an orphan page with no links as a PPC landing page.
Do you have any thoughts on the proposed “fumflockerkin†test QuadsZilla posted about? Instead of one person doing the test, 10 people might be able to participate and share results.
Ah see Dave you stepped into the trap they laid out for you,and assumed what they wanted you to assume. I’m not suggesting that they score each individual page. I’m saying your domain has a master landing page score.
I’m totally liking the fumflockerkin idea.
Bingo. The days of “we don’t rank sites, we rank pages…” is long gone as absolute and an injurous thing to assume. The domain and your google account(s) carry metadata in them about quality.
fumflockerkin away!
Hmmm, you are mentioning an interesting point. I am assuming that links and the general site authority\trust could be part of the quality score since the july update. But I am not sure about this.
Put it this way – what possible reason would Google have *not* to use the trust data they use for organic SERPs? Surely it is a case of when, not if.
Personally I wouldn’t be surprised if Google were to use organic factors (ie, domain authority) in quality scoring factors.
Actually, the surprise would be for it never to be a factor. Google has the data, so why not ensure the left hand knows what the right hand is doing?
The question would be the degree to which organic factors may have a role to play. For example, a new mom-and-pop site won’t have much in (link-based) organic factors to recommend it – but it would be absurd from an advertising point of view to freeze them out. So on-page factors could be a basic reference point here.
I’m not saying Google are using organic ranking factors for Adwords – but since they first announced quality factors, it seemed to make sense that some degree of on-page and off-page SEO factors could be factored in.
Simply 2c.
And following that thought pattern what happens when a domain has been around and in the system say 6 months or longer and it has “negative” score instead of a positive or neutral one …
Since the July update, I’ve played with putting landing pages on ‘aggresive’ MFA sites that are no longer indexed but are 1+ years old.
Comparing the same ppc keywords with substantially the same landing page on sites that are indexed
oops, fat fingered my response. Here is what I saw (note that this isn’t a scientific test but more anecdotal observations)
- with new adwords accounts, the non-MFA site had a less expense CPC by a factor of 11.
- with new accounts in a market that requires approval by Google there was no significant difference over a 3 week period.
- with an old account I did significant direct-linking (google cashing) with to CPA or affiliates bids were higher on both clean and MFA sites compared to an established account that didn’t have any direct linking in its history.
- deleting MFA content, making a 5 page minisite and submitting a google sitemap impacted the CPC costs. But the ‘old’ MFA site was still 6x more expensive initially than a ‘clean’ domain. In three and a half weeks I was able to drive bid prices down so there wasn’t a significant difference.
My spend in all these little tests was small – each campaign was generating only 200-300 clicks a day.
I tended to believe from this that domain quality is one of the adwords quality scores. There are too many variables to say this is in any way definitive then or now, but following the logic I think that more and more of the historical data G has will be connected and applied.
Comments on this entry are closed.