Forum Moderators: martinibuster
So what determines whether some third party mailicious activity gets you thrown out or not? I believe it's the quality of your account; that it's about how much Google values your account. They may not have a PR ranking system in place for publisher "honesty" but it sounds logical that they are less likely to throw some big partner like CNN out for one fraudulent click when they may well do that for a Joe Blogg who generates only $1 a day.
So what makes for a site that Google values and is less likely to penalise? I've no idea but I'll propose some possibilities:
1. Size of earnings: If you generate a large amount of revenue from the account Google earns a lot too.
2. Size of traffic: If you generate a large volume of traffic spread out to visitors from across the globe Google may be more lenient of a single user multiple clicking your ad from somewhere in China.
3. The subject area: If you are in a non-competitive subject area and there aren't a lot of other websites that can attract the ads that your site does then they may look more favourably on the odd mistake
4. Effort required to watch you: If you've got a site that doesn't generate a lot of monitoring work they may be more tolerant. Forums may lose out a bit here.
5. Quality of your site/network: If your site has a lot of quality content and doesn't appears like a made for Adsense page/site/network then they're likely to be more tolerant.
6. History: If you've been with them a long time and you have a clean record they're more likely to put some fraudulent clicking down to malicious activity.
7. Optimisation: If you don't seem to have too many SEO optimisations (they see it as trying to game the SEs) they may be more kind
(some of these may not take precedence over point #1 above ;)).
So, if you want to make your site more valuable to Adsense what do you do? I've no idea but if a site's value to Google influences how strict they are on apparently fraudulent clicks then it's in your interest to improve your site's value to the program. And it's possible to do that. Even size of earnings, size of traffic can be easily increased in more ways than one (Buy a site or two if need be, and add it to your account).
size of traffic Yes
Very important especially if not in highly optimised high paying keywords
Subject Area Yes
As you say must be some areas where G is looking for more advertisers and publishers
Effort to manage Yes
may also be areas where same users 90% of time
very sticky site / forum / blog etc.
therefore webmaster may directly or indirectly encourage clicks
Site Quality YES
Possibly the most important due to the long term impact on advertisers perceptions of Context advertising and possibly negative press at some stage if becomes to rampant
History Yes
Heavy handed SEO doorway any blackhat
most likely to incur email fits with site quality issues and advertiser perceptions
steve
I would think advertisers would want to be seen in clean sites that rank well, if not the top, of the Google serps. So even if the user doesn't click on the ads on the serps and instead clicks on the top ranked sites, then the advertisers will still have the chance to be seen by their target audience.
What impact do you guys think a site ranking well in their search engines actually have on the site's value to Adsense?
None, except for the fact that good rankings pay off in higher traffic and revenues.
Why would Google want to favor sites with high rankings? That would just encourage SEO spam (not that it needs much encouragement!).
I said the very same thing in another thread and had the notion dismissed by all. But it DOES make sense for google to look favorably on sites that are more RELEVANT and pay them accordingly, because sites that are relevant and show up in serps are closer to what the user was looking for and are more likely to result in ROI/conversion for the advertiser. And if the google algo does its job, then relevant sites WILL (along with a hefty sample of spam, of course) get to the top of serps.
But how does the smart pricing algo determine relevancy for the purpose of rewarding adsense publishers? By how well a site does in serps? I don't think so. Serps are simply the outcome of an automated algorithmn that few human hands ever intervene in. Really, the only objective relevancy-related criteria that google could use to influence payouts to publishers would be:
Pagerank
Incidence of incoming links from UNIQUE IP's (many unique links from many IP's versus 10,000 inbounds from one IP).
Matching of anchor text of inbound links to onpage content.
So, I really wonder if the smartpricing algo uses relevancy and does this by borrowing from the SE algo to determine payouts
Page rank can't be an indicator of relevance either. It's an indicator of number of incoming links. The "copyright" page of a PR8 adult site isn't a page relevant to copyright law.
Smart pricing only makes sense if it's based on conversions. That's the only metric that ultimately matters to the advertisers (ROI).
Some may not have ways of tracking this, but if you are involved in affiliate programs along with Adsense, you'll know if your site is capable of sending sales to merchants.
A site that has a strong conversion ratio ALONG WITH the other factors that Macro posted, has a better chance of staying with Adsense through any bumps.
One of the cancelled accounts claims to be earning $6000 a month so that must be ruled out.
Do you believe everything you read on a message board? The person who posted that was a first-time poster. That is quite a first post. If one of the regular contributers posted that, I would truly be shocked.
So Macro's statement stands: the size of the earnings matters when it comes to shielding you from the arbitrary boot.
Bear in mind also that if you operate in a "high value keyword" area you are more likely to attract fraudulent clicks from advertisers trying to nobble their competitors. It does happen.
I've long thought these folks are big targets, too. I wouldn't touch a hosting or webmaster site for anything.
I'm not particularly worried about being kicked out of Adsense because my sites are clean, helpful to the consumer, and they exist in areas where there are tons of retailers and little good content. And they're in niches that click-happy punks would probably never think to target. Despite that, they're lucrative enough to make a living from.
I think it all comes down to quality.
Quality=
Design
Usability
Organic Traffic
Conversion
Being Professional
& so much more ; )
Work hard and become a partner that Google will never want to stop paying. I am sure that we all know that birds of a feather flock together.
Be good to Google and they will be very good to you!
I love ADSENSE : )
My point was this: Assume I'm earning Google $1 per day and the Washington Post is earning Google $100,000 per day.... and a clever clickbot tries some funny business to the tune of $100 a day. If that happened on my account Google would think it's me (whether it was me or not). If it happened on the Washington Post account they are more likely to think that it's an advertiser targetting his competition. In the first case a termination is a good move, in the second case it's not. So, directly or indirectly, a high turnover figure may provide some protection.
Assume I'm earning Google $1 per day and the Washington Post is earning Google $100,000 per day....
That example is a bit flawed because the Washington Post, aside from being a Premium Publisher, is also a very highly respected newspaper.
So that doesn't really compare small earnings to big earnings, since there are other factors in that example.
If you had unknown publisher X earning $1 and unknown publisher Y earning $9000, I haven't seen any evidence that shows that Google favors one of the other based on dollar amount.
I see where you are coming from. But...
1. Size itself is likely to bring the reputation.
2. Size has the ability to forgive some fraudulent clicks. 0.00001% is more likely to pass muster than 10.0% and that is irrespective of whether Google takes offline reputation into account.
So substitute the Washington Post for any other large site and the argument still holds. AFAIK, there is no evidence that Google takes offline reputation into account. If number crunching is a major part of the algo's work then size of site/volume will come into play. It would make sense for the algo to have a fraud trigger set in percentages rather than raw clicks.
Anyone who believes there is a level playing field is, in my opinion, naive. It is not the way the world works.
Smart pricing only makes sense if it's based on conversions. That's the only metric that ultimately matters to the advertisers (ROI).
Then how do they measure this? Do they actually have a way to track each click to see if it converts? I can't imagine that. Are they testing a few sites for conversions then designing an algo that will help them estimate which pages will convert? Any ideas?
Then how do they measure this? Do they actually have a way to track each click to see if it converts? I can't imagine that. Are they testing a few sites for conversions then designing an algo that will help them estimate which pages will convert? Any ideas?
See Google's AdWords Conversion Tracking FAQ at:
[adwords.google.com...]
Note that "conversion" can be a synonym for "lead tracking"--it doesn't have to mean an e-commerce transaction. A travel agency that specializes in luxury cruises, for example, would hardly ever sell a $10,000 cruise through an e-form and shopping cart. The same would be true of a corporation that's selling fireproofing material for skyscrapers, silos for dairy farmers, or engineered plastic compounds for injection molders. Indeed, such leads are likely to justify far higher bids per click than any e-commerce transaction could, which is why Google needs to maintain the credibility of the AdSense network and keep it attractive to mainstream corporate advertisers and ad agencies.
Google is in a great position to define "quality visitors" in many ways. Some it can identify to the PC level (cookies), some to the ISP proxy. It could define "quality" from the conversion stats of advertisers who trust Google with that information. If these visitors click an ad, then buy, it makes them quality visitors to Google (and the advertiser). It would be too complicated to track at the individual PC level, but such information could be used to colour the IP address range as being a quality range or dodgy range.
Other quality visitors could be IP ranges that show zero to very low level of invalid clicks. These could be countries or other logical groups.
Grouping publisher sites, Google could treat sites with low clickthroughs as "low value sites" -- forums are believed to be in this category because the visitors tend to be blind to ads.
Then, if the Adsense publisher site comes up for the live-or-die judgment day, Google could apply the above measures to see what emerges. A low value publisher site that gets clicks from low quality visitors could die. A high quality site that generates a trivial amount of revenue from clicks but gets dodgy visitors and clickbots would also die, because the time needed to examine closely would cost more than the revenue it generated.
Next would be the high-revenue sites that had a high number of dodgy clicks - they get the chop mainly because of the invalid clicks
At the other end we have the major newspaper sites or those of impeccable integrity that have a whole mix of visitors but the pros far outweigh the cons, so these would be most unlikely to lose their AdSense accounts.
And, as Rodney says earlier:
A site that has a strong conversion ratio ...has a better chance of staying with Adsense through any bumps.
It is in Google's interest to improve the way they track conversions. Sure, not all products can be easily tracked (high ticket items for example are often bought over the phone) but it's in Google's interest to reward those sites that send sales.
Bear in mind that in addition to developing better systems of tracking they could give advertisers the option to block URLs. That would likely make scraper sites less profitable and benefit those sites that really work for advertisers.