Forum Moderators: martinibuster
So what does determine payout? Is there a sliding scale based on revenues? Does Google favor certain types of content over others to promote the kind of Web that it would like to see? We don't know, but here's something to think about: Wouldn't it make sense for Google to use internal referrals vs. referrals from search as a factor in computing payout?
Let's look at two hypothetical sites:
Site A is a large "content site" with a high percentage of internal referrals. It receives a significant amount of traffic from Google and other search engines, but users typically stick around and view half a dozen pages, making internal referrals a bigger source of traffic than the search engines.
Site B is a search-driven "made-for-AdSense" site that has very little original content on its pages. It consists mostly of keyword-optimized pages with minimal content and the maximum number of AdSense units. The site gets nearly all of its traffic from search.
Which site provides the greatest "added value" for Google?
Site B gives Google and the advertiser a second chance to reach search users who didn't click on an AdWord, but that's all. In essence, it's nothing more than a parasitic offspring of the search results, and most of the value (in the form of prospects) is being supplied by Google.
Site A also provides that second chance to reach search users who didn't click on an AdWord, but in addition, it gives Google and the advertiser a chance to reach prospects who were already on Site A (e.g., repeat visitors and users who showed up via links from third-party sites). In other words, Site A--unlike Site B--is creating value for Google and the advertiser, not just extracting value from the search engine.
Common sense would suggest that (greater added value) = (higher payout). And if "added value" is a factor in Google's payout formula, that would go a long way toward explaining why a heavily optimized made-for-AdSense site might see its earnings drop even though it's sending qualified clicks to Google's advertisers.
DISCLAIMER: The above is mere speculation (a "what if" scenario, if you will), so please don't assume that I'm presenting it as fact.
As an analogy, for yellow pages, the user is likely to be only interested in the product he is searching , which results in higher coonversion
Yes, but I'm not talking about conversions, I'm talking about "added value." From Google's point of view, wouldn't it make sense to pay less for clicks from users whose visits originated on a Google search page?
Then Google should be paying more for traffic from Yahoo, MSN , mailing lists, etc. (all non-Google sources)
But Google isn't getting traffic from those sources. It's getting traffic from AdSense publishers.
However the ads shown on Google's page are based on the specific search text entered, while the AdSense ads shown on the destination page are contextual and most likely will be different. So I'd call the website more of a partner in determining the best ads to be displayed, rather than a mere parasite.
Good point, but it would seem to me that the real issue here is how much value the publisher vs. Google is bringing to the party. If Google is supplying the publisher with most of its traffic, then wouldn't Google be entitled to a larger share of the revenues?
[edited by: europeforvisitors at 5:12 pm (utc) on Feb. 9, 2006]
Although the amount of page views per visitor is extremely low, I see no evidence of smart pricing. EPC is constant since I started with AdSense and prices per click roughly match the bid prices I know for the specific niches for these sites.
Combining this experience and EFV's hypothesis I would slightly change the hypothesis in such a way that not the amount of internal referrals, but the average time between the first ad impression and the actual click on an ad might be a factor in smart pricing.
Both on my type of sites with large content pages, and on sites with smaller pages but more internal referrals, the amount of time between the first ad impression and the moment a visitor leaves by clicking on an ad would be higher than on an MFA site, indicating that the visitor had more time to read the content and therefore have a higher chance of being a converting customer for the contextual advertiser.
Combining this experience and EFV's hypothesis I would slightly change the hypothesis in such a way that not the amount of internal referrals, but the average time between the first ad impression and the actual click on an ad might be a factor in smart pricing.
Lammert, you make some good points, but--just for the record--I'm talking about payout percentage, not smart pricing.
Smart pricing = discounts to advertiser, based on anticipated likelihood of conversion.
Payout percentage = the Google/publisher split of actual revenue (i.e., revenue after any or no smart-pricing discounts).
Consider the Google AdSense Search product: Google appears to be paying less for search clicks than it does for content clicks, and it reserves the right to charge a monthly fee to publishers who have a low ratio of search revenue to search traffic. Why? Obviously because there's a cost associated with providing search, and most of the value of AdSense Search is being supplied by Google.
What I'm saying is that, if a higher-than-average percentage of an AdSense publisher's impressions are being generated by Google Search, Google is providing a greater share of the value than would be the case if the publisher had a large number of internal referrals. So it wouldn't be unreasonable for Google to keep a bigger slice of that publisher's AdSense revenues.
Again, I'm not claiming that Google actually does this; I'm merely suggesting that it could be one factor (other than smart pricing) that determines how much a publisher earns from AdSense.
So the visitor likely did not find his widget information even after he surfed a lot around the site because he finally clicked ads about widgets. Which means site is MFA or bad site about widgets and doesn't deserve to rank for widgets in SERPs.
By the way, "More non-Google referrals - higher payouts" would make sense as you are transferring visitors from Y!, MSN, etc to Google empire ;)
