Forum Moderators: martinibuster
NS
1) External distractions such as wars, royal weddings, Super Bowl games, holidays, and weather patterns; and...
2) How many entrepreneurs and PPC managers are bidding on AdWords at any given moment, which will determine prices on any day or even at a given time of day.
My own average clickthrough rates and effective CPMs have seesawed over the 3-1/2 weeks that I've been in the program, but the overall trend has been upward.
As far as the "novelty" issue goes, I think that may be more of an issue on some sites than on others. On a site like mine, where there's a lot of reader turnover and most readers are researching ways to spend their money (i.e., planning vacations), the novelty issue really doesn't come into play. If John Doe is flying to Paris for the weekend and needs a hotel, the fact that he's seen AdSense ads before won't keep him from looking at (and clicking on) relevant ads.
However, on a community-based site where the same readers come back every day, familiarity may reduce the incidence of casual clicks by curious users. If that happens, it's a good thing, because poor conversion rates aren't good for advertisers (and, by extension, for AdSense and the publishers who derive income from AdSense).
As far as the "novelty" issue goes, I think that may be more of an issue on some sites than on others. On a site like mine, where there's a lot of reader turnover and most readers are researching ways to spend their money (i.e., planning vacations), the novelty issue really doesn't come into play.
I have to agree with Europeforvisitors, it all depends on the site. If for example, you have a site with high turnover where visitors are looking for something specific be it finding out about a a vacation spot for that once a year vacation or looking for financing for their business or an investment {my site} then targetted adsense ads are great. In fact, I find that most people only browse a couple of pages of my site {average is like 2.2 pages} and barely 10% of them bookmark it or come back a second time.
Regarding earnings, they fluctuate wildly, yesterday I had one of my best days, but so far today with 33% of yesterday's clicks I am making 8% of the money. Then again my site is so small, I would need 10 times the traffic to really tell trends.
If i was them i would be looking at the sites (including types of sites) that are returning best CTR and CPC, and also getting smarter and trying to work out which sites it works best on and make the most money for adsense by increasing advertiser ROI and ad effectiveness.
I would not be surprised at all if what they are paying publishers is far more than a simple percentage. We have had a few 3c clicks and i think the min spend is 5c, so if I'm right that means they are NOT paying 100%.
It may well be that what they pay publishers for impressions takes into count things like the 1st, 2nd, or 3rd click from the same IP, and maybe even a sliding scale per amount paid by advertiser.. (more for lower CPC as a percentage, and less as a percentage for the larger bids for example)
In fact they can do anything like as they have not made any promises anywhere! They have the freedom to test away. All they have to do is to provide publishers enough incentive to use adsense rather than competing revenue generators, and adwords advertisers enough ROI on content ads for them to NOT stop advertising in content sites.
That may explain some of the substantial differentials we are seeing day to day. They are testing as they go with diff algos.
I would not be surprised at all if what they are paying publishers is far more than a simple percentage.
That's what I've been thinking, too. There are only two reasons why an ad network would use a simple revenue split like 50-50 or 65-35:
1) It's easy for the ad network.
2) It's an easy concept to sell.
Unlike most ad networks, Google is a company that was built on complex algorithms. Even its AdWords rankings are determined by an algorithm that goes beyond the usual bid-for-placement scheme. Why shouldn't its revenue sharing work the same way? At its simplest, the formula might be like a traditional tiered royalty or commission structure where the top performers get a bigger piece of the pie.
Such a formula is tougher to explain than a simple 50-50 or 65-35 split, of course, so Google doesn't bother: It lets the revenue speak for itself. And for most of us, the revenue that we earn from AdSense is far better than we've ever earned from a conventional ad network.
Side note: I'd like to see Google display "RPM" (revenue per 1,000 impressions) and "Average CPC rate" along with the other numbers in the daily report. Those numbers are easy enough to figure out, but why should we have to do the math?
That's the beauty of Excel spreadsheets ;) Killroy has an excellent example that he's letting others use, it works well. But I agree, it would be nice if G offered those stats as well, so we didn't have to crunch the numbers so to speak. As well, it would be nice to know how much you make per unique visitor....
Dave.
I have no problem with this personally and probably see this "flexibility" as a strong point in favor of their policy of not stating their commission structure. Could you imagine the web publishers jumping up and down complaining "unfairness" if Google started publishing how they worked our commissions. For people who want to know exactly how their ad network or affiliate pays out, there are many, many alternatives. Amazon for example uses the publishing of their commission structure to regularly spin new branding for themselves, and encourage many options for affiliates to advertise Amazon!
We here are bit like Pavlov's dogs right now, reacting to various stimuli, with no doubt a Google employee laughing their socks off reading what we are making of it all (a big hi to that Google employee ;) )
It is possible to get closer to Google's algo by opening and monitoring multiple accounts...however I agree that they appear to be altering the payouts (as you say, it could be that it is not they who are altering the percentages, but that the advertisers budgets/bids are altering quickly over time)
My conclusion is that one waits till it settles down before trying to get closer to what is really happening with the algo for payment, and in the meantime maximise one's sites where seems most appropriate.
This non-publishing of their commission or payout structures/alogos also means that they can "incentivise" certain sites more than others.
Exactly. For example, they could offer slightly higher payouts for:
1) Sites in categories where publishers have other revenue options (such as affiliate programs).
2) Sites in categories where there's a great need for ad inventory.
3) Sites that demonstrated shown higher-than-average revenue potential.
It's worth noting that affiliate programs often make special deals with high-revenue or even high-potential Web sites. In some cases, an affiliate program may pay double the normal commission if it really wants a publisher as an affiliate. There wouldn't be anything odd about Google doing the same thing with AdSense--but by using an algorithm to determine the revenue split instead of relying on human decisions or negotiations.
Dave
Sometimes, the same keyword can fluxuate by 66% in value throughout the day. Without fail, it comes back relativly soon.
All in all, it seems pretty steady overall to me.
If a $1 click only brings $0.20 commission compared to $0.03 for a $0.05 click It would still be better to have the expensive ads, but hardly worth remodelling your website for.
SN
If a $1 click only brings $0.20 commission compared to $0.03 for a $0.05 click It would still be better to have the expensive ads, but hardly worth remodelling your website for.
It wouldn't be smart for them to have such a commission structure, because:
1) Publishers would quickly figure out what was happening, and...
2) The publishers that AdSense needs most (those with high-revenue topics) wouldn't stand for having to subsidize their less profitable peers.
If anything, a variable or tiered commission formula would almost certainly be skewed the other way as an incentive for the most profitable publishers to stay with the program.
