Forum Moderators: martinibuster
Niz
You don't agree? Why not?
So a site in one market may be doing very well with a 1% click through and a site in another doing very badly with a 2% click through.
It would seem strange to reward a site for performing badly against others in its sector and to penalise a site in another sector even if it is doing well in comparison with other sites in its sector.
What would be the point of offering a higher payout percentage linked to higher CTR and not tell anyone?
If they wanted to encourage people to get a higher CTR then they should at least promote the fact that this is what they want to do.
But, encouraging people to get a higher CTR in my mind is dangerous and would encourage spamming techniques and not quality content. The last thing Google want is to get AdSense a bad name.
If a site generates a 10% clickthrough rate, I figure its content layout or ad placement is more effective than a site with a 1% clickthrough rate. Google spends less resources generating a higher click through rate - it make sense to me to provide a better payout for the higher percentage. You don't agree? Why not?
I agree with Dillonstar's comment that paying more to sites with high CTRs would just encourage spam.
Google's long-term success depends on giving advertisers what they need--and advertisers are looking for high conversion rates or high-quality leads, not just high clickthrough rates (especially when they're paying money for every click). Indeed, AdSense's biggest weakness may be the fact that it targets only by keyword and not by audience. This is likely to create an opening for competitors who offer targeted text ads on carefully vetted or handpicked niche sites (unless, of course, Google plugs this hole by giving advertisers more control over where their ads appear).
I don't know if this is something Google will carry out later on. It will be interesting to see.
It also depends on where they get their traffic from, pure SE traffic provides higher conversions however
as an advertiser i would prefer those sites with low CTR , eg 1%. 2 reasons.
Less likely of SPAM and shows site is not optimised for advertising.
If 1 out of 100 CTR then advertiser thinks they are getting targetted traffic.
I think this to some extent is practiced by the banner ads network like Burst. Sites with low click through rates are optimized out of the campaign, and only publisher sites with a CTR acceptable to the advertiser continues to run the campaign. The idea is to allow advertisers to show their ads only to sites that "perform".
That makes sense for CPM campaigns where advertisers are paying for impressions and don't want to buy "waste circulation." But AdSense is a CPC product, so cost per impression isn't a factor.
If the ad networks could do optimization, I'm sure the PhDs of Google could do this too. It would be interesting if there will be a reward system of sorts, wherein only publishers capable of pulling better CTR will be shown the higher paying ads. And those who are not able to show high CTR will be shown only the poor paying ads.
In the same way with Adwords -- if your ad could generate higher click throughs, then you move up higher in the listing and up the search results page. If not, then your ad slides down.
Of course, there's the issue of greater incentive to spam -- which is all the more reason for the fraudulent clicks algorithm to be made stricter. Spam the system to get higher CTR and you'll be booted out of the program.
I'm not saying this could happen. Just interesting to see IF this will happen.
The Optimization being done by banner ad networks is for BOTH CPM and CPC campaigns. Not just CPM.
It makes sense for banner ad networks to optimize for CPC campaigns because, if they don't, advertisers will take advantage of them by running "branding ads" that are designed to communicate a message without inviting a response. That isn't the case with AdSense ads, which are harder to use for branding purposes alone because (a) Google rewards higher advertiser CTRS with higher position to discourage such use, and (b) Google has greater control over ad content than a banner-ad network does.
If the ad networks could do optimization, I'm sure the PhDs of Google could do this too. It would be interesting if there will be a reward system of sorts, wherein only publishers capable of pulling better CTR will be shown the higher paying ads. And those who are not able to show high CTR will be shown only the poor paying ads.
They certainly could do it, but they'd be stupid to take that approach. Why? Not just because it would create an incentive for spammers (and put pressure on non-spammers to spam), but also because it would make the program less attractive to information sites, which provide credibility for the network and (in the higher-profit categories) have alternative sources of revenue.
If the idea behind AdSense would be to attract and reward sites with high CTRs, the ads wouldn't be called "content ads" and Google wouldn't be using the term "publishers." Instead, Google would be marketing AdSense as an add-on revenue source for e-commerce sites and affiliate sites--and in doing so, Google would be turning its back on mainstream advertisers and their Madison Avenue agencies.
If Google were to implement a sliding scale of payouts (or if it's already done so), logic would suggest that the scale be tied to revenues--not to clickthrough rates. Why? Because, for example, $1,000 in clicks on chocolate ads at Chocoholic-and-proud-of-it.org are likely to be worth more to those chocolate advertisers than $1,000 in clicks at Arnies-autogenerated-directory-pages.com/chocolate--even if it takes more impressions to get there. To put it another way, the most effective targeting involves more than targeting by keyword: It means reaching a targeted audience. That's why Google uses language such as "content ads" and "publishers."
[edited by: europeforvisitors at 8:35 pm (utc) on Jan. 7, 2004]
Instead, banner ad networks used optimization to attract more advertisers -- by promising through this feature that their ad will be served only to publisher sites that are performing. This in turn helps the publishers: by ensuring that more advertisers are added to the pool in order to give publishers the chance to earn higher revenues.
Adsense, first and foremost, is a program designed to make advertising at Google more attractive to advertisers by giving advertisers a pool of web publishers targeted to their ads. Adsense earnings to publishers come from advertisers -- so the primary goal of Google is to please the advertisers. If optimization is a way to do that, then they can consider that (if not already being done).
Of course, how publishers would react to it is an entirely different ball game. Even with banner ad networks, you can hear grumbling when PUBLISHER A gets a better-paying campaign for 1 week, only to be optimize later out of the campaign for poor CTR, then hears that PUBLISHERS B to Z saw a tremendous increase in revenues due to the campaign.
I am not advocating this system. Just following OP's question by including this "what if" scenario. But it is good for Google to see -- in the event that they are thinking about this approach -- that publishers are willing to revolt against this approach if implemented (or known to be implemented).
When users come to his site, they'll see AdSense travel-agent ads for Platinum Cruises, and they'll probably click on them. Why? Because the site has little or no information--and it certainly doesn't have the information that a Platinum Cruise costs $700 a day and up. So Sam will have a great CTR, but those clicks won't translate into high-quality leads or conversion rates for advertisers.
The advertisers will have much better success with leads generated on sites like Cruise Critic and Cruise Diva where readers are more likely to (a) know something about Platinum Cruises or (b) read an article about Platinum Cruises and its fares before they click.
That's why it doesn't make sense for Google to reward high CTRs for their own sake. High CTRs might be good for Google in the short term, but they aren't necessarily good for advertisers--and in the long run, Google's success is dependent on the quality of the traffic that it sends to advertisers.
If Google is looking at the approach of optimization based on CTR, then it must first make sure that it does not accept sites that are designed solely to generate Adsense income. With or without optimization, the fact that these types of sites exist in the Adsense network is one of the main complaints of advertisers (and publishers, too).
Then Google must make sure that it periodically reviews its publisher sites to see whether it has the type and quality of content that an Adsense publisher must have. I remember a few years ago when Burst Media did an exercise of weeding out "undesirable" publishers out of its network in an effort to raise the bar as to the quality of its member publishers (and attract better advertisers).
Adsense is still a young program. With competition gearing up, we can expect to see changes in the near future. Hopefully for the better
The problem with any reward system for quality is that Google favors scaleable algorithms over human intervention. Still, if advertisers had include and/or exclude filters, the "human intervention" could be on the advertiser side--meaning that advertisers, not Google, would pay for it. :-)
Google favors scaleable algorithms over human intervention
I can understand Google wanting to avoid human intervention by Google, because it's expensive in staff costs and doesn't scale. But human input from advertisers and publishers doesn't have the same problem, as there are more of them and they do scale. There must be better ways for them to use the distributed information of the ad market to improve AdSense.
I think Europe has the right idea about what Google should be doing, but I don't think they are -- not while they're pushing ads on parked domains and other content-less pages, anyway.
Adsense is still a young program. With competition gearing up, we can expect to see changes in the near future. Hopefully for the better
It will. It has to. You all are on the right track with this thread. There is a way to make this work for everyone and it is implemented with a group I know. It's lengthy to explain but goes something like this:
The problem with the way campaigns are and content sites are treated is wrong. They are treated the same and that is the problem. It's not just cpc and ctr that has to be taken into consideration but rather many different things.
Since campaigns are treated the same across sites that means the cpc is the same across sites. If site 'A' does very well and site 'B' does less then well the campaign as a whole may fail.
Overture is on the right track by allowing two different sets of cpc for advertisers but it still doesn't differentiate between content sites.
Let's say that site 'A' who does well for the advertiser gets $1.00 per click. If one click converts and the amount the advertiser gets from the conversion is $100.00 then the campaign had a ROI of 0. I'm just keeping the math simple and assuming that they are OK with 0. Now if site 'B' has to deliver 200 clicks to get one conversion and site 'B' still makes $1.00 per click then the cost was $200.00 and a negative ROI. The campaign failed. Let's suppose that the advertiser wants all the leads they can get. How do they make it work with site 'B'? Simple. Drop the cpc to $0.50 and the campaign ROI comes back to 0.
Can AdSense do this today? Nope. Could they in the future. Yep. Now if you are site 'A' your probably thinking great I make the max $1.00 per click. If your site 'B' your probably thinking this sucks only getting $0.50 per click.
What are the alternatives? Site 'B' would lose all revenue because the campaign would fail on them and they might end up getting the dreaded email or the advertiser would pull out of the program. Either way it's a loss. To get the highest payout the easiest thing to do is work on attracting better quality traffic.
By doing this the idea of tiers becomes less important. Google is a tier one player so the less than desirable sites don't do so well with them. The same goes for second and third tier players. They don't have to be they just have to be able to differentiate between sites and account for it programmatically. The tier can be moved down to the site level where it should be. The middleman should be able to account for this and automatically make adjustments. No campaign should fail unless they fail on an individual basis.
It's coming folks. I've seen it and it works.
JAG
4 days ago, I went from 468x60 banners to 728x90 leaderboards on my site. Traffic is largely big enough to represent a relevant sample for statistics. CTR went up since there are up to 4 ads served in a leaderboard, instead of up to 2 for a 468 banner. The ads are also more visible since they are wider.
My daily clicks number have been multiplied by X.
My daily earnings has been multiplied by Y.
X is much greater than Y. So much greater that it hurts.
I question myself more and more often about what happens if my traffic is multiplied by, say, 10. Will my earnings be multiplied by 10 or at least 8? Or will it be more like 3? Will it be enough to cover the costs of serving the additional traffic?
I understand that it could be a coincidence, a large advertiser pulling out and bids falling down, but this is just TOO big to be a coincidence.
I've been serving adsense ads since August so I kind of know how its digits work, now.
I'm more and more leaning towards thinking that there IS some kind of virtual "band", that Google tries to keep your earning inside of. Of course that band is not the same for everyone, I believe. And probably that band can change depending on some factors. But still, it means that simply increasing your traffic (and click amount) doesn't mean you automatically earn that much more money.
At least this is how it is for me. I would love to be proved wrong :-) Please do, if you have the stats to back it up. It will put me in a better mood. :)
Doesn't it stand to reason the average CPC will drop? Either a little or a lot, depending on how deep the selection of matching ads is.
So advertisers should be able to track the clickthroughs that actually result in sales (or extended interest in the site), and pass that information back to Google. (They don't need to know which sites those clickthroughs came from.)
And publishers should be able to mark ads appearing on their sites as "well focused", or off-topic, and so forth.
Any individual advertiser or publisher may not provide much information, but collectively it could come to quite a lot.
So advertisers should be able to track the clickthroughs that actually result in sales (or extended interest in the site), and pass that information back to Google. (They don't need to know which sites those clickthroughs came from.)
Why not just let them vote with their feet with inclusion/exclusion filters similar to the exclusion filter that AdSense publishers have? Advertisers would be reassured by the control over where their ads appeared, and Google could track inclusion/exclusion data to determine which publishers were considered better or worse than the norm.
And publishers should be able to mark ads appearing on their sites as "well focused", or off-topic, and so forth.
That could be a time-consuming process for even a medium-sized site. (Unless Google could figure out a way to include toolbar-style smiley and frowney voting buttons in ads that a publisher is viewing on his own pages!)
A simpler solution would be to have special Google meta tags that publishers could use to include or exclude certain categories of ads. For example, if ads for beer kegs and tappers were displaying above my Munich Oktoberfest article, I could put "Munich" and "travel" under an "include" heading or "kegs" and "tappers" under an "exclude" heading to help AdSense keep ads on target. Most of the time, these probably wouldn't be needed, but they'd be useful in cases where the Mediapartnerbot is flummoxed by words that are (or aren't) on the page.
