Forum Moderators: martinibuster
publishers would be able to enable or disable ads of this type, much the same way they they can enable or disable image ads. google would also display the ads on the SERPs.
the ads would be indistinguishable from PPC ads and advertisers would specify how much they want to pay per conversion. A hybrid could also be offered, where the advertiser pays an amount per click, and much more if the click it turns into a conversion. pay-per-click ads and pay-per-conversion ads would be intermixed, and google would use eCPM figures to decide which should get priority placements.
of course the complexity is that the conversions would have to be tracked with far more accuracy than they are now, but does anyone doubt that google could achieve this in the way that commission junction, for example, does?
the effective publishers would get paid more, perhaps significantly more, and click-fraud would be reduced. it seems like this could be a win-win situation for everyone involved except those publishers that deliver low-conversion-traffic.
This method would also mean the virtual death of the pay-per-click program. The only advertisers who would find it worthwhile to pay per click rather than per conversion would be advertisers who are interested in branding. With the lowered competition for pay-per-click ad space, the price per click would deflate like a dropped souffle, and fft--there goes our income.
A variation that would be amusing is for advertisers to bid per click as they do now, but publishers only get paid per converting clickthrough. Thus if your actual clickthroughs converted at a rate of 10X the average during some time period, then you would get paid 10X as much per click, and if none of your clickthroughs converted you would get paid zero. That would remove the incentive for self-clicking - at least for sites where a conversion is based on an actual sale.
the other problem is that publishers are not paid in the most equitable way. scraper sites that have high CTR because the visitors are just trying to escape the page, probably generate good revenue (if they didn't, why would there be millions of them). meanwhile, legitimate sites that deliver targeted and sometimes primed-to-buy traffic, are not rewarded as highly as they should.
pay-per-conversion solves this.
The trouble is that this method would effectively create a bank of affiliate ads that the publisher has no control over.
google could limit the number of aff. ads to top-bidder-per-domain, just like they do on the pay per click system.
in this method, they have to trust that the marketer knows what they're doing
not necessarily... since google would have the conversion info, they would depreciate the ads that are not working. poorly performing ads are depreciated or dropped in the current system, and poorly converting ads would simimlarly be dropped in the proposed system.
This method would also mean the virtual death of the pay-per-click program
i dont think it would because not every advertiser has a service that is conducive to a pay-per-conversion model. but... if pay-per-click did die, that would be because the advertisers were happier with the model, and more money was being made. i am not suggesting that pay per click be thrown out, i am saying that the two should co-exist.
Sounds like a logistical nightmare from google's perspective. Making sure thousands of companies are tracking conversions correctly etc.
CJ does it, and they do it well. i havent thought this through completely, but the conversion, could even occur on google's server. google has the money to make this work, and since the vast majority of their revenue comes from advertising, they can spend the required money to make this function smoothly.
What's the point of Google getting into this business?
i think google could make more money with this. advertisers would make more money, because click fraud would be decreased. good publishers would make more money too. it would be good for everyone except publishers who deliver poorly-converting traffic.
As previous poster noted, publishers wouldn't have any control over bad converting advertisers
no... but google would have control over the bad advertisements... the ones that did not perform would be depreciated or dropped, just like they are now.
-----------------------------
in the end, advertisers are paying for conversions... thats how they make their money. since advertisers are the ones that drive the whole system, google needs to satisfy their needs.
share the increased revenue with the highly-performing publishers to incentivize them to keep doing what they are doing.
the current system incentivizes CLICKS. so what do you see on the internet today: vacuous sites with just ads on them to increase CTR. if you move to a pay-per conversion system, what do you incentivize? CONVERSIONS. that would lead to real content that informs users, and primes them for a sale.
the current system incentivizes CLICKS. so what do you see on the internet today: vacuous sites with just ads on them to increase CTR. if you move to a pay-per conversion system, what do you incentivize? CONVERSIONS. that would lead to real content that informs users, and primes them for a sale.
No, the scraper sites would just earn affiliate revenues instead of PPC revenues. It costs virtually nothing to create and host a scraper site, so for the scraper or other "made for AdSense" site owner, whether revenues are based on clicks or conversions doesn't change the owner's incentive. If it takes twice the impressions to earn the same amount of money, so what? It's all found money anyway.
I'd prefer that AdSense continue to fine-tune what it has now versus going the pay-per-conversion route.
When I've looked at the sites with AdSense ads on my site I've often been shocked (shocked!) by their poor technical quality. I have no confidence in their ability to actually sell anything.
Good point.
The problem is if these sites affect conversion factors used by Google when it comes to paying publishers.
But as others have said, if you want affilliate ads on your site you can always sign up with an affiliate program.
That's not Google Adsense, at least at the moment.
google could limit the number of aff. ads to top-bidder-per-domain, just like they do on the pay per click system.
But will those ads be the right ones for a given site? If there are two advertisers for a given niche, a chic-and-trendy one and a homey one, and I have a homey site, the chic-and-trendy site is going to convert much worse for my traffic. Do I have to put up with losing visitors to the chic and trendy site because they're willing to pay more for conversions (that they're not going to get from my site)? Do they have to stay on my page because they do better for other people in the program, even though they're garbage for me?
Also, there's the question of these "bids." In affiliate marketing, publishers can earn either a set amount (generally worth several dollars, at least) or a percentage of the sale. Your proposed model would pay only a set amount--that the publisher doesn't get to okay. If the advertiser wants to pay only a couple of dollars per conversion, the publisher can't just switch to another advertiser. And the publisher gets the same amount whether they primed the visitor to make a $5 purchase or a $5000 purchase. It works out to less money and less control for a higher-risk system.
And worst, given Google's current black-box management style, we wouldn't know what the going rates were or where we were getting our conversions from. It would all come down to faith. We muddle along well enough with PPC because a) Google is the best thing going at the moment and b) we see a continual stream of revenue even when the system isn't working so well. That changes as soon as you bring in affiliate marketing.
since google would have the conversion info, they would depreciate the ads that are not working. poorly performing ads are depreciated or dropped in the current system, and poorly converting ads would simimlarly be dropped in the proposed system.
That means that if there's an advertiser that would convert beautifully for my program, but not for most other people's, I wouldn't be able to use them. Finding the right affiliate for a given site is a tricky process that requires a lot of experimentation. It doesn't come down to plugging in the right algorithm.
Sounds like a logistical nightmare from google's perspective. Making sure thousands of companies are tracking conversions correctly etc.CJ does it, and they do it well.
They do? There are plenty of reports of CJ's system not working properly. Not enough to cast doubt on them, but enough to make one wonder: If a company that manages affiliate programs full-time has these kinds of problems, how could a hands-off, algorithm-happy, highly diversified company do better?
If Google wants to start up an affiliate program, well and good. It would have to be separate from AdSense, though, because affiliate marketing is a completely different beast from PPC.
EDITED TO ADD: An "AdSense for Products" program of the sort Buddha suggested in another thread might be a smash hit. You write a page about pink Australian countersham widgets, you insert a ProductSense block, Google serves up ads for specific pink Australian countersham widgets with pictures and price info. If one of your visitors buys one of the widgets, you get a percentage. Because it sells individual products instead of entire sites--among other reasons--it would sidestep a lot of the concerns about marketing that I listed before.
BUT the beauty of affiliate programs is that you can pick and choose your partner and what products to sell, integrate the links into your site's content, and pre-qualify and pre-sell surfers. That partnership is the key difference that can make affiliate programs competitive with and sometimes superior to AdSense, if you're smart enough and you work at it. Unfortunately that would be impossible with an AdSense-type system.
Some years ago, I created a web site as an affliant for 7% sales comission.
The company claimed not one conversion after 120.921 visitors.
Last week, I heared the company went bancrupt. I found out about the official receiver and talked with him, that I would like to compare customer lists with my address list to find out how much the company owes me.
Oh, not possible, because in a real idiotic behavior to make any enlightment about the bancrupt impossible, they made next day an auction with all the computers of the company.
The company claimed not one conversion after 120.921 visitors.
I could bet it was affiliated with CJ.
