Forum Moderators: martinibuster
this banner obviously outperforms conventional cpc ads, as google shows only the best performing ads. of course, on a social network site with absymal ctr this could be expected.
what a pathetic advertising medium myspace is for its owner. and what an incredible deal this guy got! he pays out several dollars for a subscription. if someone signs up, he makes secure earnings without any risk. everything else is free branding on millions upon millions of myspace pages.
concerning other websites, the issue i have with that is the following: the adsense algorithm doesn't factor in the massive benefit of visual branding. it's a serious flaw in the system. don't they think about that? publishers must be rewarded for this form of advertising.
showing high visual impact ads for free! i have no words for this.. i think it's disgusting. or just dumb..
[edited by: moTi at 6:51 pm (utc) on Sep. 5, 2007]
If it was even remotely relevant and didn't look like junk, I'd still be willing to give it a try. But the CPA ads I saw contained atrocious grammar, and/or took the visitor to the merchant's main website rather than a landing page that allows the visitor to easily complete the action they saw advertised.
I think the advertisers are viewing these ads as "freebies." I mean, why not throw in a few PPA ads if they don't have to pay to have them displayed? Why would the advertiser care if they have a poor conversion rate?
In addition, CPA ads usually rely on cookies. Some viewers have cookies turned off, or have their browser set to prompt for permission to set a cookie. In that situation, the visitor will almost certainly choose to block the cookie. In addition, some people delete cookies, because it's easy to do.
Add to that the fact that we rely on the advertiser to honestly report all of the completed actions. What if they only report 80% of the actions? How will Google know that? On the other hand, imagine if a publisher clicked on his own ads 20% of the time. That publisher would be banned quicker than you can say "click fraud."
One thing that would REALLY help out the CPA program would be to report the program's actual eCPM, like they do at Commission Junction. Right now, AdSense has a column that says "Performance on Network" along with some stars (1-5 stars) or text that indicates that the program is new. What does 5 stars mean? What does one star mean? Do the stars indicate percent of conversions, or actual money? These stars are meaningless, compared to eCPM. CJ lets you know in dollars how well the program performs on the network. Something like that forces the advertiser to put together a well-performing ad, and penalizes them for not reporting all of the conversions. When an advertiser sees that Program A pays $30 eCPM and Program B pays $2 eCPM, then Program A will rightfully get more exposure.
One thing that would REALLY help out the CPA program would be to report the program's actual eCPM
yes, that's a good idea and imo absolutely necessary to attract a fair amount of publishers. the trial and error which program works for you is way too much time consuming and not acceptable.
however, for this they have to collect the conversion data from all advertisers. and what about ppa fraud, aka false reporting of conversions in order to catch as many publishers as possible for the particular program?
and i suspect the current payout is way too low to show us a proper comparison between ppc and ppa ads. branding premium not even included, the payout would have to be not far away from the cost for the advertiser, because these are secure earnings. if the payout could compete with cpc, they could as well abandon the whole thing and let the ads solely compete with cpc ads in our ad blocks, right? ok, this already doesn't work out for cpm ads, so they deliberately let us uninformed. cheap trick to first get the advertisers on board and let the publishers do the costly testing.
[edited by: moTi at 6:34 pm (utc) on Sep. 6, 2007]
and what about ppa fraud, aka false reporting of conversions in order to catch as many publishers as possible for the particular program?
I don't think that would constitute fraud. They would be shooting themselves in the foot if they did that. However I wouldn't complain about it, since publishers would still get paid. As soon as they stopped the false reporting (i.e., a "bait and switch") then their eCPM would go down and publishers who are on top of it would eventually leave.
Got to be some other way than cookie-based.
Well one way would be for the landing page to be hosted by the affiliate provider (i.e., Google, CJ, etc) rather than the advertiser. The customer could be taken to a unique URL that has a required session ID in the query string. For example: google.com/offers?id=1234&session=4f53ds6. If the customer simply goes to google.com/offers?id=1234, the offer would not appear. The session ID would be forever linked to the publisher.
Another thought, not related to the above idea -
Since cookies can be detected by the publisher's server, one thing the publisher can do is write a script that checks for the presence of a cookie from their own website. This can be done easily using PHP or ASP. My thought is that if the website visitor didn't accept your own cookie, then they probably won't accept one from the advertiser, so it would be pointless to show them a PPA ad.
