[usatoday.com...]
Some consultants overseeing Web ad campaigns are telling clients to proceed cautiously when considering contextual ads, which are served to news and information Web sites when certain words appear in content.They say the ads can perform more like banner ads — the humbled, old next big thing of the bygone dot-com boom — than the lucrative search ads that inspired them.
I have a hard time with the lack of targetting control. I have a hard time not knowing where my ad is being shown (which could help me determine whether I want to continue or not).
I could whine even more but I'll pass the guitar so someone else can sing.
Yup, having been selling online ads on my own sites for 5 years now, I totally agree. The first thing many advertisers will do is cut out the ad broker if they can find the delivery point.
Too bad they weren't showing anything related when I viewed them...
They say the ads can perform more like banner ads — the humbled, old next big thing of the bygone dot-com boom — than the lucrative search ads that inspired them.
That's probably true of contextual ads on general news and information sites, such as the AdSense-style ads that appear in WASHINGTON POST articles.
If so, it's worse news for Overture than for Google, since Overture's "Content Match" program is geared toward large news and information sites (which offer only keyword targeting not audience targeting). Google's AdSense, in contrast, is running on thousands of niche sites, which--in many cases--offer keyword targeting plus audiences that have a demonstrated interest in the product or service being advertised.
I publicly asked an Overture exec (at the recent SES) about their ability to scale their relevance when their program is dependent on twelve human editors (see the thread about Overture taking potshots at Google), and his response was that they have over a hundred editors and that their editors have scaled quite nicely for PPC, and could easily handle the ContentMatch duties.
Of course, the statements made at the SES have to be taken with a certain amount of skepticism, as I caught some execs from other companies telling half-truths and I caught one of them telling a flat out lie.
Nevertheless, their dependence on News Outlets (Knight Ridder) does give rise to legitimate questions about the quality of the traffic.
Although as long as Google has no copyright or legal stake to sending a bot to an ad page, it can be easily copied by Overture and mined for keywords...
RCJordan wrote
Yup, having been selling online ads on my own sites for 5 years now, I totally agree. The first thing many advertisers will do is cut out the ad broker if they can find the delivery point.
The transaction costs are too substantial for at least mom-'n'-pop sites, and possibly too onerous even for larger concerns.
[Fictional] "IguanaFansSupreme.com" ("IFS") is the darn best Iguana portal on the planet. "ExoticPetDepot.com," in fact, already knows about IFS, and would love to advertise their Exotic Pet Aquariums on IFS' site.
But neither IFS or EPD has suffient technology capital or know-how to measure fraud-free clicks. EPD also doesn't feel like spending the time or money sending checks to IFS or dealing with tax issues, and so on.
But by bidding on key terms, EPD can likely appear on IFS' site via AdSense and other related sites without worrying about all the techno mumbo-jumbo and payment hassles.
Larger firms may have the chops to buy and sell advertising directly, but do they want to be spending their time doing it? They may end up spending more than the cut they give to the efficient Google AdSense team, after all.
I publicly asked an Overture exec (at the recent SES) about their ability to scale their relevance when their program is dependent on twelve human editors (see the thread about Overture taking potshots at Google), and his response was that they have over a hundred editors and that their editors have scaled quite nicely for PPC, and could easily handle the ContentMatch duties.
I can imagine Overture matching ads to categories (travel, say, or maybe even "Florida travel" or "cruise travel"), but I don't see how it would be cost-effective to have editors match words to individual stories. On general-interest news and entertainment sites, stories simply come and go too quickly to make such matching practical.
Manual matching of ads to content is likely to work best on niche sites with "evergreen" content. And it doesn't have to involve human editors at Overture or Google--it could work just fine with a combination of Google's AdSense technology and keyphrases or other input from publishers (who have a vested interest in ensuring that an article on communion wafers doesn't include an ad for silicon wafers or vanilla wafers).
Google's efficiency of scale is an import factor and a significant barrier to entry.
There is almost no overhead on either side using the current scheme. Add the research or simply processing the order and bam the costs would have to rise substantially.
