Eric Picard had an interesting article yesterday regarding AJAX and how it is throwing off the traditional ways of measuring page views or impressions of advertising at sites. Given that many of the new fancy web 2.0 sites written with ajax calls are actually changing the page without reloading the page, the traditional measurement for what counts as an “impression” cannot accurately be used. Discrete parts of the web page can pop out and change without the page itself reloading.

He is right that it is a headache right now. But I think it points to a bigger issue which is how effective are those impression banner ads anyway and should impressions even be a standard to look at? Frankly, we never use them where I work and rely 100 percent on search and pay per click advertising (and CPA deals). Ultimately, I believe every advertiser is looking to see how their ad translates to ACTION. The reason Google has worked so well is because the advertising is tied directly to performance. It’s worth paying for the clicks when you know they are looking for what you have to offer. These are the people who are most likely to convert to a sale (or the action you want) and also the ones most likely to remember you, given that you (presumably) are giving them what they need.

The problem with impression based advertising in general is that it can leave far too much up to the publisher. A publisher might design their site to generate more page views by breaking up one article into multiple pages. They can claim larger impression numbers than are actually there without any real way for the advertiser to know whether it is true or not. Further, depending on placement, these banners may not be juxtaposed in any way to pull the readers interest and lead to an action.

In cost per click, you have the middle ground between the advertiser and the publisher. The advertiser knows that they are receiving results from their advertising because they can measure all of the visitors who come to their site from the publisher. They figure out what they are willing to pay and evaluate different publisher sources based upon how that traffic performs. They don’t have to worry that too many impressions are being faked, or placed in wrong places, or that money is being burned in a big bonfire of “impressions” that will never lead to any action they can see or measure without trusting a 3rd-party source.

For the publisher, they can always take the cost per click and translate that into what they would be making on the typical CPM rate. If they don’t feel the are getting enough, they can search out more relevant advertisers who will either have a higher clickthrough rate or who are willing to pay more per click for their traffic.

You can see this with companies such as Bankrate, who sell off sections of their site such as the the comparison table for CD rates. As an advertiser, I know when placing my ad how the person got to that page and probably what is on their mind as they look at that rate information. The ad can be placed exactly at that moment in this person’s path through the Bankrate site.

The lesson: The best ads will always be both relevant to content and relevant to time. If someone is looking to invest in a CD and evaluating CD Rates on Bankrate, that is the perfect time to say “Hey, how about a money market account? Or an annuity?” That is why the best sites are always seeking synergies. A real estate site wants to have a moving advertiser because…well…you need to move when you buy a new house don’t you? And you probably want some help with that whole “moving thing.”

You can also see this approach in the new wave of blogs from smart companies. In the blog from 37signals, the web 2.0 ruby on rails darling company, they use a focused group of advertisers via “The Deck”, an advertising platform meant to target to the cutting edge web developer crowd, selling what they call “Cost per influence” not cost per impression. These ads are focused, relevant, and interesting to the readership. The Deck ad is the only ad on the page. Clean, unobtrusive, and likely to leave the right impression AND result in an action.

SEOMOZ, a search engine marketing firm run by Rand Fishkin, a regular in the SEM speaking circuit, has also been experimenting with similar approaches via Amazon recommended books. A few weeks ago they had put up a recommended book by Steve Krug on web design and usability called “Don’t Make Me Think.” Given that so many of their readers are sophisticated marketers with google adsense blindness, regular ads just wouldn’t cut it. A book on how to make your site more useable and effective however…that just might grab someone’s attention. It grabbed mine and hopefully they made a few bucks from my purchase. Good book too!

In my opinion, AJAX and these new technologies give the web a chance to redesign the way they look at advertising. The impression model seems too loosely based upon TV and mass media, and that is not what the web is. Web advertising has the ability to be highly targeted, highly specific, and highly personal. The best sites out there are going to be the ones that figure out how to use new technology to deliver the most relevant ads possible to their audience, who know what their audience is looking for and what they will be interested in at each section of their site. This leads to advertising that is not glossed over and filtered out, but to advertising that might even be enjoyed and seen as part of the value of the site.

Michael @ SEOG