"Sympathize" is the right word.

E-commerce, of which paid-search is effectively a portion, has already proven that one company's product is often a larger company's loss leader. Impressions (i.e. branding for its own sake), rather than sales lead or customer acquisition, are currently a loss leader. Google may be able to start charging for impressions if it succeeds in consolidating the online ad market further. However, if Yahoo, Microsoft, or Amazon takes market share, impressions may remain free for a long time to come.

• Posted by: Scott Rafer
• May 17, 2004 10:57 AM

Publishers now have more responsibility to prove the value of their audience and their site content. If an advertiser can reach the same audience through Google that NYTimes publishes for, then why advertise on NYTimes? Particularly, if the advertiser can target geographically through Google...?

Publishers must differentiate the value of the brand experience and the user behavior on their sites from that of Google's users. On Google.com, the ad serves a specific purpose, a very light, shallow relationship with the user. On NYTimes.com, the ad may have a less specific impact but perhaps much more meaningful. And the noise level of bad audience will definitely be lower on NYTimes.com.

Publishers will also have to build higher quality offerings for their advertisers to reach the audience such as custom programs, webcasts, or contextual opportunities that Google cannot offer given its wide reach. Again, Google has taken the low-hanging fruit, and publishers will have to show why their audience is more valuable on their site than that same audience is on Google.com.

Time for a new set of ad measurement tools in the market.

• Posted by: Matt McAlister
• May 17, 2004 11:09 AM

I did some rambling on the topic. Basically I disagree with the fundamental assumption that AdSense (text or image) is free branding. The ad shown is the one with the highest "effective CPM", so what's the difference from an ordinary media sell, other than the inventory is 100 % sold?

• Posted by: Henrik Torstensson
• May 17, 2004 1:00 PM

Agreed with Henrik.

Perhaps the other banner networks are worried about not matching Google's Effective CPM for publishers?

• Posted by: sean
• May 17, 2004 1:51 PM

think of a two inch ad in a newspaper or magazine. estimate downwards from there.

• Posted by: Robert Sayre
• May 17, 2004 5:09 PM

One advantage to Google's approach is the relative safety from having your brand besmirched by poor campaign choices. I know that there are some companies that because of horrid flashing banners in one of their campaigns have lost me as a consumer.

• Posted by: mneptok
• May 17, 2004 7:41 PM

John Battelle's Searchblog: What's the Value of Brand Advertising?. A great question. Most of the stuff on the Internet is all about direct response, but that's not the way lots of purchases work.

It's not like you think I need Tide and then you click on it 90% of the time. More like, you see Tide a bunch of times (frequency) and lots of people see it (reach) and eventually when you buy a detergent, you see it and say, yup I like it.

That would be like car guys blitzing people just in the 1 month they want to buy a car. It is often way too late by then. That's what image is all about.

That's not to say direct response and brand don't live together. They just have different purposes.

Depending on the play you are running and the playfield of course :-)

• Posted by: Rich Tong
• May 18, 2004 12:00 AM

Google is committed to the 'pay-for-performance' approach to advertising and it happens that clickthroughs are the easiest measure of performance of an ad. Impressions are not really free because if you don't get enough clickthroughs your ad will be dropped. So, there is an effective 'minimum price'

• Posted by: Seun Osewa
• May 24, 2004 2:08 AM