Cringely follows up on the Adwords experiment

[tedster]

A couple weeks ago we had a thread [webmasterworld.com] about Robert Cringely's report on an Adwords experiment. In his new column [pbs.org] of Oct 13, he gets into it further with some interesting comments from Jeff Huber at Google, who is in charge of engineering for AdWords. (note - this is the second half of his column which begins with a bit about iPod videos)

Cringely writes:

Notice how in Jeff's explanation, above, he said the AdWords system adjusts for maximum profitability for Google and "maximum user perceived quality for the advertisers." Not the maximum return on investment for their AdWords budget. The key word is PERCEIVED.

As I read the transcript from Jeff, that's a bit unfair. The only place he uses "perceived" is in regard to relevance -- that refers to the end user, and not the advertiser.

At any rate, Jeff does give a list of the factors they are using in optimizing the AdWords algo:

...the AdWords system is highly dynamic and incorporates user behavior, optimization on end-user relevance (which is continually evolving and improving), competitor behavior, historical performance of advertising campaigns, and non-linear effects based on position, campaign configurations (e.g., rate limiting to manage to advertiser-defined budgets), and policy enforcement (e.g., editorial review on ad creative changes, algorithms to encourage diversity and minimize redundant ads)

Maybe Robert X. can fill us in some more at PubCon 10 in Las Vegas [pubcon.com] this November where he's the keynote speaker.

[bts111]

Nice post! Thanks.

This does not surprise me at all. If you are a shareholder you are in good hands ;)

[markd]

<< ...the AdWords system is highly dynamic and incorporates user behavior, optimization on end-user relevance (which is continually evolving and improving), competitor behavior, historical performance of advertising campaigns, and non-linear effects based on position, campaign configurations (e.g., rate limiting to manage to advertiser-defined budgets), and policy enforcement (e.g., editorial review on ad creative changes, algorithms to encourage diversity and minimize redundant ads) >>

Try explaining that to a client that 'just wants to be No.1'...!:)

[fearlessrick]

I thought the i-pod discussion was more interesting and informative, actually. How can I advertise on an i-pod?

[tedster]

I had a laugh when Cringely described this familiar situation:

Things were going well until suddenly they changed for a five-day period ending with the publication of my first AdWords column. The AdWords black box could be optimized for a while, but then it couldn't be. And then it could be again.

But his experimenters felt that something intentional was going on during that 5-day period, even though Google said it wasn't. Makes you wonder.

[weeks]

From my experience, I think you can take Google at face value here. But, those who are looking at value to the user in terms of dollars/clicks vs in terms of relevance are going to miss it.

I spent two years of my life working on a project similar to Adword/Adsense. My hard-won conclusions were the same as Google's. Alas, as noted above, try explaining that to publishers, investors, partners or ad buyers.

The one thing I would add is it's not difficult to adjust for relevance in a meaningful way, meaningful meaning where people are clicking. If you focus on where 80 percent of the traffic is (that is, 80 percent of what people care about), you find that humans are more of herd animals than you might like to think. It's not even difficult to predict and prepare Adwords for in some cases.

Or, put another way, the mass appeal of Paris Hilton last year was really amazing and really real.