|
Monday, April 25, 2005
In the coming weeks, Google will be rolling out a significant change to the content targeting side of its AdWords program. These changes do not affect "search" ads on Google and other search network partners.
The first major change: pricing is on a CPM (cost-per-thousand-impressions) basis, not cost per click.
Apparent rationale: (1) big companies and agencies understand CPM better; (2) fraud is a bit easier to police with CPM, though of course no system is bulletproof.
The second major change: a clever device to allow advertisers to target specific sites. This is what we've been asking for since the day AdSense rolled out. Many of Google's critics adopted a sort of fatalistic attitude towards the prospects of improvement in the AdSense program, since this upgrade took so long to arrive. I was a bit more optimistic that they would do something about it, but the release of such a comprehensive new offering now catches me a bit off guard. (They would have had us believe they were going to do it piecemeal.) Looking closely at various ways they could have designed the "publisher specify" feature, it always struck me that it could get incredibly data-intensive depending on the degree of control and reporting given to hundreds of thousands of advertisers interacting with tens of thousands of publishers.
Sure, a few smaller PPC's offer what amount to prototypes of similar functionality, but the pricing is prohibitive, the volume is low, etc. Google's real-world solution understandably took longer to build.
Is it a tacit admission that the content program wasn't working? Sure. It's tough to stack up the lip-service pro-AdSense statements of dozens of Google product managers and execs over the past year with the reality that they knew full well they were designing a complete overhaul. As so often happens, their pretend-clumsy schtick (akin to having a chef present in a quarterly earnings conference call) was really just sandbagging to throw competitors off the scent. My favorite one, overheard as recently as March, is "we don't think it's best to think of targeting at the site level - we think it's better to think at the page level." This of course because that was the kind of technology Google acquired and developed. "We think the status quo is best," is what that essentially said. They had to say it louder after they acquired Sprinks and then got rid of its people and site-targeting capabilities. In any case, they must no longer think any of this, since they're completely changing that assumption in the new version of the program.
AdSense was ill-conceived from the start, but the fact is (as you know from reading Google's latest quarterly report), it helped Google grow quickly and surpass rivals. (For cautious advertisers, there was also measurable ROI from the program, in spite of its drawbacks.) Now Google can afford to design a better program. It's a happy day for advertisers and reputable publishers alike.
And a happy day for me, since I no longer have to give that same old presentation on contextual ads. Something new to talk about for SES San Jose! :)
What stays the same? You need to look at measures of ROI to gauge the effectiveness of your campaign. You should take a professional approach to determining the effectiveness of your campaign with web analytics that make sense for your company, or you should hire a pro to do so. Brand lift, latent and offline purchases aside, in most cases there is simply no longer a need to "guess" whether your campaign is working. Don't guess. Measure, test, and measure some more.
Posted by
Andrew Goodman
View Posts by Category |
|
D'oh!
|
