It appears there’s some unrest among Google’s AdWords users–or at least a small, vocal fraction of them.
Google’s little-known “session-based clicks” algorithm is starting to get noticed by advertisers who are seeing their ads displayed alongside seemingly unrelated search queries. Case in point…
…Mr. Dorfman, the New York dentist, said he wanted to show his ads when people searched for things like “cosmetic dentist nyc” or “orthodontist.” But last August, after downloading a comprehensive report of AdWords charges, he saw there were session-based charges for unrelated searches.
For example, Mr. Dorfman said he was charged between $5.36 and $7.26 per click when Google users clicked on his dental ad after searching for “penis enlargement” and “[Chinese characters] in nyc Chinatown,” respectively.
After an email tussle, Google refunded $300 in clicks but claimed that this was “expected, normal behavior” because the user had first searched for “dental implants” before searching for “penis enlargement” – I don’t know about you, but that sounds like a dude planning some major renovations!
Google argues that the ads should be displayed because the user had expressed an interest during the same search session. The advertiser argues that the user had clearly moved on and is no longer a targeted customer.
I’d argue that both sides need to take a long hard look at what they are saying. Guys searching “penis enlargement” don’t click on “dental implants” ads by mistake. Clearly there was intent there.
And Google? Either you stand by this program or you don’t. You can’t on one hand say, this is expected, normal behavior, and then turn around and refund the guy $300. Either you have confidence in your algorithm, or you don’t.
Anyway, what do you think? No, I don’t need explanations of how dentures & apertures are connected. What do you think about Google’s session-based click algo?
Update: We received this from an official Google spokesperson:
This feature is available for advertisers who use the broad match option for a keyword, and only the broad match option, and represents a single digit percentage of impressions system-wide. There are several match type options available for advertisers which gives them the choice to decide which type performs best for them. An explanation of our keyword matching types is available here. We’ve been consistent in highlighting the Search Query Performance report as the way to see which queries trigger these ads, and advise advertisers to review these reports to determine keyword performance and adjust their keywords accordingly with broad match modifier. The performance of session-based broad match ads has no effect on quality score, and – as with all Google search ads – there is only a charge if a user clicks on an ad.