An AdWords Paradox: Google Penalizes the Most Relevant Ads

Published: March 8, 2007

Author: David Rodnitzky

Google rewards ad text that has a high CTR, with the caveat that your ad text needs to make it pass a ‘evaluation phase’ (or sandbox) before it gets complete distribution across Google’s ad network. That’s the difference between a few and a few thousand clicks. Not insignificant!

The result of this rule is that there is a significant disadvantage to changing or deleting a well-performing ad. Once you get to the top of the heap, the last thing you want to do is go back down to the bottom and start over.

Google’s rationale for this rule is pretty straight-forward – ad text with an established track record of high CTR results in more clicks and thus more money for Google. This was actually a brilliant innovation away from Overture’s straight CPC model, and the market has clearly acknowledged that this will be the standard going forward.

But it’s worth considering whether this really does drive the most profit for Google, and whether it really provides the most relevancy for consumers. The problem with rewarding longevity is that it prevents what I would call “real-time advertising.”

Virtually any product or service marketed online changes frequently. New TV models are released multiple times a year, college enrollment periods come and go, mortgage rates change, seasonality influences buying behavior. As they said in Top Secret: “People change, hairstyles change, interest rates fluctuate.”

Truly savvy search engine marketers would no doubt love to take advantage of change. Consider these two hypothetical ads for the keyword “46 Inch Sony Vega TV”:

Ad 1: “46 Inch Sony Vega In Stock. Free Shipping. Shop Now!”

Ad 2: “46 Inch Sony Vega – $1299. Order by Midnight and Get it by March 9!”

I suspect that it wouldn’t take too long for Ad 2 to kick Ad 1’s butt. But by the time Ad 2 made it past Google’s trial period, March 9th would have come and gone, the price may have changed, or the model might be discontinued.

So what’s a poor Google to do? Maintain the current status quo and end up preventing the most relevant ads from having a shot, or eliminate the trial period and risk serving a lot of ads that simply don’t perform at the requisite CTRs?

I don’t have much of an answer to this paradox (but hey, I guess that’s why the folks at Google drive new BMWs and I drove here in a Hyundai). Maybe Google should reward advertisers with historical high performance with a shortened approval period? Or maybe advertisers should be able to designate certain ads as “time intensive” and get different evaluation rules?

The best idea I could think of was something analogous to Froogle, where advertisers submit feeds with multiple attributes to Google, including price, model name, delivery date, etc. It would look something like this: Buy {Product Name} Today for Only {Price} and Get Guaranteed Delivery by {Delivery Date}. Advertisers could use the Google API to automatically update attributes as they change.

Granted, if an advertiser suddenly increased price, you would expect the ad CTR to decrease. For that reason, maybe you combine attributes with a “quick trigger” review period that rapidly rewards or penalizes ads as needed.

No doubt this is something the folks at Google think about a lot. I think its only a matter of time until they come up with a solution. Then its off to the BMW dealership for another new toy!

Postscript: I want to thank Jeremy at for notifying me of a Blogger bug that auto-inserted a “nofollow” tag into my meta-tags. Hopefully I fixed it and I’ll be spidered soon. Anyway, cheers Jeremy for the Good Sam act!

25 E Washington Street
Suite 420

Chicago, IL 60602

(650) 539-4124

Want to become a client?

Contact Us decorative arrow

Want to join the team?

View Our Openings decorative arrow

Find us on social media.

Press inquiries.

Email Us decorative arrow

Expert insights for your inbox.

Accept No Limits.