Google AdWords Interview on the New Quality Score Changes
Aug, 26, 2008
I spoke to our main Google contact today about the imminent AdWords Quality Score changes. These changes have only been rolled out to a small percentage of accounts so far, but will go system-wide soon (she didn't give me a specific date).
Here's a summary of what I discovered. No, it's not in interview format, what she told me is written up in summary form. Less reading for you!Quality Score Rating Quantified!
We're going to see a 1-10 scale for quality score, along with the same old poor->great continuum... but quantified now! You'll be able to see room for improvement with keywords that are Great but not the Greatest. You could get them even Greatester if you work on it.Micro-microsites are BAD!
Lead generation sites with only one page are BAD and have been for a long time, but this is going to get worse. If you do lead generation in the future, do not created isolated new microsites or you'll pay much higher cost per click. Instead, create new pages on existing sites that have other information on them. Give it some semantic weight and something searchers can benefit from immediately.The "Ad Display Frequency Algorithm" is Complicated
AdWords has a sub-algorithm that I've never read about that determines how often ads are shown when keywords are searched for (how many impressions they get)- I'm calling it the Ad Display Frequency Algorithm. It's complicated. (That's me being confused on the left.)Automated CTR Optimization, Time of Day, and Geotargeting
Even our Google rep, who's quite smart and in the know, is not clear on all the details- but this display frequency algorithm will now be affected by historical data on click time of day and CTR in various geographic regions. The geo-part is key for our local clients, but she said the differences we'll see will be small.
STILL, the implication is we might want to relax some of our geotargeting to allow for better auto-optimized performance in locations we're not sure of either way- and we're going to have to watch state by state metrics before and after to see if anything important changes. Google Cares About Their Clicks, Not As Much About Your Conversions
All of this is CTR and relevance based, and we also care about conversion rate and booking amts, etc., so we still have to watch for high CTR low CR situations. I questioned her about why Google doesn't incorporate conversion rate for customers that track it, and she mentioned customer concerns about privacy- I said those who use AdWords conversion tracking probably aren't as worried about that- even an opt-in would be cool. I also said I understand Google gets paid based on CTR and that's their business model- so we still have to watch the conversion rate, ROAS, average revenue per conversion, etc.
We already have to keep CTR vs CR in mind when optimizing- an automated optimization of CTR makes that a bit more complicated.Advertising on Not So Relevant Keywords Will Be Virtually Impossible
AdWords will be more harsh on irrelevant keywords. For example, one of our clients recently requested a test of the "myrtle beach" keyword which isn't highly relevant to their exact offerings- only 10-20% of people searching MB are interested in that product/service. We weren't surprised to see them get 10% their typical CTR and much higher cost per click.
These sort of attempts to place ads on irrelevant keywords will be more expensive and get fewer impressions. Basically, learn that AdWords is all about relevance, or suffer zero results. And that's good.Summary
Our rep said they would roll these changes out cautiously- and that advertiser reaction could destabilize things temporarily, but over time it should result in improvements. Still, I'm uneasy about automated CTR optimization- it could be great, but then again, sometimes a really high CTR results in lower CR... so I'm going to wait until the results are in.