Google’s Evolving Algorithm

This week, Google went live with yet another change to its Quality Score system that calculates whether an ad should be shown on the first SERP (define) and the frequency and position at which an ad is shown. While Google has blogged about the change, as well as e-mailed the named advertiser contact for each AdWords account, I found that the level of knowledge about this important change is highly varied in the marketplace.

For example, at the annual summit this week in Las Vegas, many marketers weren’t even aware of the changes that went live during the show. Those who were aware had mixed opinions on the impact Google’s changes may have on their search marketing efforts.

The conference attendees included a highly diverse set of roles and job titles, from marketing directors to CMOs, plus hands-on campaign managers and those who delegate search tasks without ever having logged into a Google or even paid search application. This explains some of the diversity in level of interest or knowledge about Google’s AdWords changes.

I was surprised how little coverage there was of paid search, auction-based media, and behavioral targeting within the conference agenda. When I asked some attendees about it, some felt it was a tactical area, while others felt it had been covered ad nauseam in the past and there was little new about it. Yet much has changed in the last year in the online search space and significant changes are occurring all the time, some of which are very major, such as the Google-Yahoo linkup, which is under regulatory review.

If you’re a marketer and feel that now isn’t the time to ignore search or consider it “taken care of” because either a staffer has been hired or an SEM (define) agency retained, you may find yourself losing ground. Your competitors are innovating with new campaign structures, ads, keywords, and landing pages. Make sure you keep up, because standing still in the rapidly evolving search ecosystem is the same as falling behind.

Quality Score Dissected

Many search marketers, including some on Google’s staff, don’t really understand the specifics of what makes up the Quality Score. Though Google provides some guidance on its blog, I’ve found that the information from Google isn’t always in one place nor is it always clear about the cause and effect that campaign structure, ad relevance, keyword relevance, and starting position has on the Quality Score over time.

Generally, the easiest, most effective way to think of Google’s Quality Score is as a predicted CTR (define), normalized for position, with the addition of a few minor factors relating to post-click relevance. For example, Google collects CTR data for your ad regardless of your position and compares it against the norms for that position and possibly the norms within the SERP for that specific search query (i.e., your neighboring competition). By normalizing the data (factoring in your position), Google is attempting to evaluate ads on a level playing field.

It shouldn’t matter, then, if your ad is in position one or six; your Quality Score should be calculated accurately. However, after reviewing tens of thousands of ad groups and millions of keywords, we’ve found that one listing often gets a bit of a boost in Quality Score when it runs in the top three SERP positions because data is collected much more quickly in top positions and the normalization curve may not be perfectly accurate. This positive effect may only be temporary and may only hold for as long as Google gathers more data on your ads and the competing ads around you.

Consequently, you might blow your ROI (define) by aggressively bidding for keywords after the launch of a new AdGroup or campaign in the hopes that a good Quality Score will help you in the long run. For this reason, many search practitioners recommend an aggressive start to a campaign, both to learn Quality Score quickly and to take advantage of any inherent bias in the Google algorithm with respect to assigning an artificially high Quality Score to higher positioned ads.

In a nutshell, then, anything you can do to increase your predicted CTR, regardless of which position you happen to obtain, is a good thing. With the new changes to its Quality Score calculation, Google states “Quality Score is now more accurate — because it is calculated at the time of each search query.”

This can be interpreted in many ways, but in the very brief time that this change has been live, it looks like Google has more flexibility in terms of calculating Quality Score in real time for broad matches, as well as for adjusting Quality Score based on other factors that may correlate with predicted CTR. Many of those factors may never be disclosed by Google, but we’ve seen enough variation in ad CTRs, based on a wide variety of variables above and beyond position, that we can imagine daypart, geography, or even IP address block (ISP) are now being used as potential adjusting factors to your real-time Quality Score.

It will come as no surprise that these changes will likely result in higher Google spending, because both billed CTRs may rise on average and — if Google actually picks the more relevant ads more often — CTRs will rise as well.

Join us for SES Search Engine Marketing Training Day, September 26 at the Fairmont Dallas.

Related reading