Chapter 12. How Quality Score Works in 2014
April 2007 brought with it a major change to AdWords that effectively shut down many advertisers’ campaigns overnight. This change, also known by some as the Google Slap, was intended to get rid of low-value affiliate and one-page sales letter sites from the paid listings, but there was plenty of collateral damage. Many high-quality sites (including some affiliates) were shut out of the paid search listings, while many not-so-high-quailty sites (including some affiliates) posing as high-quality retail or review sites were awarded rankings and coverage bonuses.
Like many others, I had sites that were affected by this change. The minimum required bid for keywords where I’ve been advertising profitably for years increased to $10.00 because of Google’s insistence that my sites were no longer relevant.
I don’t believe it was Google’s intention to shut down legitimate sites that provide value for search engine visitors. Rather, they instituted this new algorithm as a means to encourage advertisers to raise the quality of their own ads. Where they went wrong was in providing little guidance to advertisers as to exactly what constituted “quality.” Furthermore, the algorithm has continued to evolve over time, making it somewhat of a moving target which advertisers constantly have to chase.
As we saw in the past two chapters, managing your quality score is a critical step in effectively managing a paid search campaign. Quality score affects your average position, coverage, clickthrough rate, and your cost-per-click. So this information contained in this section should prove invaluable.
[H1] What Are Quality Scores?
For every keyword in your AdWords campaign, Google will assign you a Quality Score between one and ten. A low score indicates a low quality ad while a high score indicates a high quality ad.
Why Is Quality Score Important?
Quality score is one of the components used to calculate your ad rank. Higher quality scores help to both push your ads higher up on the page and also increase the difference between your maximum bid and the average CPC. In other words, high quality ads cost less and gain exposure to more search traffic (impression share.)
Another benefit of having a high quality score is that it will lower the estimated bid to reach the first page of the search results. This metric, also called the first page bid estimate, approximates the CPC needed to reach the first page of the Google search results when your ad is triggered by an exact match. A high quality ad may cost as little as $0.03 to be shown on the first page, while a low quality ad can cost as much as $100.00, so it’s easy to see why you must be cognizant of your quality scores.
History of the Quality Score Algorithm
In the early days of paid search (2000), there was no Quality Score. Google sold placements in the traditional manner (through ad reps) to larger advertisers on a CPM basis. CPM cost-per-mille (thousand) meant that advertisers paid for every thousand impressions their ads received.
In 2002, Google opened up its advertising platform to anyone with a credit card under the name, “AdWords Select”. A key innovation of this platform was the incorporation of clickthrough rate into the pricing formula. Unlike the Overture (Yahoo!) straight-auction model in which the highest bidder always received top placements, advertisers who received a higher CTR on their ads would pay less for their placements.
While effective, the platform was still rife with unwanted ads. Thus, the Quality Score algorithm was introduced by Google in August, 2005 and underwent frequent iterations over the following years. The original version was not exposed to advertisers until 2007, when Google added the Quality Score designations, “Poor,” “Ok,” and “Great”. At the time, Google also added a feature which lowered the minimum bids for high-quality ads and raised the minimum bid for low-quality ones.
Quality Score worked by searching for tell-tale signs of spam within the ad copy (does it have multiple exclamation points? Does it use the word “free”? Does it match the targeted keyword phrase?) as well as the landing page (Is it a brand new domain? Who’s linking to the site? Does it take a long time to load?). There were also various metrics which played into the original Quality Score algorithm as well (keyword density, backlinks, page load time, and more). As advertisers became more aware of how Quality Score was being computed, Google continued to evolve the rules.
2008 brought several changes including the incorporation of landing page load time, real-time calculation, and an adjustment to account for the impact of average position on clickthrough rate (quality score normalization). There were potentially many other minor inputs as well (Google filed patent applications in 2007 for 44 different Quality Score factors). However, any automated criteria can be gamed exactly the same way that spammers have been gaming the search engines for years and this is exactly what happened after each of these changes. Through the use of clever tricks such as keyword stuffing, crash and burn domains, doorway pages, and so forth, spammers could find ways to get around some of these automated checks.
In response, we began to see an increase in the number of manual reviews which were designed to complement the automated checks and give advertisers a means to appeal the automated system. However, this approach did little to deter spammers (who would just close out their old account and start a new one) and turned out to be incredibly frustrating for many advertisers who were running legitimate campaigns. Imagine how it feels to wait two or three days for your ads to start running and then have someone tell you that your landing pages aren’t “high quality” enough. Moreover, there were extreme lapses of judgment and abuses of the system .
Another troubling aspect of the algorithm was that the minimum bid which Google calculated for all advertisers affected ads shown anywhere on the search network or partner sites. The “all or none” philosophy behind this approach angered advertisers and also resulted in a large drop in ad coverage . This may have negatively impacted Google’s revenues (management even said as much during their second quarter 2008 shareholders meeting) and the algorithm was changed from a static, all or nothing approach to a dynamic one in September, 2008.
The new approach calculated quality scores on the fly. This means that ads would be awarded higher quality scores for different keywords, in certain regions, and on different partner sites. While this new added flexibility addressed the “all or nothing” problem, it was yet one more change which made the Quality Score algorithm even more complicated than before.
In the previous edition of this book, I expressed hope that Google would perhaps one day dial down the complexity and make things easier on us. Well in 2011, chatter among search marketers suggested another change was taking place and we independently discovered that some of the traditional quality score factors were being downplayed. We decided to test this with our own models and found that approximately 70% of the variance in Quality Score could be explained by a single factor: clickthrough rate.
The Official Explanation of the AdWords Quality Score
A good place to start in our understanding of the Quality Score calculation is to read what Google has to say about it :
How we calculate Quality Score
Every time someone does a search that triggers your ad, we calculate a Quality Score. To calculate this Quality Score, we look at a number of different things related to your account. By improving the following factors you can help improve your Quality Score:
Your keyword's past clickthrough rate (CTR): How often that keyword led to clicks on your ad
Your display URL's past CTR: How often you received clicks with your display URL
Your account history: The overall CTR of all the ads and keywords in your account
The quality of your landing page: How relevant, transparent, and easy-to-navigate your page is
Your keyword/ad relevance: How relevant your keyword is to your ads
Your keyword/search relevance: How relevant your keyword is to what a customer searches for
Geographic performance: How successful your account has been in the regions you're targeting
Your ad's performance on a site: How well your ad's been doing on this and similar sites (if you're targeting the Display Network)
Your targeted devices: How well your ads have been performing on different types of devices, like desktops/laptops, mobile devices, and tablets you get different Quality Scores for different types of devices
The most important component of your quality score is the historical CTR of your keywords. This is Google’s bread-and-butter and hasn’t changed for many years. However, it does seem to have been increased in importance more recently.
It is not only the CTR you earn on a particular keyword that is important, but also the CTR you earn on a particular landing page. Incorporating this as a factor prevents spammers from continually creating new AdWords accounts to show their ads once they’ve been banned (they have to create unique landing pages as well). This also means you should avoid using important URLs in your AdWords campaigns unless you are prepared to live with the consequences
your past performance history can stay with you for a long time.
Landing Page CTR
Google now factors in the CTR of each unique landing page into your Quality Score. If you are using the same landing page for every keyword in your campaign, it’s likely that your lower CTR keywords could be adversely affecting the rest of your campaign. For this reason, it pays to have a variety of landing pages in your campaign.
Next up in importance is the historical CTR of your entire account. This is a grossly overlooked point. Advertisers who target thousands or hundreds of thousands of keywords with no regard to their relevance get penalized hard by Google. The same goes for those who fail to remove or improve keywords with mediocre clickthrough rates (if I can’t achieve a CTR of at least 0.5% with a keyword after several attempts to optimize it, I delete it from my account.)
Next up we have several relevance factors. The first is “keyword/ad relevance” which attempts to determine how close of a match there is between your ads and the search query which triggered the ad. Years ago, most advertisers caught on to the fact that including the search query within the text of an ad tended to improve its Quality Score. One might wonder if Google would have caught on to this practice and somehow addressed it. Our models indicate that this factor now appears to have relatively little weight, so this may indeed be the case.
The next factor is “keyword/search relevance”. This refers to the commercial intent of a particular search phrase. For instance, the search query, “South Park,” used to trigger a large number of ads. In 2012, this term stopped showing ads altogether. This was presumably because searchers weren’t clicking on the paid search ads. Generally speaking, this is not a factor that is in your control (it is more likely determined by the combined performance of all advertisers).
The next factor is “geographic performance”. This means that Google will calculate your quality score separately for each of the regions (even down to the city level) which you are targeting.
The next factor, “Your ad’s performance on a site,” only affects display advertising.
The final factor is “your targeted devices.” Just as with geographic territories, Google will calculate Quality Score separately on each type of device you are targeting (desktop, mobile, and tablets).
Do Landing Page Quality Score Factors Matter Anymore?
In the previous edition of this book, I shared the results from controlled experiments in which we attempted to identify the more important on-page factors which impacted quality score. These factors included:
Presence of popups and popunders
Load time and page size
Number of pages / scope of site
The importance of these various factors has been debated endlessly by paid search marketers for years. When the previous edition was written, the above list was essential. Today, it appears that the on-page quality factors have been reduced in importance considerably (with the exception of potential policy violations, such as the use of popups and popunders, illegal or prohibited items such as pharmaceuticals, and the presence of malware).
What we have discovered instead is that the second most important factor in determining quality score is Average Visit Duration, or the average amount of time a visitor stayed on your site after clicking your ad.
When a large number of searchers click an ad and then stay on the landing page or site for a long time, it strongly suggests a highly relevant advertiser. So combining average visit duration along with the CTR factors mentioned above makes intuitive sense. It is also a simple and elegant solution for both the search engines and advertisers as it can be easily measured through the user of redirect or wrapper URLs inserted into each paid search ad (Google now wraps all landing page URLs in their own redirect URL beginning with “http://www.google.com/aclk?...”). Should a searcher click the back button, the length of the visit can be easily calculated. And what’s more, this solution works even if the user opens the ad in a different browser window (for instance, by ctrl-clicking).
In fact, this metric was even added in mid-2012 to AdWords for users with linked Google Analytics accounts. To include it, navigate to the “Campaigns” area of your AdWords account and click on “Customize Columns”. If you have a linked Google Analytics account, you should be able to add the “Average Visit Duration” column to your AdWords reports. (Note that it is unlikely that this is the exact figure that Google uses as they are measuring bounce backs for all advertisers using the redirect URL shown above).
Figure 12-1 Adding “Average Visit Duration” to Your AdWords Reports
Best Practice: Managing Quality Scores
It’s important to understand the various mechanical factors which may impact quality score, but at the end of the day remember that it boils down to two key tasks:
Getting a higher CTR than your competitors
Maximizing your average visit duration
How to Check Your Quality Scores
In the Google AdWords interface, drill into your keywords. Clicking the small balloon icon in the status column will reveal your quality score:
Figure 12 2 You can check your quality score directly within the AdWords interface.
Limitations of AdWords Quality Score Reporting
There are two main problems which continue to linger on with the AdWords Quality Score report shown above. The first is that it is inconsistent and hardly trustworthy. While Google has added a bit more transparency by providing ratings for “Expected clickthrough rate”, “Ad relevance”, and “Landing page experience”, these ratings often do not correlate with the overall quality score assigned to the ad.
As an illustration of this, look back at the preceding figure 12-2 and note that our quality score is reported as 7 while all three of the contributing factors are shown as “average”. Compare this to Figure 12-3:
Figure 12-3 Quality Score as reported for another keyword in the same campaign
Here you can see that two of our contributing factors are shown as “above average” but now our quality score is only 4.
The other problem is a bit more insidious. Before 2009, you could be stay somewhat aware of holes in your campaign by simply looking at your AdWords reports. Back then, when Google said you had a low quality score, they meant it. Your ad wasn’t being shown.
These days when you see a low quality score, you may still have ads appearing for the keyword in certain regions or on partner sites. If Google tells you that your ad isn’t showing, another one of your ads, possibly an unrelated one, could very well be showing up in its place. Conversely, having a high quality score does not guarantee that your ad is being displayed even if the tool says that it is. It may only be showing up on partner sites (such as AOL or Ask.com), in certain geographical regions, on certain devices such as tablets or phones, or on unrelated keyword phrases. There’s no easy way to tell.
Best Practice: Monitor For Missing Keywords
We recommend that advertisers utilize a third-party monitoring service to ensure that their ads are appearing for their selected keywords within each of their targeted engines and countries. There is currently no way to protect against this using the standard search engine reports.
Quality Score On Bing Ads
Contributed by Ping Jen, Bing Ads Product Manager
The goal of Bing Ads Quality Score is to help advertisers optimize the quality of their campaigns. It’s designed to identify campaign optimization opportunities instead of measuring campaign optimization results. This approach enables Bing Ads be more transparent in providing insight to advertisers through its Quality Score.
In order to provide broader and deeper insights, in addition to tracking the performance of keywords during the auction, Bing Ads also:
Factors in marketplace competition while generating Quality Score by comparing potential click-through rate (CTR) of your keywords to potential CTR of other keywords targeting the same traffic.
Generates Quality Scores for each individual match type because different match types exposes keywords, ads, and landing pages to different audiences and different competitors.
Provides ad group and campaign level Quality Score by aggregating (impression weighted) associated keyword Quality Scores.
Bing Quality Score Components
Advertisers can access the Quality Score from within the Bing Ads reporting interface:
Figure 12-4 Bing Ads Quality Score
To help advertisers better optimize their landing pages, Bing Ads provides specific feedback on landing page relevance and landing page user experience from search users’ perspectives. In general,