Google recently released a white paper on Quality Score, aimed at dispelling myths and rumours and highlighting factors everyone should look for. I’ve taken the time to pull a number of key points from it, translating them below into what I think they mean:
The “Big-Three” component parts of quality score (Page 3)
I. Ad Relevance – the relevancy of your ad to the user query is important, but also consider the relevancy of your ad to the selected landing page, and the relevancy of query to the landing page. Understanding the relationship between these 3 major components of relevancy will ensure you maximise this aspect of quality score.
II. Expected CTR – it has long been understood that historical performance is considered when calculating quality score. In this case, historical CTR will make up the bulk of the calculation for expected CTR, in which case, it’s important to address areas of poor performance as soon as possible. Allowing keywords to collect poor data over time will only work to impede any future improvements you make.
III. Landing page experience – with more and more importance being placed on user experience at Google, you can be sure that the emphasis on landing-page experience is growing. Therefore, providing a landing page that is transparent to the user and in line with the ad that qualified that user is also getting more important. Conversion rate optimisation and usability testing are a must for new and existing websites.
Quality Score Is a Helpful Diagnostic Tool, Not a Key Performance Indicator (Page 3)
The quality score value reported in Adwords should be seen as an average metric on how effective an advertiser you have been on that term, over the period of time you have been appearing for it, not as the Quality Score value used in each auction. The fact that this rating can be vastly different from what you’re currently being awarded in auctions needs to be understood.
The various components that make up quality score can have a range of effects on the quality score value calculated in an auction: a change to a more relevant landing page or improvement of an existing landing page may result in improvements to the current quality score calculated, but it will take time to reveal the effect this has on the reported quality score.
A keyword has been ranked a quality score of 6 for each of 60,000 impressions over 12 months.
After improvements to the site and ad text, relevancy increase and expected CTR improve the quality score. Quality score is now rated 8.
After 1 month and 10,000 impressions at quality score of 8, the average quality score is now 6.29.
Showing it takes time for this reported metric to change.
Rather than look to this for improvements, consider monitoring average position and average CPC when changes are being made that could affect quality score.
Focus your efforts on High-Value areas where you can affect change (Page 3)
All Paid Search Managers should be looking to maximise their return for lowest minimum spend. The majority of this will be in relation to revenue and ROI, or conversions and CPA. What this point prompts, however, is an advertiser to think beyond optimising bids and budgets to improve cost efficiency, and consider the affect that landing page performance, ad relevancy and ad structure is having on the auction-calculated quality score, in particular, where areas of a campaign offer a big opportunity for improvements in cost efficiency. Working to improve a landing page can result in major improvements in cost efficiency through a combination of subtle improvements in relevancy, conversion rate, bounce rate and more.
Where to make changes?
Changes should be based on both the targets of your account and on the projected effect any changes to quality score will have on that particular part of an account/campaign, in relation to your targets.
Highlighting where quality score can be improved:
Consider using page statistics such as conversion rate, bounce rate and time on page to highlight pages that are performing below average, consider improving either the selection of landing page or the page itself.
How you Structure your account doesn’t matter (page 10)
This is a confusing point for Google to raise, particularly since they referred to ways you can improve ad relevance (page 7) with:
‘Move Keywords to smaller ad groups with more targeted relevance’
‘Look for ad groups with disparate keywords that can’t possibly be addressed by the same ad, and give below-average keywords their own, new ad group…’
Both of which suggest structure of an account is important… granted they may be referring to the fact that it isn’t directly considered in the calculation of quality score, but the indirect impact it has on quality score is obvious.
Ensure your campaigns are structured efficiently, with keywords grouped into tightly themed adgroups, with highly relevant ad text and equally relevant landing pages, and you won’t go far wrong.
Negative effects on metrics leading to positive effects on Quality Score (Page 8)
Halfway down the page, Google say:
‘As we mentioned about ad relevance, there may be times when a more specific ad leads to lower CTRs but higher conversion rates. Don’t assume one is always better than the other.’
Here, the importance of taking all available metrics into consideration is highlighted. In this example, the more specific ad text worked to qualify users, leading to a better experience and consistency between ad text and landing page, meaning overall a better user experience despite the drop in CTR.
Taking all metrics into account in such a way is important when understanding the actual performance of an account and the reasons behind any changes in performance.
You should work to give the user what they are looking for, and if they aren’t looking for what you offer, don’t offer it!
For Newly Launched Keywords, Performance on Related Keywords Does Matter (Page 9)
For me, this is a salute to conspiracy theorists claiming that the existence of unconfirmed quality scores, those being that of Campaigns and Adgroups. It indicates that there is a relationship between keywords in an account, in terms of quality score.
Though it refers to newly launched keywords only, we have undertook research into the effect of including high search volume brand (highly relevant) keywords on the whole account. The findings showed that, not only did the existence of highly relevant, high-search-volume keywords with strong performance metrics improve the quality scores of keywords around it, they also improve cost efficiency of the Non Brand Campaigns so much as to effectively negate the cost of the brand campaign, in doing so, negating the ever present argument of ‘Why should I pay for Brand traffic when I’m in top position organically?’