AdWords - Quality Score (QS)
- AdWords – Quality Score
- Impact & Importance of Quality Score
- Quality Score & Ad Rank
- Analysing Quality Score
- Aggregating Quality Score
- Components of Quality Score
- Dealing With Quality Score Issues
- Ways to Improve Quality Score
Quality Score is represented by a number between 1 & 10 and can be seen when viewing a keyword report (assuring you have selected the Quality Score column to show) within AdWords.
10 is the best score possible and 1 is the worst score possible. The higher the quality score, the less you pay per click (CPC) and the better your AdRank (position within the paid listings).
QS is a very important metric in all AdWords accounts due to its influence on CPC and AdRank, there are many aspects to this and we cover them all in this guide to quality score!
Google created this attribute to create a user-end monetary motivation to create more relevant ads by penalising less relevant ads. We cover a lot in this article including the difference between display and search, CPC and CPM Quality Scores, but typically unless specified otherwise we are referring to search network keyword CPCs.
Quality Score effects keywords on an individual level but having an account with a very low average Quality Score can have wider implications, dragging the whole account performance down. Thus QS impacts all levels of the account in addition to CPC.
CPC is particularly important because of its relationship to other metrics, for example; CPC effects total Cost and Clicks proportionally. If you have a budget that is being reached, higher CPC’s will mean fewer Clicks as each Click costs more. If your budget is not being reached, higher CPC’s will increase the total cost.
Higher costs mean lower ROI (Return on Investment) which is often the primary objective of this type of marketing. Whether its lower profits due to fewer sales or increased costs, the end result is poorer performance.
Quality Score is also an indication of other problems in the account such as poor CTR (Click Through Rate) and a lack of relevance for the user between what they want and what you are showing them. These are causes rather than symptoms of poor quality score, but all contribute heavily to poor account performance.
If a keyword has a QS too low, Google will prevent the keyword from triggering ads, in this case you have a critical issue with a keyword that needs to be addressed.
Overall poor quality scores will affect the starting QS of new keywords, increasing the initial period of inefficiency and the magnitude of inefficiency.
Poor QS can also make keywords too expensive to target or too expensive to target in good positions, limiting access to valuable customers.
A high average QS on keywords will allow you to bid less and achieve higher positions than competitors for the same search term and in the same auction. Reducing CPC’s can increase ROI, profit and ultimately allow you to get more customers.
This lends a huge competitive advantage to businesses with large or small spends. Paying less for the same traffic as a competitor means higher volumes, better margins and ROI all of which are valuable to businesses.
Around 60% of the QS is comprised of historical data such as CTR based on the AdRank, in addition to other components. We discuss each of these components in more detail later in this guide.
Quality score on the Search Network is a composite metric, combining the perceived relevance between a keyword an ad and a landing page. As mentioned historical CTR accounts for 60% of the score.
AdRank is calculated thusly:
CPC Bid x Quality Score
CPM campaigns are a lot less common, we mention them below but for the purposes of the rest of this guide we make very little mention of them.
The historical performance is measured in a very similar way with the exception that CTR is calculated by placement and the site in addition to an aggregate across all placements. There is often only one position for a display ad to appear in on a page so position is not as relevant. But lower QS will lead to losing this position to a competitor.
AdRank is calculated thusly:
Display Network CPC Bid x Quality Score
If you are using a CPM campaign on the Display network, CTR is not necessarily a good measure of its relevance as your primary aim is to drive Impressions and not Clicks. In this case the landing page is the primary factor in determining QS
When looking at the QS of keywords there are a number of ways you can cut or analyse the data. First let’s look at what action should be taken, if any at an individual keyword level:
10 = This is as high and as good as it can be, good work you!
9 = Still very high, typically does not require much attention and unless this is the lowest score in the account focus your attention elsewhere.
8 = A strong QS
7 = Above average QS, we would be looking to improve this on valuable keywords where possible.
6 = This is really the lowest QS you should accept. This will definitely need improving, not just on valuable keywords but on all keywords.
5 = Under average, this will still allow keywords to trigger ads but should be addressed no matter how valuable the keyword.
4 = At this level CPC will be massively inflated and the keyword is well under the minimum acceptable level. These keywords should be addressed quickly.
3 = This is as low as the QS can be on a keyword before potentially having AdWords stop it from triggering ads. These keywords should be addressed Immediately.
2 = Keywords may not be eligible to appear when QS is this low, these keywords should be paused or removed until the underpinning problem can be addressed.
1 = Keywords will not be eligible to appear when QS is this low, these keywords should be paused or removed until the underpinning problem can be addressed.
It may not be possible to get a QS of 10 for everything in an account, even mature accounts, but this should be the ultimate goal. In more competitive industries, smaller players have to pick their battles. For example, brands that advertise on major TV networks often see inflated CTR for certain terms that people relate to their brand. Brand awareness is a major factor in what people click on.
Thus if your brand cannot compete on this level there may be external limiting factors in how good your CTR will be in any position compared to some bigger brands.
When auditing an account or reviewing quality score we typically analyse the data to show any of the following ways... It should be noted that we would look at QS in some to all of these ways but always by network. Comparing QS across networks muddies the waters as they are calculated differently and should be dealt with separately:
- Average (Mean) QS of all keywords
- Average (Mean) QS of keywords by campaign
- Average (Mean) QS of keywords by ad group
- Average (Mean) QS of keywords by product or service
- Average (Mean) QS of keywords by landing page
- Distribution of QS (no. of keywords with QS10, QS9, etc) of all keywords
- Distribution of QS by campaign
- Distribution of QS by product or service
- Distribution of QS by landing page
- No. of keywords with QS under 6 - All keywords
- No. of keywords with QS under 6 - By campaigns
- No. of keywords with QS under 6 - By product or service
- No. of keywords with QS under 6 - By landing page
Analysing the QS of keywords with a structured approach such as this will provide a greater level of insight. Identifying what campaign, ad groups, products / services, & landing pages have low or high QS can direct your attention to areas with systemic problems. Finding out what is working is just as important as finding out what it not working.
Using the techniques above will provide an idea of what the priorities should be in a campaign with regards to what to address first and how severe the problem is. Aggregating data into a single location (spreadsheet or tool) will speed things up, so pulling landing pages, ads and keywords & CTR into a table that can be filtered or pivoted helps when analysing large volumes.
Segmenting this into priority order enables you to manage the job in bite size chunks. Segment the data further into stuff that’s working, stuff that’s not working and stuff that’s really not good.
Interrogate this data to identify the root cause/s of the problem... Is there one landing page for 1,000 varied keywords, is the account structured to facilitate tight clustering of keywords with dedicated ad copy? Once you have identified the reasons for the poor quality scores you can begin to work to fix them.
Finding out what is working, can provide a guide as to what needs to be done for those that are not working. Additionally, there may be scope for expansion in the areas of high QS, increasing budgets to capitalise on the optimal cost for the traffic.
We always say that a single metric is usually meaningless outside of the context of other data. A CTR of 1% doesn’t mean anything unless you know other facts like what network it’s on, what position the ad is in, etc.
When looking at averages, an average doesn’t mean much if there is only one keyword! In smaller campaigns with a handful of keywords, there is no need to perform such an analysis.
As we demonstrate above section there are several ways we can aggregate data up to campaign level or by product. Google is suspected of doing something similar, although we should note that this has not been confirmed by Google. We discuss the below topics both from the perspective that is speculated to be Google’s view; in addition to how we can learn something practical from this regardless of whether it exists.
We also cover a number of different ways that data can be aggregated or segmented when being analysed and the benefits of doing this.
The net is alive with discussion about account level quality score and whether it exists or not. Google has never confirmed the existence of this but various bloggers have put forward good cases for its existence. It is likely that there is some kind of overarching score for quality that each account gets even if it is not viewable by users.
It is proposed that an account level quality score will be based on an aggregate of historical account performance and effects the QS of new keywords when added to an account.
Regardless of whether this exists as a thing on Google’s end, the implications of having an account rife with low quality keywords is something that should be addressed. Looking at the average QS throughout an account is a moderately useful top-level view of the QS component.
This isn’t something that we are saying Google has but we do not, by looking at QS by campaign can quickly highlight what campaigns might need more attention. Thus segmenting QS data by campaign can provide a fairly top-level overview.
Ad group level quality score is more granular that campaign but again is not something we are suggesting Google has a metric for. If you have only a few related keywords per ad group a low QS may mean poor ad copy or landing page. If the ad groups have lots of keywords, the problem may be more to do with structure and an intrinsic lack of relevance between a large group of words and ad copy.
We have touched upon each of the components of what Google factors into QS, we explore these in more depth here.
Things to check are whether the keyword is used within the ad copy. There are several places that this can be done, such as the title, description, and the display URL before or after the domain. This becomes more possible when the keywords in the ad group are very similar.
This is not possible if the keyword targets a trademarked term or certain other restricted categories such as drugs; because you cannot use these terms within the ad copy. Thus targeting these terms often inevitably leads to poor or under average QS.
There is no need to stuff the keyword into every possible field in the ad copy, but having the keyword mentioned once or twice can improve relevance significantly. Ultimately how relevant is your ad copy to the intent behind the searches that trigger it, is something you need t consider.
There are various places on a landing page where using the keyword will help to build relevance, such as:
- Page title
- Header tags (H1, H2, etc)
- Meta description
- Content on the page
- Images (file names, alt tags, etc)
Having the keyword in all or most of these locations will help to improve relevance between the keyword and the landing page. We discuss landing page optimisation later in this guide and in another article.
There are a number of things to consider here, such as:
- Offers used in ad copy should be present on the landing page
- If a specific cost, price, discount, rate, etc is used in the ad copy, it should be present on the landing page
- Ad copy should be relevant to the landing page, consider using the ad copy in the Meta description in addition to the main body of content
Ensuring that you are not promising in an ad something you fail to mention on the landing page is pivotal.
Landing pages should have enough content to build relevance to the target keyword and also enough for the user to find out what they need to know. Content does not have to be unique if using dedicated SEM / PPC landing page, as they can be easily removed from Google’s index. We discuss this in more detail later on.
Click Through Rate (CTR) is the primary factor here, but also the volume of data needs to be sufficient for Google to consider the CTR history. Impressions are needed in the thousands of the Search Network and hundreds of thousands to millions in the Display network. This is referred to as the impression threshold.
So small budgets will limit the speed with which QS will improve. Whereas larger budgets allow for faster data gathering and a faster road to optimal performance.
It is likely that overall historical data is used when issuing an initial QS for new keywords.
The CTR is relative to the position the ad is in, for example the average CTR of ads in position 4 for a particular search query might be 5%; your ad may achieve 6% on average and after enough data has been gathered to come to a statistically significant position. Thus Google may keep you in that position or try you in the next spot above to see if you outperform the existing ad.
If your campaign allows for mobile traffic then your landing pages should be able to accommodate said traffic. Providing a poor user experience due to non-mobile, non-responsive or adaptive landing pages can impact overall QS. Google allow you to adjust bids for mobile devices and they allow you to prevent mobile traffic, this is a strong indication that they want you to factor this in.
There are a lot of potential causes and problems that can lead to poor quality score, we set out here the main ways that you can deal with them.
The choice is yours, we see no compelling arguments that one is better than the other, rather logic should dictate what you do. If you are planning on reactivating the keyword after work has been finished to improve things, then pausing seems legitimate. If you have no intention of reactivating the keyword, deleting / removing it also equally reasonable.
We cover this in more detail below.
The solution should obviously match the problem, if the issue is that there are too many irrelevant search terms triggering ads, optimising the landing page for 200 irrelevant terms is not the solution! Below we provide the solution given a specific problem or set of problems.
If the keyword is triggering a lot of impressions and not a lot of click, there is good reason to suspect that there may be a lot of irrelevant search terms triggering ads. Also if you notice irrelevant terms in the search term report, this would be a strong indication of this issue. Typically if phrase or broad match keywords are used and there are few to no negative keywords, this will be an issue.
Adding negative keywords and performing negative keyword research to identify those keywords can massively improve CTR and traffic quality. This will be factored into the QS once the impression threshold has been reached.
Downloading and reviewing search term reports can also highlight potential keywords that need to be added to the ad group or campaign as negative keywords. This will prevent future wasted budget and irrelevant searches triggering ads.
Identify search terms of high value that do not exist in the account, adding these and targeting them with custom ad copy will help to improve the performance of the account.
Often the biggest problem is account structure, having keywords grouped in ad groups in large numbers with no structure will inevitably lead to poor QS. Structuring an account with no more than a few closely related keywords in each ad group, facilitates much more targeted ad copy.
This has many other benefits in terms of effective account management, reporting, analysis, etc.
There is a whole suite of extensions that are available to AdWords users, all of which are designed to target people in different ways while making the ad stand out more from the page. Using ad extensions can make your ad more prominent and thus drive up CTR.
We have listed the extensions below, but we cover each of this in detail in another guide. For now it is simply worth mentioning that using these can improve CTR.
- Structured Snippets
We mentioned earlier that there are a number of places on a page that can be optimised to improve relevance. We cover this in more detail below, in addition we also look at wider landing page strategies that can drive up Quality Score. We look at landing pages in more detail in another article.
Improve Existing Landing Pages
Depending on whether you are using dedicated stand-alone SEM landing pages or whether you are using your standard website pages will affect what you can do here. If standard web pages are being used and are already optimised to target organic keywords for SEO, you will not want to go through and change all of that.
If however, these pages are either not optimised for SEO or are dedicated SEM landing pages you can look to make the following changes to improve the landing page’s relevance.
Page title: Using the keyword within the page title is a powerful tool for increasing the relevance of the landing page to the target keyword. If two or three keywords deliver traffic to one landing page, consider using all of them in the page title. Page titles should not exceed 70 characters, for more information on this follow the link.
URL: Using the primary keyword within the URL is ideal, stuffing more than one in is not good. URL’s should not be spammy, but reinforce the relevance to both the user and Google. For more information on URL’s follow the link.
Header tags (H1, H2, H3, etc) : Header tags provide an opportunity to use variations of keywords, using the most important in the H1 and so on down the scale. For more information on Header Tags follow the link.
Meta description: Using both the keyword/s and the ad copy in the meta description can also improve relevance between the ad > page & keyword > page. For more information on Meta Descriptions follow the link.
Content on the page: Often we find that this is the biggest problem on a landing page, generic content or a complete lack of content. This is where any offers of discounts, prices, figures, etc should be mentioned if used in the ad copy. The content on this page should be compelling, address the user intent behind their search, and be designed to convert the user.
Ensure the CTA (Call to Action) is prominent on the page and ‘above the fold’. This should have your best sales message and be designed for the user.
For more information on Content or conversions follow the link.
Images (file names, alt tags, etc) : These are a powerful tool when selling a product or service, they assist users in making decisions and can be optimised in a number of ways to target keywords. For more information on Images follow the link.
Video: Videos are known to assist in converting users, they are great for conveying information and selling products. Moreover they can be optimised to target keywords. For more information on Videos follow the link.
Internal Linking: Linking the landing to other relevant content on your site can also build relevance, ideally the desired content should be within a click of the landing page. For more information on Internal Linking follow the link.
Custom SEM Landing Pages
A very effective strategy that has potential crossover with SEO is to build out a whole range of custom SEM or PPC landing pages for each ad group or campaign, depending on size and scale. Building dedicated and custom landing pages will provide a better user experience and improve QS.
This can be approached in a number of ways, but a good and efficient way to roll out a strategy like this will be to; roll out landing pages with the minimum level of customisation, then over time (in order of priority) go back through these pages adding unique content and allowing them to be indexed. This provides a benefit to both the SEM and SEO endeavours.
This tactic is relevant to accounts of all sizes, small accounts can benefit without a huge amount of work, whereas large accounts can see significant improvements in ROI as a result.
Split testing (testing two ads at once) and multi-variant testing (testing more than two ads at once) can improve almost any metric including CTR. Although optimising for CTR isn’t always the priority, when CTR is very low it should be addressed.
Performing these tests is an intrinsic part of account management and the ongoing optimisation thereof. This concept can also be applied to landing pages, but this isn’t necessary for directly improving Quality Score.
Improving any component with the aim of driving CTR will take time to have an impact in most accounts. This is because more historical data needs to be collated and reviewed before the QS will be adjusted. Collating new data will take longer the smaller the budget, so improvements may take days to weeks before being noticed.