Reverse-engineering AdWords Quality Score factors

Having a high Quality Score is crucial when it comes to maximizing AdWords performance. However, it’s been very difficult to work with Quality Score in the past, as Google has not easily allowed you to diagnose all of the factors that comprise Quality Score at scale.

A recent change to Google’s API has now made this possible, so I’d like to share some valuable insights into Quality Score factors and scale-based diagnosis.

Why a high Quality Score is important

Quality Score, along with your bid and ad extension data, is how Google determines your Ad Rank. Quality Score is the second-largest component of Ad Rank, with only your bid being more important.

When you raise your Quality Score, you can expect one of the following to happen:

• Your costs-per-click (CPCs) decrease.
• Your average position increases.
• Your CPCs decrease, and your average position increases.

It is also possible to see your average position increase and your average CPC increase, as you’re now beating a different company in the auction, and thus your actual CPC can vary.

Overall, when you can’t raise your bid but you want to raise your average position, then you need to increase your Quality Scores.

The Quality Score factors

There are three main factors that make up Quality Score:

• click-through rate (CTR),
• ad relevance and
• landing page experience.

In the past, it has always been very difficult to see these factors for every keyword, as you had to hover over every keyword individually to see the Quality Score factors.

Reverse-engineering the Quality Score factors

Google recently launched a new version of its API that allows someone to download each Quality Score factor, along with its status (average, below average, above average) for every search keyword in an account. This means we can now analyze at scale exactly how Google weights the subfactors and determines visible Quality Score.

At AdAlysis (my company), we have access to a tremendous amount of data. So we downloaded a lot of keywords and started to get to work.

The first step was trying to understand the visible Quality Score. After doing a lot of analysis on the factors and their visible Quality Scores, we were able to determine the formula.

To understand the formula, you need to understand the weightings. For every keyword, you can look at this table to see how many points you get based upon your Quality Score factor and how good it is:

	Landing Page Experience	Ad Relevance	CTR
Above average	3.5	2	3.5
Average	1.75	1	1.75
Below average	0	0	0

 

The formula is simply: 1 + Landing Page Experience weight + Ad Relevance weight + CTR weight.

For instance, let’s say you have these factors:

• Landing Page Experience: average — 1.75 points
• Ad Relevance: above average — 2 points
• CTR: average — 1.75 points

Then your Quality Score is: 1 + 1.75 (LPE) + 2 (Relevance) + 1.75 (CTR) = 6.5. As Google doesn’t show fractions, this is rounded to a 7, which is seen inside your account.

Therefore, if you can improve your landing page experience or your CTR to above average from average, your Quality Score would go up 1.75 points to an 8.25, which would be displayed as an 8.

Now that we know the weightings and the formula, we can also see the actual weightings by factor:

 

This means that testing for CTR increases is more important than ad testing for ad relevance increases.

Where to start increasing your Quality Scores

The way we think about improving Quality Scores hasn’t changed in a long time. This is our overall Quality Score workflow:

 

What has changed is the data available at scale. That means that third-party tools can now add items like priority lists or Quality Score factors to their interfaces.

What’s coming in third-party tools

I’ve reached out to Google (and talked to many at the Google Dance last week) to see if this additional information would be displayed in the AdWords interface or the AdWords editor anytime soon.

Overall, it didn’t sound like Google was planning on launching interface changes, but it did leave open the possibility of this information coming to the editor at some point in time. We’ll just have to wait and see if these factors ever come to the editor.

I reached out to some tool developers to see what they were working on with these new Quality Score features. None of the third-party tool providers wanted to comment on their future plans, so I can’t speak for them. However, at AdAlysis, we have introduced in-depth Quality Score analysis using this new API feature, so I can talk about what we’ve launched so you can think of the new world of Quality Score analysis possibilities. In our tool, you can do this type of analysis:

• Easily graph Quality Score data by various metrics (impressions, clicks, CTR and so on) at the account or campaign level.
  • See the overall Quality Score factors by each of these metrics (such as percentage of impressions with low landing page experience).
• Get a priority order of what ad groups need the most help based upon spend levels and weighted Quality Score data.
• Filter by Quality Scores and their factors, and then sort by metrics such as spend or impressions.

The future of Quality Score

This API change is a big one for many advertisers. Monitoring Quality Scores has always been difficult, but possible. However, getting a handle on whether your Quality Score issues were landing-page-related or CTR-related was impossible across medium to large accounts.

With these changes, digging into each area of weakness, isolating that area, viewing common threads, and then increasing Quality Scores across your entire account is now much easier.

At the moment, you need to rely on tool providers or your own internal API development to be able to do these diagnoses at scale. However, just knowing the Quality Score weightings is a help to advertisers of all sizes, so you can start to think about how to raise your Quality Scores, lower your CPCs and increase your positions, which is going to become more important with fewer ads now showing on each page.

---

Opinions expressed in this article are those of the guest author and not necessarily Search Engine Land. Staff authors are listed here.

---