Page Authority 2.0 Is Coming This Month: What’s Changing and Why

Posted by rjonesx.

Hey folks,

I'm Russ Jones, Adjunct Search Scientist with Moz, and I'm proud to announce that this month we’ll be releasing a terrific update to our metric, Page Authority (PA).

Although Page Authority hasn't attracted the same attention as its sibling metric Domain Authority, PA has always correlated with SERPs much better than DA, serving as a strong predictor of ranking. While PA has always fluctuated with changes in the link graph, we’re introducing a whole new method of deriving the score.

Learn More About Page Authority 2.0

What's changing

Long gone are the days of just counting backlinks a couple of ways and hoping they correlate well with SERPs. As Moz tends to do, we’re pioneering a new manner of calculating Page Authority to produce superior results. Here are some of the ways we’re changing things up:

The training set

In the past, we used SERPs alone to train the Page Authority model. While this method was simple and direct, it left much to be desired. Our first step in addressing the new Page Authority is redefining the training set altogether.

Instead of modeling Page Authority based on one page's ability to outrank another page, we now train based on the cumulative value of a page based on a number of metrics including search traffic and CPC. While this is a bit of an oversimplification of what’s going on, this methodology allows us to better compare pages that don't appear in the SERPs together.

For example, imagine Page A is on one topic and Page B is on another topic. Historically, our model wouldn't get to compare these two pages because they never appear on the same SERP. This new methodology provides an abstract value to each page, such that they can be compared with any other page by the machine-learned model.

The re-training set

One of the biggest problems in building metrics is not what the models see, but what the models don't see.

Think about this for a minute: what types of URLs don't show up in the SERPs that the model will use to produce Page Authority? Well, for starters, there won't be many images or other binary files. There also won't be penalized pages. In order to address this problem, we now use a common solution of running the model, identifying outliers (high PA URLs which do not in fact have any search value), and then feeding those URLs back into the training set. We can then re-run the model such that it learns from its own mistakes. This can be repeated as many times as is necessary to reduce the number of outliers.

Ripping off the Band-Aid

Moz is always cognizant of the impact the changes to our metrics might have on our customers. There is a trade-off between continuity and accuracy. With Page Authority, we’re focusing on accuracy. This may cause larger-than-normal shifts in your Page Authority, so it’s more important than ever to think about Page Authority with respect to your competitors, not as a standalone number.

What actions should we take?

Communicate with stakeholders, team members, and clients about the update

Just like our upgrade to Domain Authority, some users will likely be surprised by changes in their PA. Make sure they understand that the new PA will be more accurate (and more useful!) and that the most important measurement is relative to their competitors. We won't release a Page Authority which isn't better than the previous version, so even if the results are disappointing, understand that you now have better insight than ever before into the performance of your pages in the SERPs.

Use PA as a relative metric, like DA

Page Authority is intrinsically comparative. A PA of 70 means nothing unless you know the PA of your competitors. It could be high enough to allow you to rank for every keyword you like, or it could be terribly low because your competitors are Wikipedia and Quora. The first thing you should do when analyzing the Page Authority of any URL is set it in the proper context of its competitor's URLs.

Expect PA to keep pace with Google

Just as we announced with Domain Authority, we’re not going to launch the new PA and just let it go. Our intent is to continue to improve upon the model as we discover new and better features and models. This volatility will mostly affect pages with unnatural link profiles, but we would rather stay up-to-date with Google's algorithms even if it means a bit of a bumpy ride.

When is it launching?

We’ll be rolling out the new Page Authority on September 30, 2020. Between now and then, we encourage you to explore our resources to help you prepare and facilitate conversations with clients and team members. Following the launch of the new PA, I’ll also be hosting a webinar on October 15 to discuss how to leverage the metric. We’re so excited about the new and improved PA and hope you’re looking forward to this update too.

If you have any questions, please comment below, reach out to me on Twitter @rjonesx, or email me at russ@moz.com.

To get prepared and learn more about the upcoming change to Page Authority, be sure to dig into our helpful resources:

Visit the PA Resource Center

Sign up for The Moz Top 10, a semimonthly mailer updating you on the top ten hottest pieces of SEO news, tips, and rad links uncovered by the Moz team. Think of it as your exclusive digest of stuff you don't have time to hunt down but want to read!

Page Authority 2.0 Is Coming This Month: What’s Changing and Why

Posted by rjonesx.

Hey folks,

I'm Russ Jones, Adjunct Search Scientist with Moz, and I'm proud to announce that this month we’ll be releasing a terrific update to our metric, Page Authority (PA).

Although Page Authority hasn't attracted the same attention as its sibling metric Domain Authority, PA has always correlated with SERPs much better than DA, serving as a strong predictor of ranking. While PA has always fluctuated with changes in the link graph, we’re introducing a whole new method of deriving the score.

Learn More About Page Authority 2.0

What's changing

Long gone are the days of just counting backlinks a couple of ways and hoping they correlate well with SERPs. As Moz tends to do, we’re pioneering a new manner of calculating Page Authority to produce superior results. Here are some of the ways we’re changing things up:

The training set

In the past, we used SERPs alone to train the Page Authority model. While this method was simple and direct, it left much to be desired. Our first step in addressing the new Page Authority is redefining the training set altogether.

Instead of modeling Page Authority based on one page's ability to outrank another page, we now train based on the cumulative value of a page based on a number of metrics including search traffic and CPC. While this is a bit of an oversimplification of what’s going on, this methodology allows us to better compare pages that don't appear in the SERPs together.

For example, imagine Page A is on one topic and Page B is on another topic. Historically, our model wouldn't get to compare these two pages because they never appear on the same SERP. This new methodology provides an abstract value to each page, such that they can be compared with any other page by the machine-learned model.

The re-training set

One of the biggest problems in building metrics is not what the models see, but what the models don't see.

Think about this for a minute: what types of URLs don't show up in the SERPs that the model will use to produce Page Authority? Well, for starters, there won't be many images or other binary files. There also won't be penalized pages. In order to address this problem, we now use a common solution of running the model, identifying outliers (high PA URLs which do not in fact have any search value), and then feeding those URLs back into the training set. We can then re-run the model such that it learns from its own mistakes. This can be repeated as many times as is necessary to reduce the number of outliers.

Ripping off the Band-Aid

Moz is always cognizant of the impact the changes to our metrics might have on our customers. There is a trade-off between continuity and accuracy. With Page Authority, we’re focusing on accuracy. This may cause larger-than-normal shifts in your Page Authority, so it’s more important than ever to think about Page Authority with respect to your competitors, not as a standalone number.

What actions should we take?

Communicate with stakeholders, team members, and clients about the update

Just like our upgrade to Domain Authority, some users will likely be surprised by changes in their PA. Make sure they understand that the new PA will be more accurate (and more useful!) and that the most important measurement is relative to their competitors. We won't release a Page Authority which isn't better than the previous version, so even if the results are disappointing, understand that you now have better insight than ever before into the performance of your pages in the SERPs.

Use PA as a relative metric, like DA

Page Authority is intrinsically comparative. A PA of 70 means nothing unless you know the PA of your competitors. It could be high enough to allow you to rank for every keyword you like, or it could be terribly low because your competitors are Wikipedia and Quora. The first thing you should do when analyzing the Page Authority of any URL is set it in the proper context of its competitor's URLs.

Expect PA to keep pace with Google

Just as we announced with Domain Authority, we’re not going to launch the new PA and just let it go. Our intent is to continue to improve upon the model as we discover new and better features and models. This volatility will mostly affect pages with unnatural link profiles, but we would rather stay up-to-date with Google's algorithms even if it means a bit of a bumpy ride.

When is it launching?

We’ll be rolling out the new Page Authority on September 30, 2020. Between now and then, we encourage you to explore our resources to help you prepare and facilitate conversations with clients and team members. Following the launch of the new PA, I’ll also be hosting a webinar on October 15 to discuss how to leverage the metric. We’re so excited about the new and improved PA and hope you’re looking forward to this update too.

If you have any questions, please comment below, reach out to me on Twitter @rjonesx, or email me at russ@moz.com.

To get prepared and learn more about the upcoming change to Page Authority, be sure to dig into our helpful resources:

Visit the PA Resource Center

Sign up for The Moz Top 10, a semimonthly mailer updating you on the top ten hottest pieces of SEO news, tips, and rad links uncovered by the Moz team. Think of it as your exclusive digest of stuff you don't have time to hunt down but want to read!

Title Tags SEO: When to Include Your Brand and/or Boilerplate

Posted by Cyrus-Shepard

If your websites are like most, they include a fair amount of extra "stuff" in the title tags: things like your brand name or repeating boilerplate text that appears across multiple pages.

Should you include these elements in your titles automatically?

To be fair, most sites do.

Alternatively, could it help your SEO to actually include less information in your titles? (Or at least in specific circumstances?)

We know from a handful of studies that titles of a certain length tend to perform better. A now-famous study from the engineers at Etsy showed how shorter titles performed better than longer ones. SEOs speculate that this could be because shorter titles can have more focused relevancy (by focusing on core keywords), might earn higher click-through rates, or some other reason we can't imagine.

When choosing which part of a title to shorten, brand names and boilerplate text are obvious choices. But how do you determine if this is something you should consider for your own SEO?

Here's an example of a brand's site name at the end of every title:

We've all seen sites like this. Heck, most of us do this on our own sites. The question is, does having our brand/site name at the end of every title actually help, or hurt?

But first, we also have to consider other types of boilerplate.

What is boilerplate? Boilerplate simply means standardized, non-unique pieces of text that are used over and over again. This often includes things like categories, product categories, author tags, and taglines.

In this example below, the boilerplate text on every title includes "Tomatoes - Vegetable Seeds - Shop."

Sometimes boilerplate material can become quite long. The comic book review site Major Spoilers (awesome name!) often includes the same 65-character boilerplate on many pages:

"Major Spoilers – Comic Book Reviews, News, Previews, and Podcasts"

Of course, at this length, it's so long that Google truncates every single title:

The problems that boilerplate can cause your SEO are threefold:

1. Relevancy: Unnecessary words can make your title less relevant, both to search engines and users. For search engines, this could mean lower rankings. For users, this could result in fewer clicks.
2. Uniqueness: Titles that share the same repeating text, and only vary from one another by a word or two, aren't very unique. While this isn't necessarily a problem, it goes against most SEO best practices, where uniqueness is key.
3. Length: Boilerplate means you have less room to display other words in your title, and Google will often cut these off if they go beyond a certain length.

Experiment #1: Remove category from title

We decided to run a couple of boilerplate experiments here at Moz, to see if we could increase our rankings and traffic by removing some of the repeating parts of our titles.

We started with our Whiteboard Friday blog posts. Every time Moz publishes a new Whiteboard Friday, we traditionally include "Whiteboard Friday" in the title.

What would happen if we removed this from the titles?

Using an A/B split test methodology — where we rolled the test out on 50% of the titles and used the other 50% as a control — we saw an amazing 20% uplift from this experiment.

This chart represents the cumulative impact of the test on organic traffic. The central blue line is the best estimate of how the variant pages, with the change applied, performed compared to how we would have expected without any changes applied. The blue shaded region represents our 95% confidence interval: there is a 95% probability that the actual outcome is somewhere in this region. If this region is wholly above or below the horizontal axis, that represents a statistically significant test.

Honestly, the results surprised us. Whiteboard Friday is a popular brand (so we thought) but removing this boilerplate from our titles produced a significant uplift in traffic to those pages.

At this point, we got cocky…

Experiment #2: Remove brand from title

If removing the category name from Whiteboard Friday posts produced such a significant uplift, what if we removed our brand name from all titles?

For this A/B experiment, we did exactly that—removing the word "Moz" from 50% of our titles and measuring the results.

Crazy, right? If it worked by removing "Whiteboard Friday" would we see the same uplift by removing "Moz?"

Sadly, Google had other plans:

While this A/B test never reached full statistical significance, we actually saw a 4% decline in traffic by removing our brand from our title tags.

Boo!

So why did this test not produce the same gains? To be honest, I've removed the brand name from other site's titles and seen as much as a 20% uplift.

It turns out that whether or not removing brand/boilerplate will be beneficial to your SEO depends on a few key factors, which you can gauge in advance.

How to know if removing boilerplate may succeed

Over 10 years of experience and literally millions of title tags, I've found that there are basically four factors that influence whether or not removing boilerplate from your titles might be beneficial:

1. Brand Strength: Popular brand names in titles almost always perform better than unknown brands, even when people aren't searching for your brand specifically. Amazon's brand recognition, for example, likely gives a significant boost to including "Amazon" in every title, even when people aren't specifically searching Amazon. Less recognizable brands, however, don't always get the same boost, and can actually lead to fewer visits based on relevancy, length, and clickability (described next.)
2. Relevancy: Are your boilerplate/brand keywords relevant to what your users search for? For example, if you're site is about television repair, then boilerplate titles that say "Brad's TV Repair" are going to be much more relevant than boilerplate that simply say "Brads." (We'll explore a way to determine your boilerplate's brand strength and relevancy in the next section.)
3. Length: In general, long boilerplate has the potential to do more harm than short boilerplate/brand words. Long boilerplate can dilute the relevance of your titles. So if you include "Buy Brad's TVs, Television Repair, High Definition Servicing, Audio and Visual Equipment for Sale in Houston Texas and Surrounding Areas" - you may want to rethink your boilerplate.
4. Clickability: Sometimes, boilerplate can make your titles more clickable, even if they aren't terribly relevant. Words like "Sale", "Solved", "Free", "2020", "New", and many others can lead to an increase in click-through rates (CTR.) Sometimes you can't tell until you test, but in many cases even adding clickable elements to your boilerplate can lead to significant gains.

Simple technique for determining your brand strength and boilerplate relevancy

This simple technique will also show why removing "Whiteboard Friday" led to an increase in traffic while removing "Moz" from titles did not.

Here's what you want to do: for each piece of boilerplate, determine the number of URLs on your site that rank/receive traffic for those keywords.

For this, we'll use Google Search Console.

Simply enter your boilerplate/brand as a query filter (you may need to break it into chunks for longer boilerplate) and see how many URLs receive traffic for queries that include that keyword.

When we filter for keywords that contain our "moz" brand name, we find thousands of ranking URLs.

People are searching for things like:

• Moz DA Checker
• Moz Pro
• Moz SEO
• Moz Blog
• Etc., etc.

As our brand name is part of so many queries and leads to visits across thousands of pages, this tells us that "Moz" is a very strong brand, and we'd likely be smart to include it as part of our title tags.

"Moz" is also very short at only 3 characters, which doesn't hurt either.

So what happens when we try this same technique with "Whiteboard Friday" — the boilerplate that led to a 20% uplift when we removed it? We see a very different result:

In this case, almost all the traffic for "Whiteboard Friday" search terms goes to only one or two pages.

For most Whiteboard Friday posts, the term is simply irrelevant. It's not what people are searching for, and the brand isn't strong enough to produce additional uplift.

Also, at 17 characters long, this boilerplate added significant length to each of our titles, in addition to possibly diluting the relevancy for what the posts were ranking for.

Final thoughts + bonus free title tag webinar

These tips can't tell you definitively whether you should or shouldn't include boilerplate or brand in your title tags, but they should give you a pretty good idea of when you should test things out.

Remember: Always test and evaluate before making any SEO change permanent. At least know the impact of the change you are making.

Also, please don't be under the impression that you should always remove boilerplate from your titles. In some instances, actually adding boilerplate can produce an uplift, particularly when the boilerplate is:

1. Recognizable: For example a strong brand
2. Relevant: The right keywords
3. Clickable: Encourages a high CTR
4. Succinct: Not overly long

---

If you found value in the tips, and want to learn even more ways to optimize your title tags, we've made available a free webinar for you: SEO Master Class: Advanced Title Tag Optimization (For Any Site).

If you've got 40 minutes, it's definitely worth a watch.

Watch Free Webinar

Best of luck with your SEO!

Sign up for The Moz Top 10, a semimonthly mailer updating you on the top ten hottest pieces of SEO news, tips, and rad links uncovered by the Moz team. Think of it as your exclusive digest of stuff you don't have time to hunt down but want to read!

Page Authority 2.0 Is Coming This Month: What’s Changing and Why

Posted by rjonesx.

Hey folks,

I'm Russ Jones, Adjunct Search Scientist with Moz, and I'm proud to announce that this month we’ll be releasing a terrific update to our metric, Page Authority (PA).

Although Page Authority hasn't attracted the same attention as its sibling metric Domain Authority, PA has always correlated with SERPs much better than DA, serving as a strong predictor of ranking. While PA has always fluctuated with changes in the link graph, we’re introducing a whole new method of deriving the score.

Learn More About Page Authority 2.0

What's changing

Long gone are the days of just counting backlinks a couple of ways and hoping they correlate well with SERPs. As Moz tends to do, we’re pioneering a new manner of calculating Page Authority to produce superior results. Here are some of the ways we’re changing things up:

The training set

In the past, we used SERPs alone to train the Page Authority model. While this method was simple and direct, it left much to be desired. Our first step in addressing the new Page Authority is redefining the training set altogether.

Instead of modeling Page Authority based on one page's ability to outrank another page, we now train based on the cumulative value of a page based on a number of metrics including search traffic and CPC. While this is a bit of an oversimplification of what’s going on, this methodology allows us to better compare pages that don't appear in the SERPs together.

For example, imagine Page A is on one topic and Page B is on another topic. Historically, our model wouldn't get to compare these two pages because they never appear on the same SERP. This new methodology provides an abstract value to each page, such that they can be compared with any other page by the machine-learned model.

The re-training set

One of the biggest problems in building metrics is not what the models see, but what the models don't see.

Think about this for a minute: what types of URLs don't show up in the SERPs that the model will use to produce Page Authority? Well, for starters, there won't be many images or other binary files. There also won't be penalized pages. In order to address this problem, we now use a common solution of running the model, identifying outliers (high PA URLs which do not in fact have any search value), and then feeding those URLs back into the training set. We can then re-run the model such that it learns from its own mistakes. This can be repeated as many times as is necessary to reduce the number of outliers.

Ripping off the Band-Aid

Moz is always cognizant of the impact the changes to our metrics might have on our customers. There is a trade-off between continuity and accuracy. With Page Authority, we’re focusing on accuracy. This may cause larger-than-normal shifts in your Page Authority, so it’s more important than ever to think about Page Authority with respect to your competitors, not as a standalone number.

What actions should we take?

Communicate with stakeholders, team members, and clients about the update

Just like our upgrade to Domain Authority, some users will likely be surprised by changes in their PA. Make sure they understand that the new PA will be more accurate (and more useful!) and that the most important measurement is relative to their competitors. We won't release a Page Authority which isn't better than the previous version, so even if the results are disappointing, understand that you now have better insight than ever before into the performance of your pages in the SERPs.

Use PA as a relative metric, like DA

Page Authority is intrinsically comparative. A PA of 70 means nothing unless you know the PA of your competitors. It could be high enough to allow you to rank for every keyword you like, or it could be terribly low because your competitors are Wikipedia and Quora. The first thing you should do when analyzing the Page Authority of any URL is set it in the proper context of its competitor's URLs.

Expect PA to keep pace with Google

Just as we announced with Domain Authority, we’re not going to launch the new PA and just let it go. Our intent is to continue to improve upon the model as we discover new and better features and models. This volatility will mostly affect pages with unnatural link profiles, but we would rather stay up-to-date with Google's algorithms even if it means a bit of a bumpy ride.

When is it launching?

We’ll be rolling out the new Page Authority on September 30, 2020. Between now and then, we encourage you to explore our resources to help you prepare and facilitate conversations with clients and team members. Following the launch of the new PA, I’ll also be hosting a webinar on October 15 to discuss how to leverage the metric. We’re so excited about the new and improved PA and hope you’re looking forward to this update too.

If you have any questions, please comment below, reach out to me on Twitter @rjonesx, or email me at russ@moz.com.

To get prepared and learn more about the upcoming change to Page Authority, be sure to dig into our helpful resources:

Visit the PA Resource Center

Sign up for The Moz Top 10, a semimonthly mailer updating you on the top ten hottest pieces of SEO news, tips, and rad links uncovered by the Moz team. Think of it as your exclusive digest of stuff you don't have time to hunt down but want to read!

Title Tags SEO: When to Include Your Brand and/or Boilerplate

Posted by Cyrus-Shepard

If your websites are like most, they include a fair amount of extra "stuff" in the title tags: things like your brand name or repeating boilerplate text that appears across multiple pages.

Should you include these elements in your titles automatically?

To be fair, most sites do.

Alternatively, could it help your SEO to actually include less information in your titles? (Or at least in specific circumstances?)

We know from a handful of studies that titles of a certain length tend to perform better. A now-famous study from the engineers at Etsy showed how shorter titles performed better than longer ones. SEOs speculate that this could be because shorter titles can have more focused relevancy (by focusing on core keywords), might earn higher click-through rates, or some other reason we can't imagine.

When choosing which part of a title to shorten, brand names and boilerplate text are obvious choices. But how do you determine if this is something you should consider for your own SEO?

Here's an example of a brand's site name at the end of every title:

We've all seen sites like this. Heck, most of us do this on our own sites. The question is, does having our brand/site name at the end of every title actually help, or hurt?

But first, we also have to consider other types of boilerplate.

What is boilerplate? Boilerplate simply means standardized, non-unique pieces of text that are used over and over again. This often includes things like categories, product categories, author tags, and taglines.

In this example below, the boilerplate text on every title includes "Tomatoes - Vegetable Seeds - Shop."

Sometimes boilerplate material can become quite long. The comic book review site Major Spoilers (awesome name!) often includes the same 65-character boilerplate on many pages:

"Major Spoilers – Comic Book Reviews, News, Previews, and Podcasts"

Of course, at this length, it's so long that Google truncates every single title:

The problems that boilerplate can cause your SEO are threefold:

1. Relevancy: Unnecessary words can make your title less relevant, both to search engines and users. For search engines, this could mean lower rankings. For users, this could result in fewer clicks.
2. Uniqueness: Titles that share the same repeating text, and only vary from one another by a word or two, aren't very unique. While this isn't necessarily a problem, it goes against most SEO best practices, where uniqueness is key.
3. Length: Boilerplate means you have less room to display other words in your title, and Google will often cut these off if they go beyond a certain length.

Experiment #1: Remove category from title

We decided to run a couple of boilerplate experiments here at Moz, to see if we could increase our rankings and traffic by removing some of the repeating parts of our titles.

We started with our Whiteboard Friday blog posts. Every time Moz publishes a new Whiteboard Friday, we traditionally include "Whiteboard Friday" in the title.

What would happen if we removed this from the titles?

Using an A/B split test methodology — where we rolled the test out on 50% of the titles and used the other 50% as a control — we saw an amazing 20% uplift from this experiment.

This chart represents the cumulative impact of the test on organic traffic. The central blue line is the best estimate of how the variant pages, with the change applied, performed compared to how we would have expected without any changes applied. The blue shaded region represents our 95% confidence interval: there is a 95% probability that the actual outcome is somewhere in this region. If this region is wholly above or below the horizontal axis, that represents a statistically significant test.

Honestly, the results surprised us. Whiteboard Friday is a popular brand (so we thought) but removing this boilerplate from our titles produced a significant uplift in traffic to those pages.

At this point, we got cocky…

Experiment #2: Remove brand from title

If removing the category name from Whiteboard Friday posts produced such a significant uplift, what if we removed our brand name from all titles?

For this A/B experiment, we did exactly that—removing the word "Moz" from 50% of our titles and measuring the results.

Crazy, right? If it worked by removing "Whiteboard Friday" would we see the same uplift by removing "Moz?"

Sadly, Google had other plans:

While this A/B test never reached full statistical significance, we actually saw a 4% decline in traffic by removing our brand from our title tags.

Boo!

So why did this test not produce the same gains? To be honest, I've removed the brand name from other site's titles and seen as much as a 20% uplift.

It turns out that whether or not removing brand/boilerplate will be beneficial to your SEO depends on a few key factors, which you can gauge in advance.

How to know if removing boilerplate may succeed

Over 10 years of experience and literally millions of title tags, I've found that there are basically four factors that influence whether or not removing boilerplate from your titles might be beneficial:

1. Brand Strength: Popular brand names in titles almost always perform better than unknown brands, even when people aren't searching for your brand specifically. Amazon's brand recognition, for example, likely gives a significant boost to including "Amazon" in every title, even when people aren't specifically searching Amazon. Less recognizable brands, however, don't always get the same boost, and can actually lead to fewer visits based on relevancy, length, and clickability (described next.)
2. Relevancy: Are your boilerplate/brand keywords relevant to what your users search for? For example, if you're site is about television repair, then boilerplate titles that say "Brad's TV Repair" are going to be much more relevant than boilerplate that simply say "Brads." (We'll explore a way to determine your boilerplate's brand strength and relevancy in the next section.)
3. Length: In general, long boilerplate has the potential to do more harm than short boilerplate/brand words. Long boilerplate can dilute the relevance of your titles. So if you include "Buy Brad's TVs, Television Repair, High Definition Servicing, Audio and Visual Equipment for Sale in Houston Texas and Surrounding Areas" - you may want to rethink your boilerplate.
4. Clickability: Sometimes, boilerplate can make your titles more clickable, even if they aren't terribly relevant. Words like "Sale", "Solved", "Free", "2020", "New", and many others can lead to an increase in click-through rates (CTR.) Sometimes you can't tell until you test, but in many cases even adding clickable elements to your boilerplate can lead to significant gains.

Simple technique for determining your brand strength and boilerplate relevancy

This simple technique will also show why removing "Whiteboard Friday" led to an increase in traffic while removing "Moz" from titles did not.

Here's what you want to do: for each piece of boilerplate, determine the number of URLs on your site that rank/receive traffic for those keywords.

For this, we'll use Google Search Console.

Simply enter your boilerplate/brand as a query filter (you may need to break it into chunks for longer boilerplate) and see how many URLs receive traffic for queries that include that keyword.

When we filter for keywords that contain our "moz" brand name, we find thousands of ranking URLs.

People are searching for things like:

• Moz DA Checker
• Moz Pro
• Moz SEO
• Moz Blog
• Etc., etc.

As our brand name is part of so many queries and leads to visits across thousands of pages, this tells us that "Moz" is a very strong brand, and we'd likely be smart to include it as part of our title tags.

"Moz" is also very short at only 3 characters, which doesn't hurt either.

So what happens when we try this same technique with "Whiteboard Friday" — the boilerplate that led to a 20% uplift when we removed it? We see a very different result:

In this case, almost all the traffic for "Whiteboard Friday" search terms goes to only one or two pages.

For most Whiteboard Friday posts, the term is simply irrelevant. It's not what people are searching for, and the brand isn't strong enough to produce additional uplift.

Also, at 17 characters long, this boilerplate added significant length to each of our titles, in addition to possibly diluting the relevancy for what the posts were ranking for.

Final thoughts + bonus free title tag webinar

These tips can't tell you definitively whether you should or shouldn't include boilerplate or brand in your title tags, but they should give you a pretty good idea of when you should test things out.

Remember: Always test and evaluate before making any SEO change permanent. At least know the impact of the change you are making.

Also, please don't be under the impression that you should always remove boilerplate from your titles. In some instances, actually adding boilerplate can produce an uplift, particularly when the boilerplate is:

1. Recognizable: For example a strong brand
2. Relevant: The right keywords
3. Clickable: Encourages a high CTR
4. Succinct: Not overly long

---

If you found value in the tips, and want to learn even more ways to optimize your title tags, we've made available a free webinar for you: SEO Master Class: Advanced Title Tag Optimization (For Any Site).

If you've got 40 minutes, it's definitely worth a watch.

Watch Free Webinar

Best of luck with your SEO!

Sign up for The Moz Top 10, a semimonthly mailer updating you on the top ten hottest pieces of SEO news, tips, and rad links uncovered by the Moz team. Think of it as your exclusive digest of stuff you don't have time to hunt down but want to read!

Page Authority 2.0 Is Coming This Month: What’s Changing and Why

Posted by rjonesx.

Hey folks,

I'm Russ Jones, Adjunct Search Scientist with Moz, and I'm proud to announce that this month we’ll be releasing a terrific update to our metric, Page Authority (PA).

Although Page Authority hasn't attracted the same attention as its sibling metric Domain Authority, PA has always correlated with SERPs much better than DA, serving as a strong predictor of ranking. While PA has always fluctuated with changes in the link graph, we’re introducing a whole new method of deriving the score.

Learn More About Page Authority 2.0

What's changing

Long gone are the days of just counting backlinks a couple of ways and hoping they correlate well with SERPs. As Moz tends to do, we’re pioneering a new manner of calculating Page Authority to produce superior results. Here are some of the ways we’re changing things up:

The training set

In the past, we used SERPs alone to train the Page Authority model. While this method was simple and direct, it left much to be desired. Our first step in addressing the new Page Authority is redefining the training set altogether.

Instead of modeling Page Authority based on one page's ability to outrank another page, we now train based on the cumulative value of a page based on a number of metrics including search traffic and CPC. While this is a bit of an oversimplification of what’s going on, this methodology allows us to better compare pages that don't appear in the SERPs together.

For example, imagine Page A is on one topic and Page B is on another topic. Historically, our model wouldn't get to compare these two pages because they never appear on the same SERP. This new methodology provides an abstract value to each page, such that they can be compared with any other page by the machine-learned model.

The re-training set

One of the biggest problems in building metrics is not what the models see, but what the models don't see.

Think about this for a minute: what types of URLs don't show up in the SERPs that the model will use to produce Page Authority? Well, for starters, there won't be many images or other binary files. There also won't be penalized pages. In order to address this problem, we now use a common solution of running the model, identifying outliers (high PA URLs which do not in fact have any search value), and then feeding those URLs back into the training set. We can then re-run the model such that it learns from its own mistakes. This can be repeated as many times as is necessary to reduce the number of outliers.

Ripping off the Band-Aid

Moz is always cognizant of the impact the changes to our metrics might have on our customers. There is a trade-off between continuity and accuracy. With Page Authority, we’re focusing on accuracy. This may cause larger-than-normal shifts in your Page Authority, so it’s more important than ever to think about Page Authority with respect to your competitors, not as a standalone number.

What actions should we take?

Communicate with stakeholders, team members, and clients about the update

Just like our upgrade to Domain Authority, some users will likely be surprised by changes in their PA. Make sure they understand that the new PA will be more accurate (and more useful!) and that the most important measurement is relative to their competitors. We won't release a Page Authority which isn't better than the previous version, so even if the results are disappointing, understand that you now have better insight than ever before into the performance of your pages in the SERPs.

Use PA as a relative metric, like DA

Page Authority is intrinsically comparative. A PA of 70 means nothing unless you know the PA of your competitors. It could be high enough to allow you to rank for every keyword you like, or it could be terribly low because your competitors are Wikipedia and Quora. The first thing you should do when analyzing the Page Authority of any URL is set it in the proper context of its competitor's URLs.

Expect PA to keep pace with Google

Just as we announced with Domain Authority, we’re not going to launch the new PA and just let it go. Our intent is to continue to improve upon the model as we discover new and better features and models. This volatility will mostly affect pages with unnatural link profiles, but we would rather stay up-to-date with Google's algorithms even if it means a bit of a bumpy ride.

When is it launching?

We’ll be rolling out the new Page Authority on September 30, 2020. Between now and then, we encourage you to explore our resources to help you prepare and facilitate conversations with clients and team members. Following the launch of the new PA, I’ll also be hosting a webinar on October 15 to discuss how to leverage the metric. We’re so excited about the new and improved PA and hope you’re looking forward to this update too.

If you have any questions, please comment below, reach out to me on Twitter @rjonesx, or email me at russ@moz.com.

To get prepared and learn more about the upcoming change to Page Authority, be sure to dig into our helpful resources:

Visit the PA Resource Center

Sign up for The Moz Top 10, a semimonthly mailer updating you on the top ten hottest pieces of SEO news, tips, and rad links uncovered by the Moz team. Think of it as your exclusive digest of stuff you don't have time to hunt down but want to read!

Title Tags SEO: When to Include Your Brand and/or Boilerplate

Posted by Cyrus-Shepard

If your websites are like most, they include a fair amount of extra "stuff" in the title tags: things like your brand name or repeating boilerplate text that appears across multiple pages.

Should you include these elements in your titles automatically?

To be fair, most sites do.

Alternatively, could it help your SEO to actually include less information in your titles? (Or at least in specific circumstances?)

We know from a handful of studies that titles of a certain length tend to perform better. A now-famous study from the engineers at Etsy showed how shorter titles performed better than longer ones. SEOs speculate that this could be because shorter titles can have more focused relevancy (by focusing on core keywords), might earn higher click-through rates, or some other reason we can't imagine.

When choosing which part of a title to shorten, brand names and boilerplate text are obvious choices. But how do you determine if this is something you should consider for your own SEO?

Here's an example of a brand's site name at the end of every title:

We've all seen sites like this. Heck, most of us do this on our own sites. The question is, does having our brand/site name at the end of every title actually help, or hurt?

But first, we also have to consider other types of boilerplate.

What is boilerplate? Boilerplate simply means standardized, non-unique pieces of text that are used over and over again. This often includes things like categories, product categories, author tags, and taglines.

In this example below, the boilerplate text on every title includes "Tomatoes - Vegetable Seeds - Shop."

Sometimes boilerplate material can become quite long. The comic book review site Major Spoilers (awesome name!) often includes the same 65-character boilerplate on many pages:

"Major Spoilers – Comic Book Reviews, News, Previews, and Podcasts"

Of course, at this length, it's so long that Google truncates every single title:

The problems that boilerplate can cause your SEO are threefold:

1. Relevancy: Unnecessary words can make your title less relevant, both to search engines and users. For search engines, this could mean lower rankings. For users, this could result in fewer clicks.
2. Uniqueness: Titles that share the same repeating text, and only vary from one another by a word or two, aren't very unique. While this isn't necessarily a problem, it goes against most SEO best practices, where uniqueness is key.
3. Length: Boilerplate means you have less room to display other words in your title, and Google will often cut these off if they go beyond a certain length.

Experiment #1: Remove category from title

We decided to run a couple of boilerplate experiments here at Moz, to see if we could increase our rankings and traffic by removing some of the repeating parts of our titles.

We started with our Whiteboard Friday blog posts. Every time Moz publishes a new Whiteboard Friday, we traditionally include "Whiteboard Friday" in the title.

What would happen if we removed this from the titles?

Using an A/B split test methodology — where we rolled the test out on 50% of the titles and used the other 50% as a control — we saw an amazing 20% uplift from this experiment.

This chart represents the cumulative impact of the test on organic traffic. The central blue line is the best estimate of how the variant pages, with the change applied, performed compared to how we would have expected without any changes applied. The blue shaded region represents our 95% confidence interval: there is a 95% probability that the actual outcome is somewhere in this region. If this region is wholly above or below the horizontal axis, that represents a statistically significant test.

Honestly, the results surprised us. Whiteboard Friday is a popular brand (so we thought) but removing this boilerplate from our titles produced a significant uplift in traffic to those pages.

At this point, we got cocky…

Experiment #2: Remove brand from title

If removing the category name from Whiteboard Friday posts produced such a significant uplift, what if we removed our brand name from all titles?

For this A/B experiment, we did exactly that—removing the word "Moz" from 50% of our titles and measuring the results.

Crazy, right? If it worked by removing "Whiteboard Friday" would we see the same uplift by removing "Moz?"

Sadly, Google had other plans:

While this A/B test never reached full statistical significance, we actually saw a 4% decline in traffic by removing our brand from our title tags.

Boo!

So why did this test not produce the same gains? To be honest, I've removed the brand name from other site's titles and seen as much as a 20% uplift.

It turns out that whether or not removing brand/boilerplate will be beneficial to your SEO depends on a few key factors, which you can gauge in advance.

How to know if removing boilerplate may succeed

Over 10 years of experience and literally millions of title tags, I've found that there are basically four factors that influence whether or not removing boilerplate from your titles might be beneficial:

1. Brand Strength: Popular brand names in titles almost always perform better than unknown brands, even when people aren't searching for your brand specifically. Amazon's brand recognition, for example, likely gives a significant boost to including "Amazon" in every title, even when people aren't specifically searching Amazon. Less recognizable brands, however, don't always get the same boost, and can actually lead to fewer visits based on relevancy, length, and clickability (described next.)
2. Relevancy: Are your boilerplate/brand keywords relevant to what your users search for? For example, if you're site is about television repair, then boilerplate titles that say "Brad's TV Repair" are going to be much more relevant than boilerplate that simply say "Brads." (We'll explore a way to determine your boilerplate's brand strength and relevancy in the next section.)
3. Length: In general, long boilerplate has the potential to do more harm than short boilerplate/brand words. Long boilerplate can dilute the relevance of your titles. So if you include "Buy Brad's TVs, Television Repair, High Definition Servicing, Audio and Visual Equipment for Sale in Houston Texas and Surrounding Areas" - you may want to rethink your boilerplate.
4. Clickability: Sometimes, boilerplate can make your titles more clickable, even if they aren't terribly relevant. Words like "Sale", "Solved", "Free", "2020", "New", and many others can lead to an increase in click-through rates (CTR.) Sometimes you can't tell until you test, but in many cases even adding clickable elements to your boilerplate can lead to significant gains.

Simple technique for determining your brand strength and boilerplate relevancy

This simple technique will also show why removing "Whiteboard Friday" led to an increase in traffic while removing "Moz" from titles did not.

Here's what you want to do: for each piece of boilerplate, determine the number of URLs on your site that rank/receive traffic for those keywords.

For this, we'll use Google Search Console.

Simply enter your boilerplate/brand as a query filter (you may need to break it into chunks for longer boilerplate) and see how many URLs receive traffic for queries that include that keyword.

When we filter for keywords that contain our "moz" brand name, we find thousands of ranking URLs.

People are searching for things like:

• Moz DA Checker
• Moz Pro
• Moz SEO
• Moz Blog
• Etc., etc.

As our brand name is part of so many queries and leads to visits across thousands of pages, this tells us that "Moz" is a very strong brand, and we'd likely be smart to include it as part of our title tags.

"Moz" is also very short at only 3 characters, which doesn't hurt either.

So what happens when we try this same technique with "Whiteboard Friday" — the boilerplate that led to a 20% uplift when we removed it? We see a very different result:

In this case, almost all the traffic for "Whiteboard Friday" search terms goes to only one or two pages.

For most Whiteboard Friday posts, the term is simply irrelevant. It's not what people are searching for, and the brand isn't strong enough to produce additional uplift.

Also, at 17 characters long, this boilerplate added significant length to each of our titles, in addition to possibly diluting the relevancy for what the posts were ranking for.

Final thoughts + bonus free title tag webinar

These tips can't tell you definitively whether you should or shouldn't include boilerplate or brand in your title tags, but they should give you a pretty good idea of when you should test things out.

Remember: Always test and evaluate before making any SEO change permanent. At least know the impact of the change you are making.

Also, please don't be under the impression that you should always remove boilerplate from your titles. In some instances, actually adding boilerplate can produce an uplift, particularly when the boilerplate is:

1. Recognizable: For example a strong brand
2. Relevant: The right keywords
3. Clickable: Encourages a high CTR
4. Succinct: Not overly long

---

If you found value in the tips, and want to learn even more ways to optimize your title tags, we've made available a free webinar for you: SEO Master Class: Advanced Title Tag Optimization (For Any Site).

If you've got 40 minutes, it's definitely worth a watch.

Watch Free Webinar

Best of luck with your SEO!

Sign up for The Moz Top 10, a semimonthly mailer updating you on the top ten hottest pieces of SEO news, tips, and rad links uncovered by the Moz team. Think of it as your exclusive digest of stuff you don't have time to hunt down but want to read!