WordPress is publishing a maintenance release that fixes issues that broke millions of websites.
The post WordPress 5.5.1 Fixes Millions of Broken Sites via @martinibuster appeared first on Search Engine Journal.
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.
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:
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.
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.
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.
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.
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.
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.
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:
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!
Here's how agencies can take a thorough look at SEO forecasting and better communicate to clients how an SEO goal can impact their business.
The post Why Do SEO Forecasting Models Keep Failing Us? via @seomonitor appeared first on Search Engine Journal.
Pinterest is currently testing a number of new shopping features designed to draw more attention to products pins.
The post Pinterest Tests Product Ratings, “Best Seller” Labels, More via @MattGSouthern appeared first on Search Engine Journal.
As 58% of voice search users employ it to run local business queries, a new study looks at factors that go into ranking the results.
The post 2020 Local Voice Search Ranking Factors [STUDY] via @MattGSouthern appeared first on Search Engine Journal.
Here is a recap of what happened in the search forums today...
Google is now displaying a “Licensable” badge in image search results on photos that can be purchased for legal use on other websites.
The post Google Shows “Licensable” Badges in Image Search Results via @MattGSouthern appeared first on Search Engine Journal.
Google has removed both the Labeled for Reuse and Labeled for Reuse with Modification' features from its image search tools.
The post Google Removes ‘Labeled for Reuse’ Options from Image Search Tools via @karenneicy appeared first on Search Engine Journal.
Google announced a new Home Activities Rich Results. Currently triggered by fitness related queries on mobile.
The post Google Announces Home Activities Rich Results via @martinibuster appeared first on Search Engine Journal.
Google introduces a personalized entity repository, similar to a user-specific knowedge graph to predict actions and suggestions for users of Google.
The post A Personalized Entity Repository in the Knowledge Graph via @bill_slawski appeared first on Search Engine Journal.
Need to create paid search offers that don't just convert, but sell? This step-by-step process gives you the structure to maximize profits.
The post How to Create Paid Search Offers That Are Actually Profitable via @amyppc appeared first on Search Engine Journal.
Many people are suspicious of information they find online, but a very basic understanding of how algorithms and machine learning work can help.
The post How Do Algorithms Work? A Basic Primer for Non-Marketers via @CarolynLyden appeared first on Search Engine Journal.
Join this webinar and learn how to identify internal link optimization opportunities that can elevate your organic presence to the next level.
The post Internal Linking: How to Unlock Your Site’s Hidden Potential [Webinar] via @brentcsutoras appeared first on Search Engine Journal.
