Google‘s Recent Page Speed Insights Interface Overhaul
Users of Google’s PageSpeed Insights didn’t always understand the context of the data they were presented with. If you’re new to search engine optimization (SEO), you could be one of these users. Fortunately for us, Google has taken the time to address people’s confusion and rolled out changes to the tool to make data easier to parse and understand.
At that point, URL-level field data was shown for the Core Web Vitals if it was available. If not, lab data was used instead. The item at the head of the tool’s results was the PageSpeed rating (imaged above as “43”). The PageSpeed rating was still pushed as the most important metric visually, with the Core Web Vitals serving merely as feed-ins.
Core Web Vitals data was relatively top-line, and the main focus was still on the PageSpeed rating score. As per the new layout, there seems to have been a fundamental shift in focus:
With the new layout, there’s a much stronger focus on the Core Web Vitals metrics, which have been pushed higher and given an overall synopsis; in this case: “Core Web Vitals Assessment: Failed.”
Core Web Vitals Expanded
Not only are the Core Web Vitals more prominent, but they can also be expanded:
The expansion boxes give granular detail as per Google’s attempted page loads. When checking a page in Google’s PageSpeed Insights, the check takes significantly longer to perform than when you load the checked page in a browser tab. That’s because Google attempts to load the same page numerous times when generating PageSpeed results.
For the page in our example, we can see that the Core Web Vital called Largest Contentful Paint (LCP) is slow, with 51 percent of page loads in the good range, 31 percent in the “needs improvement” range and 18 percent in the poor range. This tells us LCP is only successfully met for this page around half the time (and thus the page fails the test).
Lab Data Prominence and Origin Data Toggle
At the top of the interface, lab data seems to have been removed. If a page has insufficient “real-world” data, no lab data is offered as a substitute at the top of the read-out, unlike the old version of the tool:
However, if you scroll down the tool’s read-out, estimated lab data (and the associated PageSpeed rating) are still available:
Near the top of the page, there’s a new “Origin” toggle:
This allows you to move away from data bound to the checked URL, and instead to view “origin” data. Origin-level data is an average of all data collected for all pages on the specified website – not only the specific web page being checked but also the whole website!
Toggling between “This URL” and “Origin” can quickly help a webmaster ascertain whether an individual URL is performing above, in line with or below other pages on the site (from a Core Web Vitals perspective).
Sometimes, if there is insufficient URL data, Google’s PageSpeed Insights will default to origin-level data instead. As previously mentioned, if there is insufficient real-world URL-level and origin-level data, Google will show the “No data found” error. In such a situation, simply scroll down to where the tool says “Values are estimated and may vary.” Below that point, you have a small “METRICS” heading, which is Google’s emulated lab data.
You can still use Google’s estimated data if real-world field data is unavailable. This is probable for sites with low traffic, which are likely to see the “No data found” error nearer the top of the tool’s output. Google’s real-world field data mostly comes from the CrUX report. Google states that this data comes from Chrome users who “have opted-in to syncing their browsing history, have not set up a Sync passphrase and have usage statistic reporting enabled.” This is clearly not everyone, so traffic has to reach a certain peak before field data is available.
So, What‘s Changed?
The design of the page has been altered, and the tool looks much slicker. But in terms of things that SEO specialists need to care about, here are changes that we’ve observed so far:
- The PageSpeed rating (actual numeric read-out) has been pushed further down into the estimated metrics section. This shows less focus on the overall page speed rating, though it’s important to remember that this still does matter to an extent.
- This video by Google’s Martin Splitt explains some PageSpeed-related SEO myths. He states that the PageSpeed rating is still important, but not in an extremely granular way. For example, if your rating changes from yellow or orange (average) to green (good), then a page is likely to see very slight boosts in Google’s rankings. Content still matters more, but if two pages have equally good content and one is in a faster page speed, that page is likely to rank above the other. If your PageSpeed rating changes from 90 to 95 (good), then don’t expect to see any ranking differences at all.
- Core Web Vitals are paramount for PageSpeed optimization. If real-world field data says that the Core Web Vitals assessment has “failed,” that’s going to be your main problem and your main issue to address.
- Field data is now much more prominent and, therefore, likely much more important than lab data. If the lab data says your page is okay and you have an average page speed, but the field data says things aren’t good and that your “Core Web Vitals Assessment” has “failed,” that needs addressing.
- Field data now have accessible origin-level data, which may be defaulted to if there is not enough URL-level field data. If there is enough URL-level field data, you can still navigate to the origin-level field data. This gives performance analysts the power to compare an individual URL’s performance (for Core Web Vitals) against the site’s average norm.
- Core Web Vitals metrics can be expanded to show performance during different page loads, which were initiated by Google’s PageSpeed Insights tool.
There have been lots of interesting changes from Google’s side. The re-arranging of the user experience (UX) to put Core Web Vitals first makes perfect sense after Google’s recent algorithm updates. It’s pretty great to be able to expand individual Core Web Vitals metrics to see data by page load. The newly accessible origin-level data for field data, specifically, is a nice touch. Previously, there was an “origin summary,” but it wasn’t so useful.
So there you have it! As Google overhauls its tools and the various associated interfaces, we continue to be vigilant and interpret these changes.