Google Rolls Out Search Console Speed Report

[engine]

Google is rolling out its search console speed report.

the Speed report automatically assigns groups of similar URLs into "Fast", "Moderate", and "Slow" buckets. The data is based on the Chrome User Experience Report,

[webmasters.googleblog.com...]

[Shepherd]

Was looking at this earlier, not really confident in data they are showing. Doesn't help that the dev tools, lighthouse is not working in chrome today...

[NickMNS]

Interesting. My report doesn't look too good. The great thing is that it shows the results across many pages. With millions of pages testing a few pages is hardly representative, but with this new report it provides a representative sample that will allow me to target some fixes. More work...

[SweetPotato]

Hard to have green pages when AdSense takes 3 seconds to load. Is Google suggesting to ditch AdSense?

[JorgeV]

Hello-

Hard to have green pages when AdSense takes 3 seconds to load. Is Google suggesting to ditch AdSense?

The GSC takes in consideration only the FCP and FID, so Adsense (or other third part async scripts) has no impact on these measures.

[aristotle]

This thing doesn't appear to be working properly. It could be showing incorrect information. Might be best to wait a while before reaching any conclusions about your site's performance.

[JorgeV]

Hello -again

Also, keep in mind that these statistics are based on data collected from Google Chrome users, worldwide, and who accepted that Chrome sends "anonymous" data. (I don't know if this is the default for EU users).

It means real users, with all kind of connection quality and speed.

When you test your pages, with online tools, they have excellent connectivity, which is resulting in better results than average users. When you choose to simulate mobile connection, it's just throttling the speed, or just doing a calculation, but it's not taking in consideration network quality, lost packets and so on, which can happen more frequently on mobile devices.

Same remark when you test your site using the audit tool of Chrome/Chromium. It's still ideal conditions, even with the simulation turned on.

[elos42]

What this is showing is downright alarming.
Suddenly, on Oct 30, it shows most of my pages falling into slow category on mobile. What could it mean?
I had put the site behind cloudflare a week before that, and not on Oct 30.
I wonder if it had anything to do with it, or whether it's sort of like the day Google started collating data or something.
You can see my report here:
[drive.google.com ]

[riccarbi]

What this is showing is downright alarming.
Suddenly, on Oct 30, it shows most of my pages falling into slow category on mobile. What could it mean?

My suggestion? Don't give a damn about it.
The importance of speed for ranking is another Google b..it; like many other similar b..its of the past - such as responsive design, https, better ads standards, and so on - only some self-appointed "SEO expert" still care about Google's "suggestions". I've never seen any ranking change consequent to my websites' speed increase or decrease for the last 10 years.
This move is just another fig-leaf aimed to cover Google's secret eleventh commandment: "Do you want to rank better? Want more traffic? Give us some money!" and/or to force webmasters to use AMP.

PS: I have zero slow URLs as per the GSC "speedometer", so my complaints are not sour grapes...

[Selen]

Is it safe to assume that all AMP pages will be marked as "Fast" by default, even if they are slow?

[RedBar]

Do / dare / can, I assume that if I am "100" that everything is ok according to G?

[dethfire]

Yikes, most of my pages are labeled as slow in mobile. I check them in gtmatrix and pingdom tools and they check out as fast and loading in a few seconds. I don't know what to do really to make them fast to Google.

I also see the Adsense is a major speed issue. I even removed Analytics, but Adsense is half my bottleneck.

[dethfire]

Also I agree not to read too much into it. Put big sites into gtmatrix like CNN.com and see what disasters they are and yet no penalties for them.

[RedBar]

I also see the Adsense is a major speed issue.

I used to run AdSense, removing that improved things dramatically then removing lightbox turned me into 100% on all sites.

I'm a code, how shall I descibe myself, perfectionist hacker, have been since whenever, I was really surprised about lightbox however my new rendering of images is actually much better, this test likes it and the SEs love it even more:-)

[RedBar]

Put big sites into gtmatrix

However, like many big sites, they are so code-bloaty I wonder if anyone, anywhere actually knows what their sites are doing?

[aristotle]

Well this thing clearly isn't working properly. I just checked one of my sites again, which has 32 pages for which the test shows results.

For desktop it shows all 32 pages jumping from day to day back and forth between fast and moderate. Makes no sense that all 32 pages would be fsat one day, then moderate the next day, then fast the next day, and so on. (On mobile, all are rated moderate for all days)

Actually all of these pages are extremely fast, loading in a fraction of a second even at slow connection speeds.

There was a previous google speed test that gave numerical scores from 0 to 100. If I remember correctly, all the pages on any of my sites that I ever checked scored between 98 and 100. I don't know if that test is still available, but couldn't find a link for it.

Can anyone explain why this new speed checker doesn't give numerical scores like the old one did. "Moderate" could mean that a page is just barely faster than "slow", or just just barely slower than "fast". There's no way to know. A numerical score would be far more useful.

[elos42]

I think I may be paying a penalty for the fact that 90% of my customers come over mobile connections in India, which can be quite unpredictable. So, if Google's going by load times on those mobiles, and compare those times to a 'global benchmark', the load-times for my site would be very high.
On the other hand, if load times are benchmarked against 'national averages', I wouldn't do so bad. Similarly, if mobile download times are compared with mobile download times, instead of a blended average, I wouldn't do so bad either. I don't know how they do this.
As far as I can see, sites whose consumers are located in high-speed countries such as Korea, Japan, USA and EU should automatically be marked fast, compared to an identical site on the same host targeting a country like India or Pakistan, which would be marked slow.

[aristotle]

automatically be marked fast

I'm not sure what you mean by "automatically".

Aren't these results supposed to be based on "real-world" measurements made by the chrome browser? I rarely use chrome but expect that all the major browsers perform at about the same level, so this shouldn't be a factor.

In any case, a terribly-slow site will be terribly slow for users in all countries. It wouldn't be "automatically be marked fast".

[riccarbi]

Let's face it. Fast or slow means nothing.
On desktops, I expect almost all modern websites to be fast enough to provide reasonably good user experience, these days. If a website is slow on desktops, that means it either uses an obsolete technology or it's badly coded; these reasons should be enough to cause Google to penalize it in SERP, independently from speed.
On mobile devices, page loading time depends much more on local mobile network speed, signal strength, device power, and so on than it depends on a website's theoretical performances; so it's a largely meaningless indicator.

One may say that Google's speed report is a comparative tool, measuring how a website performs against all other websites. Again, what sense does it make to compare totally different websites? And who cares about a page loading in 1 second versus a page that loads in 3 seconds, as long as they are both able to provide a good user experience?

The only webmasters (and search engine owners) obsessed with page speed I know are those whose websites lack good content.

[tangor]

The tool is only as good as the service that measures it ... has anyone considered that g, itself, is messing with the results SIMPLY because it is overburdened, over-extended, or overly SLOW itself?

Every observer INTRODUCES a change in the results SIMPLY because they are observing. In this case it gets worse because g is observing g accessing a third party, complicating things.

If, on the other hand, a third party was observing g doing "ordinary" (ie, accessing pages) that would produce a different metric, and possibly more believable. Most of these g "tools" are placebos to get webmasters to do different things ... things that benefit g first...

Sounds tinfoil ... but not really. :)

.

[Shepherd]

or overly SLOW itself?

interesting observation.

My desktop traffic reports as "fast"

My mobile traffic reports as "moderate"

google is getting this information from Chrome

Maybe Chrome for mobile is a bad product. (I've never used it, don't know)

[tangor]

The measurement itself will introduce errors ... and that's the takeaway.

And one reason to ignore this new "tool".

YMMV

.

[koan]

Your site could be lightweight and lightning fast in terms of local code but if your traffic comes from places with unreliable network performance (rural areas, poorer countries), then it will categorized it as slow. I feel the results should provide that kind of information.

[Roie]

We should all go back to 90s early 2000s tech. My next website will be built using Macromedia Dreamweaver and Flash, it will be fast!

[RedBar]

And one reason to ignore this new "tool".

Right, I am going to be totally dumb ... again!

Where is this damn-fangled new tool?

I thought you were talking about the "old" one where I score 100, obviously not, where is this effing thing?

If it's a logged-in G thing then I won't be going there.

[elos42]

Your site could be lightweight and lightning fast in terms of local code but if your traffic comes from places with unreliable network performance (rural areas, poorer countries), then it will categorized it as slow.

That's what I said too. However, it's possible for Google to avoid the 'poor reader effect' by comparing apples to apples. A site should be compared to a bunch of peers who share the same or very similar audience (mobile, third world etc)
And yes, we don't know if the current 'fast', 'slow' desig is based on such a comparison or not.

[elos42]

I've put an update on this issue here -- [webmasterworld.com...] Basically trying to see if anyone has anything to contribute towards the definitions that GSC seems to be using as far as FID, FCP etc are concerned. However, let's continue the general discussion on the new tools here.

[No5needinput]

@RedBar [developers.google.com ] However, the tool has changed from what it was years ago, now it appears to be closely related to LightHouse data.

Edited: Oops, my bad, the one they are talking about is in GSC

[RedBar]

the one they are talking about is in GSC

Thanks, that's what I thought:-)

[engine]

More dots to connect...
Google Chrome to Badge Site Speeds [webmasterworld.com]