New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Is page speed insights a good assessment of a hosting provider?
Neuromante
Member
in Help
If I test hosting pages there, is a goog indicator of the speed in his customers?
I guess it is not, but are any other free and reliable ways to know?
My interest in this is for SEO purposes.
Thanked by 1supakika
Comments
Good o' SEO and page speed, a relic from 2000.
If a providers site is fast for YOU as a user, then do your regular checks on them.
Pagespeed insights doesn't always reflect 100% what the user experience is like. I can view a mobile site which is pretty snappy, but only scores 40 on Pagespeed.
Also slow pages is usually caused by your software and not by the host.
I'v saw pages wich position first with web.dev and page speed insights very low, and very different too. Even a page of some SEO guru.
Can you explain more about this? It is not speed an important factor?
If your target audience is bots, yes, it matters.
It's a marketing gimmick to pray on fools.
Update: Read my (hopefully useful) comment below jar's post for answers.
We do not publish at first for google?
This. Almost any oversold host can serve a fast website in the right conditions, and almost no host can serve a poorly built website quickly.
always thought that content was published for Humans, not google. Gosh I could be wrong.
In any case, I am not a SEO expert by any means, but here's a fun experiment:
Look up Top websites on Alexa. Or 50, or 500 if you set up a script or a bot.
a. Look up the page speed insights of That page itself. (Page speed ranking =1)
c. Now, see the page speeds for mobile for some top sites.
YouTube(pagespeed raking 41)
Facebook (pagespeed ranking = 80)
tmall (pagespeed raking = 42)
Reddit(pagespeed ranking = 35)
Then think hard for a moment: is page speed really that important, or quality of content, depth of content, diversity and relevance, seo optimized pages, repeat visitors, stickyness, etc.. etc.. also matter? I am not saying page speed is not important, but is that the only thing you should be worried about?
Update: LET scores an awesome 18 on google pagespeed. No wonder users do not like to keep coming back here. (Or was it the 522 errors?)
You would be fine as long as the site is actually fast for the real users. You can get insights from page speed but don't get cover concerned with it, if you do you would be simply wasting your time. If you are using wordpress having light speed cache would make a big difference for the end user.
if your hosting provider is hetzner, then definite no.
As far as I understand the first reader is some google bot. If it likes mysterious things in the page, it ranks.
If page do not ranks, humans can't read.
While I think the same like you, I see a lot of like SEO optimized sites.
PLOT TWIST: provider puts their web site on a CDN.
What if... the provider is the CDN provider
@Neuromante
Where are the customers in relation to the server? All of them. Or 90%. Or maybe 50%? Is that even known in a given case, e.g. yours?
Also consider that processing time (the time needed by the server to process a request and to find or build an answer to be sent) is insignificant compared to round trip time. Even in relatively pampered Europe with a low round trip time the processing time is insignificant, say 2 ms vs 60 ms.
Plus, bots work quite differently from browsers. For a human user TTFP (time to first paint) or TTPFP (time to page fully painted) are relevant. For a bot time to first response packet and time to page fully loaded or time to response fully processed are relevant. Those may sound similar but actually almost completely different things, with the users part (painting) usually being the slow one.
So, a not at all uncommon case is the server needing 2 ms, TCP round trip is 60 ms but user waits > 2000 ms for the page to be fully loaded and painted. And THAT is what the user cares about.
TL;DR Dont put too much effort into finding a fast server on a high end network but rather
[1] I've seen many cases where pages themselves would have loaded quickly but the 3rd. party ads made them slow.
Are you sure your advice results in good google ranking?
TTFL (Time to First Lulz) for this thread was ... 41 hours 47 minutes
@Neuromante - if I understand the subtext of @jsg
's analysis - in the long run, having actual humans reading, appreciating, remembering, and possibly even eventually linking to your site will be your best bet for ranking in search engine results.whatever else you do to boost your site's "bot appeal" may or may not have some minimal effect, but it really won't matter unless you happen to be some mystic master of supposed SEO mojo - in which case you wouldn't be asking here, at least not with any expectation of receiving a straight answer.
tl;dr: if you don't have good content presented to build a real human audience, don't bother rearranging deck chairs on a sinking ship
@jsg suggested sticking to the fundamentals- keeping the human audience in mind, also reiterated by @uptime. Unfortunately that is the slow and difficult path. But time tested in the short 30 odd year history of internet or even shorter for blogging.
Again this is not the right place for insights for SEO, but the collective advice from the posts should give any newbie a headstart.
For search terms that don't have a very high competition, page content quality matters by far the most. For search terms that have a very high competition, domain authority, along with page content, plays a crucial role (though even with high domain authority, good quality content is still crucial for ranking high).
Speed is among the ranking factors, yes. If, for example, top ten search results are closely matched in search engine ranking for a certain term, then small differences (speed among them) can make a difference - adding a fraction of a "ranking point", enough to make a difference. However, I'd say that content and domain authority are a lot more important (by a huge margin).
As long as the page isn't too large in megabytes and doesn't load too slowly (like 5+ seconds), it should be fine. Exception are websites with sales, there, from what I could read, short attention span and little patience of an average visitor has an impact and there load speed makes a difference (so they say) - haven't tested this.
Page speed info can provide good info on how to improve the load speed, it shouldn't be disregarded. But it's not the most important thing. In fact, I personally find GTmetrix info a lot more useful.
Wrote in more details on SEO here:
https://io.bikegremlin.com/5944/optimization-06-seo/
Speed is mentioned, near the end, in the matter/context that it matters/deserves.
@jsg Google ads do slow page load times - sometimes a lot. The good thing is they can be set to load after all the other content has been loaded (my ongoing experiment with Google AdSense ad placement).
When doing speed tests, it's obvious how much of a difference (in terms of lower speed and optimization) the added adverts make.
If you focus on your human audience you can win because you are a human too and because you highly likely share some common ground with them, plus at the end of the day the question of real relevance for human beings is the holy grail in search engine ranking too.
If you focus on SEO, i.e. on tricking bots and AI, you are bound to loose. Google and similar simply have way, way more resources and engineers than you and to make it worse a considerable part of those resources is spent on fighting "SEO optimization". Additionally you'll p_ss off many humans in your target audience.
Oh and: Yes I am sure about my advice. Because it works. I don't care AT ALL about SEO but only about my audience and my site has been continually growing and it's quite high up in search results too, often in fact on page one.
No
@Neuromante If you have a visitors from Russia (lets say) and your hosting servers are in USA you will have latency. You can not use test speed from your locations to determinate how fast the site will load from the client side. If you go for SEO you may take some points from unic content, visitors, the text on the page should not be less than 400-600 words. optimize your code and use the old well known alt in the img tag, because this is used from machines to read the content to people who are not able to. A very good strategy is to have the text on your pages in many languages. That will sure bring you visitors.However the optimization is a little more specific. You can not apply the same optimization exactly for all sites.
And to not forget the best practices is to make the visitors happy if they are happy they will return if they return this means your go is a little bit closer
@jsg is right that you must care first for the visitors and everything else is a step aside
If you live in China, all providers from here are going to have 300+ ms rtt latency, for you.
Which part of China could well add or reduce extra 100ms.
First of all: what is the relationship between "provider" and pagespeed results?
Pagespeed Insight is focused on code/cms/development feature.
Provider, and since we are on LEB, we think about unmanaged root hosting , do not have anything to do with pagespeed results.
And by "anything" i mean they just need to have a good network..
But for elements like:
-ttfb
-cache
-css/js bundling and prioritization
-etc
Provider has nothing to do. I have website hosted on a fairy unknow provider but the development agency build the website so good they have achived 100 on pagespeed...