Site Speed: What SEOs Need to Know
[MUSIC PLAYING] JOHN MUELLER: Welcome,
everyone, to a special episode of Ask Google Webmasters. Here today is Martin. MARTIN SPLITT: Hi, everyone. And this is John. JOHN MUELLER: Hi. Yeah, we’re going to
be answering questions on the topic of speed
that were submitted with the #AskGoogleWebmasters
on Twitter. So let’s start with
a question from Rohit. What is the ideal page
speed of any content for better ranking on Search? MARTIN SPLITT: Oh, you’re
asking me that question. OK. So basically, we are
categorizing pages, more or less, as really
good and pretty bad. So there’s not
a threshold in between, it’s just we are, more or
less, roughly categorizing the speed experience for users.
And how are we
doing that? Where do we get the data from? JOHN MUELLER: Yeah. That’s important. So we mostly get
data from two places. On the one hand, we try to
calculate the theoretical speed of a page using
lab data, and then we also use real field data
from users who have tried to use those pages. And that field data is similar
to the Chrome User Experience Report data. MARTIN SPLITT: Cool. So we have hypothetical
data and practical data. So we don’t have
a threshold to give away, but basically, the
recommendation I would say is just to make sites
fast for users. That’s what it boils down to. JOHN MUELLER: That sounds good. Yeah. MARTIN SPLITT: The next
question comes from Olga. Olga is wondering if a
website’s mobile speed is best looked at using the Test
My Site tool, the GTmetrix tool, or PageSpeed Insights. Hm. That’s a really good question. What’s the most
important tool for SEO? JOHN MUELLER: We
have multiple tools that measure multiple things.
And I can understand
that that can be kind of confusing at times. In general, these
tools measure things in slightly different ways. So what I usually recommend is
taking these different tools, getting the data that
you get back from that, and using them to discover
kind of low-hanging fruit on your web pages– so
things that you can easily improve to give
your page a speed bump. How does that sound? MARTIN SPLITT: Yeah. That sounds pretty good. Also, the tools
are differently aimed. Some of these tools,
like Test My Site, are pretty high-level, so
everyone understands roughly what’s going on there. Whereas, GTmetrix is
a lot more technical. And PageSpeed
Insights, I think, is kind of in the middle of that. So depending on who
you are catering to, and who you are trying to give this
report to get things fixed, you might use one or the other. So figuring out what is
the low-hanging fruit and using the tool that
gives you the best insight into that for the audience that
you are trying to convince.
Is it a C-level? Is it another marketer? Is it someone from the tech
side, like is it a developer? Then, you’d probably
pick a different tool. JOHN MUELLER: The next
question comes from Onur. Onur is asking, I am testing
an almost empty page on DevTool audits, and it usually
gives me minimum results, which are 0.8 milliseconds
for everything, and 20 milliseconds for FID. What is FID? MARTIN SPLITT:
First input delay. JOHN MUELLER: First
input delay, of course. But sometimes, it gives worse
results for TTI, FCI, and FID. MARTIN SPLITT: Right. OK, let’s talk
about these metrics. FID, we have covered
the first input delay. TTI is the time to interact. That’s when you can first
interact with the page. And FCI is first CPU
idle, which means that there is no more JavaScript
work or other work that needs to be done by the CPU.
JOHN MUELLER: So
it’s the same page, same code, different numbers. Why would that happen? MARTIN SPLITT: Well,
first things first. These measurements
aren’t perfect, right? So if it’s between 0.8
milliseconds and 20 milliseconds, 20 milliseconds
is a lot more than 0.8, but it’s still quite a
short amount of time if you think about it. You roughly have 10 milliseconds
for a single frame to draw. So yeah, 20 milliseconds
isn’t too bad. So you will always see
some, basically, noise in that measurement. And also, don’t get too hung up
on these metrics, specifically. If you see that there’s
a perceptible problem, and there’s an
issue that your site stays working on the main
thread and doing CPU work for a
minute or 20 seconds, that’s what you
want to investigate. If it’s 20 milliseconds,
it’s probably fine. Our next question
comes from Drew. Drew asks us, what
are the best metrics or metrics to look at when
you want to decide if a page is fast or slow? And why or why
not would you just look at things like FCP, which
is the first contentful paint, and FMP, which is the first meaningful paint,
instead of just the scores that these tools give you? JOHN MUELLER: Wow.
I don’t know, Martin. You need to tell me
some more about that. MARTIN SPLITT: Right. OK, so I guess the question
here really boils down to, what’s the metric that
you should look at? And that’s a
tricky one because I guess it depends on the site. It’s the typical
“it depends” answer. If you have just a website where
people are reading your content and not interacting
as much, then I think first meaningful paint
or first contentful paint is probably more important
than first input delay or time to interact. But if it’s an interactive
web application where you want people to immediately
jump in and do something, then probably that metric
is more important. So don’t try to break it down. And that brings
us to the scores. The problem with the scores
is they are oversimplifying things, aren’t they? JOHN MUELLER: Yeah,
it sounds like it. I mean, all of
these measurements sound like they’re
measuring different things and ultimately
trying to understand what a user would perceive
when they access the page.
So a score might be,
I guess, a simple way to look at it overall, but it’s
probably not all of the details that you’d need. MARTIN SPLITT: It just gives
you a ballpark. It’s like, how
fast is this page? Five. What does that mean? It doesn’t convey
meaning, doesn’t it? So I would say, use that to
figure out how you’re roughly doing, and then use
the specific insights the different tools
give you to figure out where you have to improve
or what isn’t going so well.
JOHN MUELLER: Wow. Yeah, sounds like speed
is a tricky topic. And you kind of have to
know what you’re measuring, so that you can take
action on the right things. MARTIN SPLITT: Yep. JOHN MUELLER: So
would that explain why there is no simple number
that Google is just giving? MARTIN SPLITT: Yeah, so
that explains it. Because if you think about
it, you can’t break down speed into one simple number. It is a bunch of factors. If I am painting
quickly, but then my app is all about interaction,
it’s a messenger, so I show everything, I
show the message history. But if I try to answer the
message that I just got and it takes me 20 seconds
until I actually can tap on the input field and
start typing, is that fast? Not really. But is it so important that
I can use the contact form at the bottom of a blog post
within the first 10 seconds? Not necessarily, is it? So how would you put
that into a number? You don’t.
So I guess it’s hard. JOHN MUELLER: Speed? Speed sounds hard. What do you think? Will this get easier? MARTIN SPLITT: I guess
it will get easier, but it will never go to
a point where you just have a score that you optimize
for and be done with It, right? It is such a broad topic that
it’s really hard to break that down into one number. JOHN MUELLER: OK. So you imagine the more
advanced people will continue to focus on the counting metrics
and counting milliseconds, and others will look at a
bigger overview picture? MARTIN SPLITT: I guess so. JOHN MUELLER: And
together, we’ll try to find ways to improve the
speed of the pages, overall. MARTIN SPLITT: I think browsers
are also doing a lot of work to make things
faster, in general, and easier to understand.
But generally speaking,
you will still need to go and do the
work of figuring out what matters to you, your
audience, and your website, right? Is it interactive? Is it contentful paints? Depends. JOHN MUELLER: That sounds cool. Yeah. So I expect more
questions on speed on the #AskGoogleWebmasters. And as we get those
questions, we’ll ask an expert, like
Martin, who knows all of this three-letter
abbreviations, and who can help us figure out
which ones are the right ones. So thanks, for submitting
all of these questions. And hopefully, see
you again on one of the future episodes, Martin. MARTIN SPLITT: Hopefully. Thank you, very
much, for having me. And thanks, for
all the questions. [MUSIC PLAYING].
𝗢𝗿𝗮𝗰𝗹𝗲 𝗼𝗳 𝗛𝗲𝗮𝘃𝗲𝗻𝘀 ᶦˢ ʸᵒᵘʳ ᵍᵘᵃʳᵈᶦᵃⁿ ᵃⁿᵍᵉˡ ᵗʳʸᶦⁿᵍ ᵗᵒ ˢᵉⁿᵈ ʸᵒᵘ ᵃⁿ ᵘʳᵍᵉⁿᵗ ᵐᵉˢˢᵃᵍᵉ? ɪꜰ ʏᴏᴜ ꜱᴇᴇᴋ ɢᴜɪᴅᴀɴᴄᴇ ᴀɴᴅ ɪɴꜱɪɢʜᴛꜱ ɪɴᴛᴏ ᴛʜᴇ ᴘᴀꜱᴛ, ᴘʀᴇꜱᴇɴᴛ, ᴀɴᴅ ꜰᴜᴛᴜʀᴇ ᴡɪᴛʜ qᴜᴇꜱᴛɪᴏɴꜱ ᴀʙᴏᴜᴛ ʟᴏᴠᴇ, ʀᴇʟᴀᴛɪᴏɴꜱʜɪᴘꜱ, ᴏʀ ᴍᴏɴᴇʏ – ᴄᴏɴɴᴇᴄᴛ ᴡɪᴛʜ ʏᴏᴜʀ ᴀɴɢᴇʟ ᴛᴏᴅᴀʏ https://aef5aa-t-ztics23v7-ljxbw4j.hop.clickbank.net/
