Is more content better? SEO Mythbusting
LILY RAY: So what would
you say for a publisher type of website, that might
talk about the same topic every year, maybe the content
is a little bit different, but it's largely the
same conversation? Like, let's say they're talking
about a certain type of skin care treatment. And they talk about it in
2017 and 2018 and 2019. MARTIN SPLITT: Right. LILY RAY: Do you
think they should take the same piece of content
and update it each year, or should they have three
different pages for that topic? MARTIN SPLITT: Well,
if it's– it depends on if it's an incremental change
that happened, as in, like, if the skincare routine is
pretty much the same as it was last year, you can maybe
rephrase it a little bit. But I would say you
update the existing page and maybe just reposition
it somewhere more prominently on your website
for the visitors to see. But I wouldn't create a
new page that basically says the same thing, because,
especially when they're really similar, we
might just think one is a duplication
of the other and then canonicalize
them together, no matter what you're
doing canonical tags.
[MUSIC PLAYING] Hello and welcome
to SEO Mythbusting. In this episode, my guest
is Lily Ray, SEO director at Path Interactive,
and we're going to discuss an interesting
topic that you might want to learn more about. What is it that you
brought for us today? LILY RAY: We're going to talk
about if too much content is a good thing for SEO or not. MARTIN SPLITT: Hmm. All right. So what is it that people
believe about this? What is the questions that you
keep hearing and wonder about? LILY RAY: Yeah.
I think a lot of companies
think that maybe content is good for SEO, so we should
produce a lot of it, because it will help us rank
for a lot of different keywords. And maybe we should
put out a new blog post every single week. To the point where their website
has thousands of blog posts and maybe they're not
performing really well. So I think a lot of
people have a question about how much content should I
really have and to what extent does this actually
help my performance. MARTIN SPLITT: That's
a really good question. So I think, just going
back to the basics, your key is to provide
information to your users, right? How much content
is good for that depends a little bit
on what you're doing. If you're a news
site, then sure, cover as much of the
happenings that you can, but if your website is
about a specific product, then there is only so
much you can say about it. And just keeping rambling
on in a single page is not helping you much.
LILY RAY: Right. So you would think that
maybe having a blog that talks about industry
updates or things that are relevant for that
company is worthwhile, but maybe not to
just produce content for the sake of
producing content. MARTIN SPLITT: Not for the
sake of producing content. If you have something
like, if you have a product that is very
versatile, and different users or different customers are
using it in very different ways, then this would be an
interesting thing to provide, basically say, like,
oh, look, our product can be used for this. Our product can
be used for that. But just for the
sake of content, that's basically the same as
having light content or useless content, really, and then
you're just spending– crawling, and you're
spending resources on things that are
not performing much.
LILY RAY: Right. Is, like, the presence of
a blog and showing Google that you're producing new
content something that helps your performance overall,
as a kind of site-wide factor? MARTIN SPLITT: Not necessarily. I mean, it is not
a site-wide factor, but if, again, your
blog or your website is about something that
is basically happening on a regular basis or it
has a lot of updates to it, then that can help you bring
relevant content to users that would otherwise maybe not
find your website, especially if your users don't know
about what you're doing, then the blog that reports on
current events or developments can actually help
people understand that, oh, there's
this company that does this interesting thing.
LILY RAY: Right. MARTIN SPLITT: But it doesn't
change your search performance or ranking or anything. It's just providing
something relevant and useful to users that is
going to help you. If you're just putting
it out to have a blog or if you're just
like, hey, we will just have content that keeps updating
and changing without actually giving value to the user,
then that's not helping much. LILY RAY: So if you have
an older piece of content, would you recommend that
if it's a high quality piece of content, do you
need to go back in there and make updates? Or should you only do that
when something significant has changed? MARTIN SPLITT: I think
you should update it if something significant
has changed for sure.
If nothing has
really changed, what you can do instead is you can
write something different, new content, have a
fresh piece of content, and just link that
other piece of content to say, like, hey,
by the way, this is referring00 this is
not about, necessarily, search relevancy or
anything, but it's more helping the user
understand that there is other interesting
content for them. And it's keeping
them on your website, making sure that they
get the information they were looking for. LILY RAY: Definitely. Is there any way
that Google tells us if there's too much
content, or maybe that content's underperforming? Can we look at our crawl stats
to figure something like that out? MARTIN SPLITT: So crawl
stats are a bad place to look at this. Because the fact that we are
not crawling something again does not mean that we
are thinking it's bad or it's good if we're
crawling it often. What's more interesting would
be to look at the Performance Report, for instance,
in Search Console. If you are seeing that you
get a lot of impressions but not that many clicks, you
might want to change something about the content.
If you are getting a lot
of clicks through it, but then you see
in your analytics that actually not
much action happens, then you can ask yourself,
is the traffic worth it? Or do I need to change
my content there? There's no such thing
as too much content. It's just– again, think
from the user's perspective. What is the thing that I
want the user to understand, and is the user interested
in spending time on a page where they need 27
minutes to read everything? LILY RAY: Right.
MARTIN SPLITT:
You get to decide. LILY RAY: Definitely. If there's a lot of content
that's not necessarily performing well on the
website, could that be something that
kind of brings down the overall trustworthiness
or authority of the website from Google's perspective? MARTIN SPLITT:
That depends a bit on what is the reason
for it not performing. If it's spammy content,
if it's very thin content, then that can bring you down
a little bit in terms of, we might just spend
crawl budget on pages that are not performing or not
even being indexed anymore. And you might actually
want to avoid having spammy content and bad
content, especially if you get penalties
or manual actions, you want to definitely
clean up there. But besides that, it is
usually a good idea to see, oh, this piece of content
really does not perform well. Let's take it down or at
least change it, right? LILY RAY: Sure. And what would you
think for companies that have something
like a help center, where there's a
lot of content that answers very specific questions,
but maybe it's only one or two sentences per page,
and maybe they have 500 pages of that nature? Would you say that's something
that they should remain indexed on Google, or how does Google
treat those types of pages? MARTIN SPLITT: That's
a really good question.
That might be treated
as light content, as like thin content and
not necessarily useful. I would try to
group these things and structure them
in a meaningful way. If it's a question about a
specific range of products, then you can group all these
questions together to one page. Or if you have questions in
the category of troubleshooting or operating the thing
that you're trying to sell, try to group this
together to have more dense and helpful
pages in one go. Because how likely is it that
I have exactly one question? LILY RAY: Right. MARTIN SPLITT: If I
have one question, I might have a
follow-up question or I might have a
similar question. So putting these all
together is a good idea. LILY RAY: So
grouping it together. I think that's one
common theme that we talk about a lot in
the SEO world now, is kind of consolidation.
MARTIN SPLITT: Yes. LILY RAY: So do
you think there's a case to be made for one– maybe you have two pieces
of pretty similar content and they would be better
as one single article. So doing a lot of
merging and redirecting. MARTIN SPLITT: Definitely. LILY RAY: That's
something that Google's kind of appreciative when
we do those types of things? MARTIN SPLITT: We have
less crawling to do. That's great. We also know where to
send the users, then.
And there's a chance that
if you have similar things, that these come from
organizational reasons. Like it's one department
thinks about it and another department thinks
about it and none of these two talk to each other. So if you consolidate that,
you bundle more relevancy and information in one place,
and that makes it easier for us to figure out, oh, yeah. This is a good site. Check this out. And get the user
this information, rather than cannibalizing
each other or just like being duplication. LILY RAY: And what
about word count? SEOs are always
asking, is word count a ranking factor, which I
think Google's talked about quite a bit recently. MARTIN SPLITT: We've talked
about this quite a lot.
And it's not a ranking factor. If you can say what the user
needs to know in 50 words, that is fantastic. If you need 100
words, that's cool. If you need 2,000 words, that's
also absolutely acceptable. It's, again, about
trying to figure out what's the intention. If you see yourself repeating
yourself multiple times and saying the same
things over and over again in the same document or on the
same page, what's the point? LILY RAY: Yeah. Well, let's say you're in
a situation where you've written 500 words
for a specific topic or keyword that you're
trying to rank for, and you see all your competitors
have 4,000 words or something like that. Even though word count's not
technically a ranking factor, that's probably
a good indication that you need longer
content, right? MARTIN SPLITT: I
mean, it depends. Just because other
people are doing it doesn't mean that
they're doing it right. LILY RAY: Right. MARTIN SPLITT: Right? So if you see them
rank high, that might not continue to be
that way just because they have a high word count.
Again, try to understand what
is it that the users need. Maybe the larger word
count just accidentally hits the right
bits of information that people are looking
for, and actually fits the query
intention of the user better than what you're writing. In this case, if you
can reformulate it so that your 500 words are
better, then go with that. Don't be the school kid
that goes like, furthermore, as I was saying, just,
like, to fill in– LILY RAY: Unnecessary
language, yeah.
[LAUGHTER] And what's the criteria for
determining if something is spammy or auto-generated? So take, for example, if you
have 50 location pages for 50 states and you want to talk
about the business, which is largely the same
in all those places, but you basically just swap
out the name of the city and maybe add a couple of facts
about that city, for example, how does Google
perceive those pages? MARTIN SPLITT: That's
a really tricky one.
Because either they
work or they don't. LILY RAY: Right. MARTIN SPLITT: So generally
speaking, if you are using generated content and that
is really relevant and good and a human sees this and
goes, oh, I like this, you're probably on
the right track. That can work for
these pages where you have different information
for different cities, but it's pretty much the same
kind of formula behind it. If you have enough
facts around it and there's relevant
information in there that changes city to city,
that might really work. It might also not. If it's too similar,
and you basically– we see that in
places like Germany, sometimes, that there are
literally two sides of a river. And then they are having two
different pages for this, but they say pretty much
the exact same things.
Maybe, like, five
words are different or something like that,
and maybe a few numbers here and there, like a
different number of people there, or whatever. Then we might just consider
one a duplication of the other and not put it in the index. We might de-dupe it and
eliminate it from the index. And then there's
not much you can do. If we think it's the same
kind of content, then what's the point? Why would we show the same
content on multiple URLs? Then we come back to
canonicalization, really.
But if you have
information that is good enough and different enough
from the other bits and pieces, go for it. LILY RAY: OK. So you would
encourage businesses that are in that position where
they do need to target highly localized keywords, that
it's OK to have those pages, but really invest in making
them as unique as possible. MARTIN SPLITT: Make them
relevant and helpful for the user. The user is the
key here, really. And if you're just copying data
over from one place to another, is that that helpful? LILY RAY: Not exactly. MARTIN SPLITT: Not exactly. LILY RAY: Can you
talk a little bit more about how Google treats, or
how does Google determine what duplicate content is? What's the threshold
for duplication? MARTIN SPLITT: I'm actually not
sure what the threshold really is, but I know that we are
basically fingerprinting the content, and
the fingerprint is done in a way that allows
us to say how similar is it.
We use different
similarity metrics and figure out, OK, so
this is pretty much– 95% of this is the same thing. We see– again, we see that in
the German speaking countries a lot where, for instance, a
shop operates in Switzerland and Germany and Austria. All of this is in
German, and then they have the same
products, and the price is slightly different
due to whatever reason. Switzerland has a
different currency, but that's pretty much
the entire difference.
Maybe they use a
few different words because the local
dialects are different. LILY RAY: Sure. MARTIN SPLITT: So
you have 1,000 words in each of the
product descriptions, maybe, and reviews and whatnot. But the price is
different, the currency might be different
if in Switzerland, and there might be like five
different words or something like that. We consider them
all to be the same. And then you can actually
shoot yourself in the foot when you are trying to
canonicalize all of them, because we're like, that's
not a helpful signal. Because we determined
that what you think is individual pieces
of content is actually kind of the same thing. But hreflang and help,
again, and make sure that we are surfacing
the right version. So we might only be indexing
and canonicalizing one of them.
But we will be showing the
different versions of these, depending on where the person
who's searching is located. LILY RAY: Got it. So use hreflang. There's different dialects
and different regions– MARTIN SPLITT: Definitely,
if that's your issue, but if it's literally just the
content is slightly different because maybe you have different
prices or something like that, then we would consider
that the same content. LILY RAY: Got it. MARTIN SPLITT: Thank you
so much for being here and talking with me through
all these questions regarding content and what is good
content and what's too much and what's not enough content. And this was really helpful and
interesting and thanks so much for making it here. LILY RAY: Thanks for having me. I think you answered a lot
of questions that I have, and my clients have.
So I really appreciate it. MARTIN SPLITT: Hopefully this
will be useful for everyone out there, and thank you
very much for watching. LILY RAY: Thank you. [MUSIC PLAYING] MARTIN SPLITT: So
the next episode we have Barry Schwartz
visiting, and we're going to be talking about
SEO community and Google and how we can, you know,
make the relationship better, hopefully. BARRY SCHWARTZ:
Looking forward to it..
