0:07
In this lesson, I plan to review industry studies that show how people online
engage with content, and what types of content they share and or link to.
By the end of this lesson, you'll have some perspective on the best types of
content for getting both shares and links.
So first, here's a study that was done by Chartbeat about whether or
not we read the articles we share.
And it's fascinating if you look at this chart closely, because most
of the articles that have high read time get very low social sharing.
You see the high read time articles are on the bottom right there.
And our top right quadrant is almost empty.
And in addition, most highly socially shared articles get really low read time.
So you see that seems to be a disconnect here.
The things we share a lot, we don't read so much, and the things we read a lot,
we don't share so much.
Goodness, why could that be?
So, I want to show you data now from a second study.
This one was done by Moss and BuzzSumo.
They looked at over 100,000 random posts to try and
see what it took to get links and shares and what kinds of content did that.
And at this first level chart that I'm showing you,
75% of those posts got zero external links.
Remember when I told you that good was not good enough
when it came to content marketing?
Well this is why, the great majority of stuff just simply fails, and
that same 75% by the way got 39 or fewer shares on social media.
Let's take a deeper look at the data, though.
First of all, there's no correlation between shares and links, and
that's what I'm going to show you now.
Before I go into that, let's understand how this was measured.
I used a technique called Pearson correlation coefficient.
You don't have to worry too much about what that is.
But, just understand that it ranges from -1,
which is a total negative correlation to +1, which is a total positive correlation.
And you're looking for things obviously that have positive correlation.
And once you get to 0.3 and higher, +0.3 or higher,
that starts to represent a reasonable or significant positive correlation.
Across all 100,000 posts,
what was the correlation between shares and external links?
0.01.
Or about one-thirtieth of a significant correlation.
Effectively, the correlation was insignificant.
So, interesting thing is that no social network does well with this.
Facebook, Twitter, LinkedIn, Pinterest, Google+, they all do pretty badly.
Twitter actually does slightly better than the rest,
but you know what, it's just nothing to write home about.
It's not really about what social network you are choosing,
the dynamic is different than that.
So ultimately it gets to be a top level view like this.
Most content that gets either share or links, gets only shares or links.
And in this chart, where I'm showing the intersecting bubbles,
I probably have them overlapping too much.
But the overlap isn't that strong.
But we are going to dig into what kind of content gets both shares and links.
Here is a summary of all the posts sampled.
In this summary, it was actually 757,000.
They actually hit 69,000 posts with over 10,000 shares, and
even those still only got a correlation of 0.1 or
1/3 of what we considered significant correlation.
However, there are some sites with a very high correlation.
Here are a few that they found in the study,
0.9 through 0.73, those are all very high levels of correlation.
Yep, the sample size for these particular sites is pretty small.
The Economist isn't bad,
but the number of articles in the others is relatively small.
But I'm going to give you another view of some other sites that also still have
a strong correlation with a large number of articles in the sample size.
So if you look at this,
these have a strong correlation between links and shares, well why?
Well let's think about that for a second.
They all focus on opinion-forming journalism.
4:16
Now, keep in mind this only works if you've got the reputation or
somebody cares about your opinion.
Someone like the New York TImes, of course, does that.
But, if you are in that position, you've been able to establish that
kind of reputation, then you can create content that gets both shares and links.
Here's some other sites that also have a very strong correlation.
And, still some pretty strong numbers here.
So what did they have in common?
Well, they focus on data-driven research studies.
This is something else that people appear to really attach to.
Again, this only really works if you're credible as a source for that data.
But if you are, then this can be a great way to go as well.
So what about content length.
Length, how much does that matter?
Well, in this study, there were a whole bunch of articles less than 1,000 words
which is probably the great majority across the entire web, 86% in this sample.
But you look at it, very few get longer than 3,000 words in all the study.
And why this is interesting is longer form content seems to do better in terms
of having a correlation between shares and links than shorter form content.
Now, let's think a minute about why that may be.
The real reason here is that you don't throw 3,000 random words out there.
There's some value and depth to the content.
For now what you need to think about is in-depth content
having a better chance of getting both shares and links as well.
When you begun this lesson, you might have thought that content that gets a lot of
social shares was likely to be the content that got lots of links as well, but
that's not the case.
One reason for
this is people like to share things that cause an emotional reaction.
However, most emotion-evoking things are not necessarily authoritative on important
topics.
As we saw,
it's more authoritative pieces of content that are most likely to get some links.
In the next module,
we'll discuss the concept of anchor content as this is the type of content
that makes up the backbone of a successful content marketing program.
Eric EngeCo-Author of "The Art of SEO"
