We Analyzed 912 Million Blog Posts. Here's What We Learned About Content Marketing

We Analyzed 912 Million Blog PostsHere's What We Learned About
Content Marketing

We Analyzed 912 Million Blog Posts. Here’s What We Learned About Content Marketing
Brian Dean

by Brian Dean · Updated Feb. 19, 2019

We analyzed 912 million blog posts to better understand the world of content marketing right now.

Specifically, we looked at how factors like content format, word count and headlines correlate with social media shares and backlinks.

With the help of our data partner BuzzSumo, we uncovered some very interesting findings.

And now it’s time to share what we discovered.

Here is a Summary of Our Key Findings:

1. Long-form content gets an average of 77.2% more links than short articles. Therefore, long-form content appears to be ideal for backlink acquisition.

2. When it comes to social shares, longer content outperforms short blog posts. However, we found diminishing returns for articles that exceed 2,000 words.

3. The vast majority of online content gets few social shares and backlinks. In fact, 94% of all blog posts have zero external links.

4. A small percentage of “Power Posts” get a disproportionate amount of social shares. Specifically, 1.3% of articles generate 75% of all social shares.

5. We found virtually no correlation between backlinks and social shares. This suggests that there’s little crossover between highly-shareable content and content that people link to.

6. Longer headlines are correlated with more social shares. Headlines that are 14-17 words in length generate 76.7% more social shares than short headlines.

7. Question headlines (titles that end with a “?”) get 23.3% more social shares than headlines that don’t end with a question mark.

8. There’s no “best day” to publish a new piece of content. Social shares are distributed evenly among posts published on different days of the week.

9. Lists posts are heavily shared on social media. In fact, list posts get an average of 218% more shares than “how to” posts and 203% more shares than infographics.

10. Certain content formats appear to work best for acquiring backlinks. We found that “Why Posts”, “What Posts” and infographics received 25.8% more links compared to videos and “How-to” posts.

11. The average blog post gets 9.7x more shares than a post published on a B2B site. However, the distribution of shares and links for B2B and B2C publishers appears to be similar.

We have detailed data and information of our findings below.

Long-Form Content Generates More Backlinks Than Short Blog Posts

When it comes to acquiring backlinks, long-form content significantly outperforms short blog posts and articles.

Long-form content generates more backlinks than short blog posts

You may have seen other industry studies, like this one, that found a correlation between long-form content and first page Google rankings.

However, to our knowledge no one has investigated why longer content tends to perform so well. Does the Google algorithm inherently prefer long content? Or perhaps longer content is best at satisfying searcher intent.

While it’s impossible to draw any firm conclusions from our study, our data suggests that backlinks are at least part of the reason that long-form content tends to rank in Google’s search results.

Key Takeaway: Content longer than 3000 words gets an average of 77.2% more referring domain links than content shorter than 1000 words.

The Ideal Content Length For Maximizing Social Shares Is 1,000-2,000 Words

According to our data, long-form content generates significantly more social shares than short content.

However, our research indicates that there’s diminishing returns once you reach the 2,000-word mark.

The ideal content length for maximizing social media shares is 1,000 to 2,000 words

In other words, 1,000-2,000 words appears to be the “sweet spot” for maximizing shares on social media networks like Facebook, Twitter, Reddit and Pinterest.

In fact, articles between 1k-2k words get an average of 56.1% more social shares than content that’s less than 1000 words.

Key Takeaway: Content between 1k-2k words is ideal for generating social shares.

The Vast Majority of Content Gets Zero Links

It’s no secret that backlinks remain an extremely important Google ranking signal.

Google recently reiterated this fact in their “How Search Works” report.

Google – How search works

And we found that actually getting these links is extremely difficult.

In fact, our data showed that 94% of the world’s content gets zero external links.

94% of content published gets zero external links

It’s fair to say that getting someone to link to your content is tough. And we found that getting links from multiple websites is even more challenging.

In fact, only 2.2% of content generates links from multiple websites.

Only 2.2% of content generates links from multiple websites

Why is it so hard to get backlinks?

While it’s impossible to answer this question from our data alone, it’s likely due to a sharp increase in the amount of content that’s published every day.

For example, WordPress reports that 87 million posts were published on their platform in May 2018, which is a 47.1% increase compared to May 2016.

Number of posts published (WordPress)

That’s an increase of 27 million monthly blog posts in a 2 year span.

It appears that, due to the sharp rise in content produced, that building links from content is harder than ever.

A 2015 study published on the Moz blog concluded that, of the content in their sample, “75% had zero external links”. Again: our research from this study found that 94% of all content has zero external links. This suggests that getting links to your content is significantly harder compared to just a few years ago.

Key Takeaway: Building links through content marketing is more challenging than ever. Only 6% of the content in our sample had at least one external link.

A Small Number of “Power Posts” Get a Large Proportion of Shares

Our data shows that social shares aren’t evenly distributed. Not even close.

We found that a small number of outliers (“Power Posts”) receive the majority of the world’s social shares.

Specifically, 1.3% of articles get 75% of the social shares.

"Power posts"

And a small subset of those Power Posts tend to get an even more disproportionate amount of shares.

In fact, 0.1% of articles in our sample got 50% of the total amount of social shares.

Top subset of "power posts"

In other words, approximately half of all social shares go to an extremely small number (0.1%) of viral posts.

For example, this story about shoppers buying and returning clothes from ecommerce sites received 77.3 thousand Facebook shares.

"Shoppers are buying clothes" post

This single article got more Facebook shares than the rest of the top 20 posts about ecommerce combined.

Key Takeaway: The majority of social shares are generated from a small number of posts. 75% of all social shares come from only 1.3% of published content.

There’s Virtually No Correlation Between Social Shares and Backlinks

We found no correlation between social shares and backlinks (Pearson correlation coefficient of 0.078).

In other words, content that receives a lot of links doesn’t usually get shared on social media.

(And vice versa)

And when content does get shared on social media, those shares don’t usually result in more backlinks.

This may surprise a lot of publishers as “Sharing your content on social media” is considered an SEO best practice. The idea being that social media helps your content get in front of more people, which increases the likelihood that someone will link to you.

While this makes sense in theory, our data shows that this doesn’t play out in the real world.

That’s because, as Steve Rayson put it: “People share and link to content for different reasons”.

So it’s important to create content that caters to your goals.

Do you want to go viral on Facebook? Then list posts might be your best bet.

Is your #1 goal to get more backlinks? Then you probably want to publish infographics and other forms of visual content.

We will outline the differences between highly-linkable and highly-shareable content below.

But for now, it’s important to note that there’s very little overlap between content that gets shared on social media and content that people link to.

Key Takeaway: There’s no correlation between social media shares and links.

Long Headlines are Correlated With High Levels of Social Sharing

Previous industry studies have found a relationship between “long” headlines and social shares.

Our data found a similar relationship. In fact, we discovered that “very long” headlines outperform short headlines by 76.7%:

Long headlines are correlated with increased social sharing

We defined “very long” headlines as headlines between 14-17 words in length. As you can see in the chart, there appears to be a linear relationship between headline length and shares.

And this same relationship played out when we analyzed the headlines in our dataset by character count.

Long headlines (100+ characters) are correlated with social shares

As you might remember from 2014, clickbait-style headlines worked extremely well for publishers like Buzzfeed and Upworthy.

And their posts tended to feature headlines that were significantly longer than average.

"This man turned a comment into an awesome lesson" post

Although clickbait isn’t as effective as it once was, it appears that long headlines continue to be an effective tactic for boosting social shares.

There are, of course, exceptions to this rule. For example, this post with a 6-word headline received over 328k social shares.

Keto no-bake cookies post

But when you look at the headlines across our dataset of 912 million posts, it’s clear that content that uses longer headlines get more social shares.

Why long headlines work so well is anyone’s guess. However, I have two theories that may partly explain things.

First, it could be the fact that longer headlines pack more information in them compared to short headlines. This “extra” information may push people to read a piece of content or watch a video that they otherwise wouldn’t, increasing the odds that it goes viral.

Also, longer headlines contain more terms that can “match” keyword searches in Google and on social media sites where people commonly search (like Twitter). Again, this results in more eyeballs, which can lead to more shares.

Twitter search

Key Takeaway: Very long headlines (14-17 words in length) get 76.7% more social shares than short headlines.

Titles That End With a “?” Get an Above Average Amount of Social Shares

One interesting nugget from our data was that “question headlines” seem to be working well right now.

In fact, headlines with a question mark get 23.3% more social shares than non-question headlines.

"Question headlines" get an above-average amount of social shares

For example, here’s a post with a question headline that boasts 3.3M shares:

"What state should you move to" post

Question titles may work because they add an element of intrigue that’s well-documented to increase click-through-rate. Put another way, you might decide to read a post in order to answer the question posed in the headline.

Obviously, question titles aren’t a magic bullet. But using questions in certain headlines may help increase shares and traffic.

Key Takeaway: Question headlines get 23.3% more social shares than non-question headlines.

There’s No “Best Day” to Publish New Content

What’s the best day to publish a blog post?

Well, according to our data, the day that you publish doesn’t make much of a difference.

(At least in terms of social shares)

Social shares by day of the week

We did find that Sunday had a slight edge over other days of the week. However, the difference in shares from content published on Sunday vs. the other 6 days of the week was only 1.45%.

Several industry studies and case studies have set out to answer the “best time to publish content” question. But most are either old (one of the most-cited industry studies I found was published back in 2012) or used a small sample size.

And this is likely the reason that the findings from those studies are so conflicting.

Considering that there’s no advantage to publishing content on a certain day, I recommend researching and testing the best publishing time for your industry and audience.

For example, after extensive testing, we found that publishing on Tuesday morning (Eastern) works best for the Backlinko blog. But I’ve heard from other bloggers that their publishing on Saturday works best for them.

So the “best” day to publish is ultimately whenever your audience is available to consume and share your content, something that’s best determined by testing.

Key Takeaway: There’s no “best” day for new content to come out. Shares are essentially equal across different days of the week.

List Posts and “Why Posts” Get a High Level Of Shares Compared to Other Content Formats

We investigated the relationship between content format and social shares.

Our data shows that lists posts and “Why Posts” tend to get more shares than other content formats.

List posts and "why posts" get a high level of shares

For example, this Why Post from Inc.com was shared on Facebook 164 thousand times:

Why reading books should be your priority

On the other hand, how-to posts and infographics don’t get shared on social media very often.

That’s not to say you should avoid any particular content format. There are infographics and how-to posts out there that generate tens of thousands of shares.

However, our data does suggest that focusing on list posts and Why Posts may increase the odds of your content getting shared on social media.

Key Takeaway: List posts perform well on social media compared to other popular content formats. Our study found that list posts generate 203% more shares than infographics and 218% more shares than how-to articles.

“Why Posts”, “What Posts” and Infographics Are Ideal Content Formats for Acquiring Backlinks

We found that “Why Posts”, “What Posts” and infographics get linked to more often than other content formats.

"Why posts", "What posts" and infographics are heavily linked to

What’s interesting is that, while there’s some overlap, there’s a significant difference in the content formats that people share and link to.

Referring domains .vs. Average social shares

While our study found that list posts were the top content format for social sharing, they’re dead last in terms of getting backlinks from other websites.

For example, this list post has 207.8k social shares.

20 amazing writing prompts

But according to BuzzSumo, despite all those shares, this article has zero backlinks:

BuzzSumo – boredpanda.com – Shares

It’s a similar situation with infographics. Our data shows that infographics tend to get very few shares relative to list posts, “what posts” and videos.

However, when it comes to links, infographics are a top 3 content format.

This supports our other finding from this research that there’s no correlation between shares and links.

My theory on this is that certain formats are primed to get shared on social networks like Facebook and Twitter. And other formats designed to get linked to from the small group of “Linkerati” that run and contribute content to websites.

Infographics illustrate this contrast perfectly.

Although the occasional infographic may go viral, it’s fair to say that their novelty has worn off in recent years. Which may explain why infographics aren’t shared very much compared to other formats (like list posts).

However, due to the fact that infographics contain highly-citable data, they work as an effective form of “link bait”.

Also, unlike a list post or how-to post, infographics can be easily embedded in blog content. This further increases the chances of acquiring links.

Key Takeaway: “Why Posts”, “What Posts” and infographics appear to be ideal for link building. These three formats receive an average of 25.8% more referring domain links than how-to posts and videos.

B2B and B2C Content Have a Similar Share and Link Distribution

We analyzed a subset of content from our dataset that was published on B2B websites. Our goal was to find out if share and link behavior differed in the B2B and B2C spaces.

First, we did find that “normal” content generates significantly more shares than B2B content. In fact, the average amount of shares for all the content in our dataset is 9.7x higher than content published in the B2B space.

B2C content gets shared 9.7X more than B2B content

This finding wasn’t surprising. B2C content tends to cover topics with broad appeal, like fitness, health and politics. On the other hand, B2B content on hiring, marketing and branding only appeal to a relatively small group. So it makes sense that B2C content would get shared more often.

However, when we analyzed the distribution of B2B shares and links vs. all published content, we found that they largely overlapped.

For example, 93% of B2B content gets zero links from other websites.

93% of B2B content gets zero external links

The amount of B2B content without any links (93%) is similar to the figure (94%) from our full dataset.

The percentage of B2B posts get linked to from multiple websites also overlaps with B2C.

Only 3% of B2B content gets linked to from more than one website.

Only 3% of B2B content generates links from multiple websites

This largely matches the 2.2% that we found in our mixed dataset of B2B and B2C content.

Overall, B2B and B2C link distribution largely overlaps.

Similar share and link distribution

When it comes to B2B social shares, we found that 0.5% of B2B articles get 50% of social shares.

B2B "power posts"

And 2% of B2B articles get 75% of social shares.

B2B subset of "power posts"

Like with B2C content, B2B publishers have a small number of “Power Posts” that drive the majority of social sharing.

B2B and B2C shares stem from a small number of "power posts"

Key Takeaway: Although B2B content doesn’t get shared as often, the distribution of shares and links in B2B and B2C appears to be similar.

Conclusion

I learned a lot about content marketing from this study, and I hope you did too.

I’d like to again thank BuzzSumo (in particular Henley Wing) for providing the data that made this research possible.

For those that are interested, here is a PDF of how we collected and analyzed the data for this research.

And now I’d like to hear from you:

What’s your #1 takeaway lesson from this study?

Or maybe you have a question.

Either way, leave a comment below right now.

424 Comments

  1. Brian, top man! I always suspected that content divided into topics generated more engagement, and now you’ve proven it!

    1. Brian Dean Avatar Brian Deansays:

      Thanks Natanael.

  2. Very thought provoking indeed Brian. Some definite surprises in there and seems that a lot of content doesn’t achieve much. Will have to read post a second time as a lot to digest there! 🙂

    1. Brian Dean Avatar Brian Deansays:

      Thank you, Paul. I was also surprised by how few posts got any links. Goes to show that blasting out content without promotion doesn’t make much sense.

  3. Great stuff! 912 million! Thats some analysing! Keep up the great work Brian, recommended to all our clients.

    1. Brian Dean Avatar Brian Deansays:

      Thanks Matt. I hope you learn some cool new stuff.

  4. Very nice case study and report!

    It’s interesting that only 6% of the content analyzed even has a single link, but it’s still getting enough traction on BuzzSumo to get pulled into this study.

    Thank-you for putting this together.

    1. Brian Dean Avatar Brian Deansays:

      Thanks Lisa. Good point there. I’m not 100% sure how BuzzSumo finds content. But I think it crawls social media sites (like Twitter) vs. finding content through links on other websites.

    1. Hi! Susan from BuzzSumo here…we index everything that is socially shared, basically adding 2.5 million URL’s /day to our database. The latest total was 5 billion, but we are on track for 6 billion in the next 6 months.

      1. Wow Susan, that’s an amazing number! Great work and thanks for the excellent, eye-opening article Brian!

      1. Johan Avatar Johansays:

        Thank you for this huge work. Was this research made on US content ? or also on french, german, italian content ? and about “best day” : there were no difference between B2B and B2C ?

  5. Once again, another great article, Brian!

    1. Brian Dean Avatar Brian Deansays:

      Thanks Jason 👍 👍 👍

  6. Awesome findings, kinda concurs with my own findings. The post with the most social signals on my site is actually not that long. While I have put more effort in some other posts which are longer and they dont have that much social shares. Im going to keep it in the back of my mind when I create new content.

    1. Brian Dean Avatar Brian Deansays:

      Thanks Koen. At the end of the day, people will share if content’s shareable. And publishing super long stuff isn’t a magic bullet, which our research shows. But for links, longer content (3k+ words) probably does help.

  7. Hadassah Avatar Hadassahsays:

    Thanks for this brilliant research, Brian. I had been itching to read a new post from you (I always anticipate your emails, lol) and you just gave me the secrets I have been looking for. Content marketing is different in 2019. Thank you for going all the way once again!

    1. Brian Dean Avatar Brian Deansays:

      Thank you 🙂

  8. Really interesting! I thought that B2C posts would actually get more backlinks than B2B ones. Thanks.

    1. Brian Dean Avatar Brian Deansays:

      You’re welcome, François. To be clear: we didn’t actually look at the raw number of links content in the B2B and B2C spaces got (although that would be interesting). But the overall distribution was the same.

  9. Amazing research Brian! Your tips have helped me get content for my SaaS ranked on Google 🙂

    1. Brian Dean Avatar Brian Deansays:

      Nice! Lots of golden nuggets in here that should help your rankings even more.

  10. Ahmed Avatar Ahmedsays:

    I can barely wrap my head around that number! Good Stuff man!

    1. Brian Dean Avatar Brian Deansays:

      It’s a big one! Big props to Henley from BuzzSumo who is a master at big data.

  11. Thanks for your hard work and advice Brian!

    1. Brian Dean Avatar Brian Deansays:

      You’re welcome, Javier. Hope you get some value out of today’s study.

  12. Great research, Brian!

    Have you covered any legal blogs?

    If so, what might be their share?

    Thanks!

    1. Brian Dean Avatar Brian Deansays:

      Thanks Dmytro. We didn’t look at legal blogs in particular. Although that would be a really interesting niche to investigate.

  13. Great information, as always Brian! Will dig in deeper & absorb it all. Thanks for all your hard work.

    1. Brian Dean Avatar Brian Deansays:

      You’re welcome, Darshana. Let me know what you think after you have a chance to review everything.

      1. I have been following & including all your key summary findings in my blog:

        2000 words or less
        Alt tag for all images
        Engaging title
        Value-based content

        What I haven’t included is a question mark on any of my blogs & that I will do from now on since there’s increased percentage of social shares (great tip from your research).

        1. Brian Dean Avatar Brian Deansays:

          Sounds good. Keep me posted.

  14. I would think that the majority of the content which gets a lot of social media shares has no link value such as content on Buzzfeed or Bored Panda so that could skew the data. Personally, I’ve seen a lot of success using social media to promote content to gain backlinks so I am not sure I would be ready to write that off yet.

    1. Brian Dean Avatar Brian Deansays:

      Hi Ronald, That’s true to a certain extend for sure. But considering out data set included 900+ million posts, it’s not likely that the results were skewed by a handful of viral sites.

  15. I just did a quick scan of this gem of a post, Brian. First, THANK YOU! It’s filled with valuable and counterintuitive insights, essential info for anyone in the content marketing game. Much obliged!

    1. Brian Dean Avatar Brian Deansays:

      Thanks Alan. That’s very true: a lot of these findings are surprising (like the fact that people share and link to content for different reasons).

  16. Great study! Among other things, this is further support for making sure your content strategy is customized for specific channels, rather than just “blasting” the same version of content out everywhere.

    1. Brian Dean Avatar Brian Deansays:

      EXACTLY. Content that people link to doesn’t usually get shared. That’s why your strategy has to fit with your goal. If it’s links, you go one way. If it’s shares on social media, it’s time for a completely different approach.

  17. Great detailed post as usual. So many takeaways from this post. I noted why and what post fare better.Thanks for this awesome content

    1. Brian Dean Avatar Brian Deansays:

      Happy to help, Vijay.

  18. “Vast amount of content have zero links” main takeaway but the article highlights so many factors to reconsider

    1. Brian Dean Avatar Brian Deansays:

      That was a big one for me too. I knew that most content didn’t get linked to. But I had no idea that the number was so low.

  19. Great research Brian

    1. Brian Dean Avatar Brian Deansays:

      Thank you, Felix.

  20. Wow, awesome case study Brian. Looks like still long form content performing the best. Working on similar types of content right now. Fingers cross.

    thanks for this awesome case study. Seems I m on right path.

    1. Brian Dean Avatar Brian Deansays:

      You’re welcome, Navin. Sounds like you’re on the right track.

  21. Brian, it’s great information on what content gets attention and links. But it’s almost disheartening to think that someone might spend two weeks researching and writing a really great post – only to have no one ever find it.

    I suppose the biggest takeaway for me is that if you have a really great idea for a post, article, etc., AND you want people to find/link to it, you probably need to spend hours up-front strategically defining your content rather than just sitting down and banging it out. It almost seems like true authors need to either acquire an additional skill set OR work with a partner who looks at the content exclusively from the visibility stand-point.

    Thanks for another great post.

    1. Brian Dean Avatar Brian Deansays:

      You’re welcome, Bob. Well said. I agree that you need a strategic approach if you want your content to get seen. That includes findings the topic, outlining your content and promoting it. Otherwise, it gets buried.

  22. I love content and fresh data is so vital to keep it relevant. Great piece – as always Brian! Thank you!

    1. Brian Dean Avatar Brian Deansays:

      You’re welcome, Jodie. Data is where it’s at

  23. Spyros Demetrios Avatar Spyros Demetriossays:

    Hi Brian. Do you sell a course? I want to get into YouTube. And need the right mentor

    1. Brian Dean Avatar Brian Deansays:

      Hi Spyros, I do.

  24. Goes to show that no matter how much we hate it, clickbait headlines remain crucial to garnering social shares.

    1. Brian Dean Avatar Brian Deansays:

      What makes you say that?

  25. Thank you, sir, for the quality stuff you’ve been sharing…I really appreciate.

    1. Brian Dean Avatar Brian Deansays:

      No problem.

  26. Hi Brian,

    Great information as always. I do find it very interesting that long-form content gets more social shares and views. Normally, when I publish longer articles, they tend to not do so well. However, when I post short posts (which are usually news related posts), they tend to do better. 300 to 500 words each post. Perhaps my long-form stuff is too long? I do a golf blog, and posts about players seem to do decent enough because the main keyword is easy to spot (in these cases, the player’s name).

    However, if it’s not about players, they tend to not do as well minus a few exceptions. One other thing I’m struggling with is bringing more followers to my golf blog’s Twitter page. Of course, tweeting out good content helps, but I feel like there’s more to it than that. Am I wrong?

    Finally, I’m planning on changing the domain name of golf blog, but the name I wanted was taken with “.com”, but was available with “.co”. Is using a “.co” domain a good idea or no?

    Thanks so much for your time and help, Brian, and keep up the great work!

    1. Brian Dean Avatar Brian Deansays:

      Hey Colin, in that case it’s probably more about the topic than the length. I’d be curious to see how your news posts would do if they were on the longer side.

      1. Thanks, Brian! I’ll definitely consider longer stories. I’m curious as well. Also, I sent a similar comment like the one above, so please kindly disregard that one.

  27. Hello Brian, Thanks for this huge data. It seems I’m outdated at headlines things since I use short form headlines for most of our sites.

    Also, I publish less number of question type posts since I thought the search volume is lower but seems to be a great mistake.

    1. Brian Dean Avatar Brian Deansays:

      You’re welcome. One thing about question headlines is that they’re great for shares, but maybe not so much for title tags. Not saying questions can’t work as title tags, but we only looked at how they related to social shares.

  28. “However, to our knowledge no one has investigated why longer content tends to perform so well. Does the Google algorithm inherently prefer long content? Or perhaps longer content is best at satisfying searcher intent.”

    Google uses user signals like bounce rate and time on page through Chrome, doesn’t it? Longer article -> in general longer time on page, that is to Google more fitting search result -> higher ranking. (Of course compared to the competition, so if they have millions of backlinks, those signals won’t help neither.)

    1. Brian Dean Avatar Brian Deansays:

      Hi Boris, I think they do use that data. But assuming that’s the reason that longer content works is very different than investigating it with data.

  29. The content is great and it will definitely help me improve traffic on to my website.

    1. Brian Dean Avatar Brian Deansays:

      I hope so, Umesh. Let me know how it goes.

  30. Lorenzo Avatar Lorenzosays:

    Really interesting. Thanks for all this stuff Brian 🙂

    1. Brian Dean Avatar Brian Deansays:

      You’re welcome, Lorenzo.

    1. Brian Dean Avatar Brian Deansays:

      👍 👍 👍

  31. Amazing. Thanks for all that effort, and for sharing. Much time will be required to fully absorb your findings. But I have learned that a long headline, ending with a “?” and preferably preceding a list, will likely gain the most results.

    1. Brian Dean Avatar Brian Deansays:

      You’re welcome, Dr. B. For social shares, definitely.

  32. Some jaw dropping stats in there! Thanks Brian. It would appear that a large % of the World is wasting valuable time preparing content that never gets viewed, liked or shared!

    1. Brian Dean Avatar Brian Deansays:

      I couldn’t have said it better myself, Simon!

  33. Great post, Brain. An important take away from this post for me is to use ‘?’ at the end of headlines. In a lot of my earlier blogs, although they were why/what posts I didn’t use the question mark in the title, just left it blank. Will change the headlines and study if there is any noticeable change in the number of shares. Thanks again for this post. Cheers!

    1. Brian Dean Avatar Brian Deansays:

      You’re welcome, Kritesh. Let me know how it goes!

  34. This is outstanding and more or less confirms what believe I see on my blog site. I would love to see some additional data in two areas:

    1. Sponsored content v. non-sponsored.
    2. How important SEO is for niche industry blogs.

    Right now I run a pretty successful niche industry blog that makes good revenue primarily from sponsored content. We have a pretty good email list we send to every day (5 day week) and our sponsors are happy with the results… I would like it to be better. Any thoughts on any of this?

    1. Brian Dean Avatar Brian Deansays:

      Thanks Steve. I know sponsored content is big right now so it would be interesting for someone to look specifically at their performance. To your question: I’ve worked with or helped sites in 50+ industries. And in my experience, content is content. Every niche has their own wrinkles. But the same basic rules tend to apply across different industries. Hope that helps.

  35. This is such a brilliant post. I read EVERY word.

    This was what I wanted at this time to validate my content strategy.

    Thank you Brain. This post shows you were not just writing for the money or for the shares or backlinks but to actually help folks like us.

    Thanks again.

    Love,
    Ojo Iszy

    1. Brian Dean Avatar Brian Deansays:

      You’re welcome, Ojo.

  36. Hey Brian, thanks for the insights.

    I’m surprised by the finding that longer headlines = more social shares, because various blog posts, headline generators & plugins like Yoast say to keep it 10 words or less.

    Also, regarding long-form content getting more backlinks: does that apply to new blog posts only, or updating content along the way?

    eg. publishing a 3,000 word piece all at once, or updating a 1,000 word piece over a period of several months till it reaches a higher word count?

    Thanks!

    1. Brian Dean Avatar Brian Deansays:

      You’re welcome, Heenay. Yoast is (rightfully) really focused on SEO. So in that case, it would make sense to keep your headlines on the shorter side so your title tag doesn’t get cut off. But social shares appear to be a different animal.

      Regarding long-form content, good question. I believe the word count is measured when the page is first indexed by BuzzSumo. But I’m not 100% sure.

      1. Ah, thanks for that last bit of info. Appreciate it!

  37. I wonder if the reason long content gets more links is because the people writing short content don’t care about or don’t know about link building.

    1. Brian Dean Avatar Brian Deansays:

      LOL! I actually never considered that. It does seem that most peeps in the know have all started to publish longer stuff.

  38. Dilshad saifi Avatar Dilshad saifisays:

    First of all Thank you for sharing this blog.

    I have a question:-
    If i want boost my webaite ranking so i have to focus on social shares or backlinks.

    Means i have to write list content on my website blog or article that contains “why” & “what” in title???

    1. Brian Dean Avatar Brian Deansays:

      Definitely focus on backlinks.

      1. Dilshad saifi Avatar Dilshad saifisays:

        Thank you

  39. can not be refuted long content is very powerful in SEO techniques

    1. Brian Dean Avatar Brian Deansays:

      Pretty much. Obviously, the content has to be great. But longer does seem to do better in general.

  40. Continue to write good content…that’s gonna give me the best opportunity to have my content shared. Thanks for the info!

    1. Brian Dean Avatar Brian Deansays:

      Writing good content is a good start, Bret. But as this research shows, it’s usually not enough.

  41. Incredibly useful information! Thank you for this. You have also confirmed mine (business partner and I) theory that there is just too much content. Its a hard road to getting back links. My partner keeps shutting me down when I say we need to create a few posts…. This seems to be the reason why.

    I am also almost sure that the power posts generally get pushed out by big sites that have a huge following. Which in turn gives them trust and the reader a reason to share… The youth of today (my sons for example) generally start a sentence by saying… I read “XYZ” on Reddit/Quora and they say… Which makes it even tougher for the small guys to get social shares or even backlinks…

    1. Brian Dean Avatar Brian Deansays:

      You’re welcome, Brad. Glad you enjoyed it.

      You’re onto something there: big publishers definitely have an inherent advantage over small publishers. But my hope with data like this is that little guys like you and me can gain an edge on them.

  42. The information in the PDF is also great! I was thinking, if a correlation could be formed between the 6% of content that atleast gets some external links and if they’re the ones that majorly dominates the Google Search Results. 🙂
    This article surely sets the tone for best practices of Content Writing for 2019 based on Scientifical analysis. Loved it as always 🙂

    1. Brian Dean Avatar Brian Deansays:

      Thank you. We didn’t look into it, but considering how important backlinks are, I’d guess there’s a strong correlation there.

  43. chechu Avatar chechusays:

    I’m working on get a deeper understanding on Headlines, are absolutely top for me. Thanks for sharing this stuff!!

    1. Brian Dean Avatar Brian Deansays:

      You’re welcome. Headlines are HUGE

  44. Incredible stuff. Thank you so much. Gave me serious clarity that too very early on my blogging career. In fact, you inspire me to produce top quality content.

    1. Brian Dean Avatar Brian Deansays:

      Thank you

  45. Great info, Brian. You never fail to amaze me. This helps us to focus on the right things to rock blogging. Will be waiting for more surprising case studies.

    1. Brian Dean Avatar Brian Deansays:

      Thank you, Siva.

  46. Tawfek Elsayed Avatar Tawfek Elsayedsays:

    So great case study. I really love it 😊. But I have one thing that i don’t understand it which is “list post”
    What do you mean by list post?

    1. Brian Dean Avatar Brian Deansays:

      Thanks. 20 Ways to X, 17 Tips for Y. Stuff like that.

  47. Hi, Brian!
    This post has great value.
    Thank you 🙂

    1. Brian Dean Avatar Brian Deansays:

      You’re welcome.

  48. Wow, this good information to process. Interesting about the longer titles and longer content. Appreciate your hard work and research.

    1. Brian Dean Avatar Brian Deansays:

      No problem, Tisha. Glad that you learned some new stuff that you can use.

  49. Duc Thang Bui Avatar Duc Thang Buisays:

    Interesting! I just thought that B2C posts may be more easier to be viral than B2B ones, and this means more backlinks. But, I’m wrong. Hahah

    1. Brian Dean Avatar Brian Deansays:

      Well, B2C content does get more overall shares. But the distribution of shares and links is almost identical in B2B, which surprised me.

Leave a Comment

Your email address will not be published. Required fields are marked *