We Analyzed Nearly 1 Million Headlines. Here’s What We Learned

July 22, 2014 - Get free updates of new posts here

This is guest post by Garrett Moon, Founder at CoSchedule

Here at CoSchedule, we help our customers plan their blog and social media content on a single drag and drop calendar. This is cool, but what it really means is that we have access to a ton of data about blog headlines, including where those headlines get shared online.

After reaching nearly 1 million blog post headlines in our system, we began wondering about what they could reveal about growing traffic and writing better headlines. Specifically, we wondered:

What is it that makes one headline more shareable than another?

The Data

For the purposes of this post, we began with a dataset of nearly 1 million headlines. I trimmed that number down to English-only posts that had already been published. Then I whittled it down further to a group of headlines that had received at least 100 total shares across all of the major social networks.

From there, I created another group of headlines that had more than 1,000 shares. This lead to my first major insight.

Most content doesn’t get shared all that much.

Most Content Does Not Get Shared

Just let this chart sink in for a minute. This means that 89% of the content that is created is never shared more than 100 times! Besides coming off as a bit depressing, this should be major motivation for you to do things differently.

In this post, I am going to specifically focus on the headlines that fell in the top 11% of these results. You are going to see what makes high performing headlines work, so that you can apply the ideas to your own content and put yourself in the top-tier of high-performing blogs.

What common words/phrases are used in highly-shared headlines?

Most Popular Words Or Phrases in Highly Shared Headlines

To start, I went through a few common words and phrases used in headlines that were shared more than 1,000 times. The results were telling.

Takeaway #1 – List Posts Are Huge

One of the first things that I saw was that lists posts are huge and were the most likely type of post to be shared more than 1,000 or even 100 times. More interestingly, list posts only made up 5% of the total posts actually written, which means that we don’t create enough of these posts to begin with. An immediate takeaway here is to start creating more list posts.

Takeaway #2 – Use ‘You’ & ‘Your’ A Lot

Posts that used words like ‘you and your’ in their headline performed extremely and were shared frequently. In contrast, posts that used ‘I and Me’ we three times less likely to be shared. This suggests content that written in the second person – the point of view you take when you are speaking directly to the reader – is far more likely to be shared than content that comes from a first person narrative. Obviously, readers like to see themselves in what they are reading.

Takeaway #3 – Help Your Readers Imagine A Better Life

People really like content that helps they do something awesome, like win something for free or learn something new. We can easily see this in the frequent occurrence of words like ‘free,’ ‘giveaway,’ and ‘how to.’ Use promising words to your readers, and that will make them more likely to share your content with others.

How do common headline words change based on social network?

I wanted to see how the frequency of keywords changed in the headlines when they were broken down by social network. I was pretty surprised by the results.

Fact: Both Facebook and Google+ are surprisingly home-oriented, with top words like ‘recipe’ and ‘homemade.’ Twitter on the other hand tends to be more business and technology focused.

Common Words Phrases in High Shared Headlines

Takeaway #4 – List-posts Do Best On Facebook, Twitter, and LinkedIn

We already know that list posts get shared like crazy, but which networks reward them the most? In a list of some of the most common terms for used on network, there were several words that seemed to indicate the use of a list-based post. On Facebook, Twitter, and LinkedIn, for example, word’s like ‘thing, should, and reasons’ seem to clearly indicate the use of a list.

You can easily imagine the corresponding headlines.

  • 5 Things You Can Do To Write Better Headlines
  • 4 Reasons You Should Wash Your Hair Every Day
  • 8 Things Every Mom Says To Her Kids

Interestingly, these list-like terms are also highly emotional terms, a strong signal that the data will touch on more about bit later.

Takeaway #5 – Video Is Most Popular On Facebook

Facebook was absolutely the most popular network featuring video content, and the only network that had the word included in its top headlines. This likely has something to do with the way that Facebook itself embeds videos directly in the news feed.

Takeaway #6 – Customize Headlines For Each Social Network

Each network has its own audience and demographic and should be catered to individually. For example, in our results, Facebook and Pinterest tended to be home-oriented, whereas Twitter and LinkedIn tended to stay more business focused. Different audiences require different types of content.

One way to accommodate this would be to write custom headlines for each social network that caters to the specific audience rather than just sharing the same old post title on each network. We try to make this type of social sharing as easy as possible with CoSchedule because we know how important it can be.

Where do the world’s most popular headlines get shared?

Part of understanding how shareable headlines work comes from understanding how users will be sharing our content. When we took a look at which social networks were contributing the most shares, the results were pretty astounding. Pinterest totally killed it.

Share Distribution By Network

Takeaway #7 – Pinterest Offers HUGE Shares If You Can Reach The Audience

Among headlines shared more than 1,000 times, Pinterest commanded an astounding 90% of total shares. This simply blew my mind. Pinterest is huge! Of course, you have to have the right type of content to scale this network (see above), but it is ripe for the picking if you know how to do it.

Share Distribution By Social Network w/o Pinterest

Takeaway #8 –At The End Of The Day, It’s Hard To Beat Facebook

Once Pinterest is removed from the scenario, Facebook comes away as the most popular network for social sharing.  In their epic slide deck The Sweet Science Of Virality, Upworthy also makes this claim. If you want big shares, Facebook and Pinterest seem to be where it’s at.

* * *

Can you predict the popularity of a headline? (YES!)

Digging in a little more, I wanted to see if there was an easy way to “rate” a headline and essentially predict (as best I can) if it is going to be a well shared post or not.

To figure this out, I started testing some of the most well-shared headlines using the Emotional Marketing Value Headline Analyzer as made available by the Advanced Marketing Institute. This handy little tool promises to tell you how “emotional” your headline is by counting the number of emotional words that are used in the phrase.

Emotional Marketing Value Headline Analyzer

The headline analyzer is easy to use. Simply copy/paste your headline into the box and will give you a calculated score of your headline’s EMV Score. Here is the result for the headline of this post:

Headline Analyzer Score

Here’s how the headlines analyzer works:

This score indicates that your headline has a total of 30.00% Emotional Marketing Value (EMV) Words. To put that in perspective, the English language contains approximately 20% EMV words.

And for comparison, most professional copywriters’ headlines will have 30%-40% EMV Words in their headlines, while the most gifted copywriters will have 50%-75% EMV words in headlines.

A perfect score would be 100%, but that is rare unless your headline is less than five words.

So, the question is: does it really work as a way to predict popularity? To test, I took an average sampling of headlines in three different sharing groups.

  1. Posts with greater than 1,000 shares
  2. Posts with 500 shares
  3. Posts with 100 shares

Here were the results:

Average EMV Score for Headlines Based on Shares

Posts with a high number of shares frequently reached an EMV Score of 30 or 40, several points higher than posts with fewer shares.

HUGE Takeaway #9 – Emotional Headlines Get Shared More

What I found was that as the number of shares also increased, so did the EMV score of the headline. This means that headlines with a higher EMV Score are more likely to be shared more that posts with a lower EMV Score. Essentially, there is a direct link between the number of emotional words used in a headline and the likelihood it will be shared more than 1,000 times. This is a HUGE takeaway that we can all put to work on our posts right away.

Of course, I wanted to verify my data on this one, since it is such a huge point. I wondered: what would happen if I compared the 5 most shared and the 5 least shared posts on some of the world’s most popular blogs? Would the EMV Score continue to be a good indicator of sharing?

The answer is yes.

Most Popular vs Least Popular

Takeaway #10 – You Can Easily Quantify The Emotional Value Of A Headline

On three of the most highly shareable blogs out there, the posts with more shares had, on average, a higher EMV Score than those posts with fewer shares. Wow! Based on these results, we should all be shooting for an EMV headline value of around 30 or above.

Of course, an EMV of 40 or more will significantly improve our chances of getting more than 1,000 shares for our post, as indicated from the results above.

* * *

At the end of the day, your headline will make a huge difference in the number of shares that a post receives, but there are several things that we can do to help ‘manufacture’ that virility. A good way to start, might be by analyzing the average EMV Score for some of your most popular posts. You can use this handy method to gather the data that you need for free. You can also take a look at this post, which takes a similar look at extremely viral headlines.

Bonus Takeaway #11 – Get A Free Mega-Trial of CoSchedule!

For 1 person who comments, they’ll get a lifetime premium plan to CoSchedule.

For OkDork readers, they are offering an exclusive (and free) 45 day mega-trial of CoSchedule starting right away. It is a pretty great place to put some of the things in this post into practice, so give it a try.

Garrett Moon is a founder at CoSchedule, a WordPress editorial calendar that allows you to schedule your blogs posts and social media together on an easy drag-and-drop calendar. Follow him on Twitter or Google+.

Leave a Reply

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

173 responses to “We Analyzed Nearly 1 Million Headlines. Here’s What We Learned

  1. Eric Reply

    Pretty shocking that Pinterest is getting on their knees and blowing away the competition when it comes to sharing. But since the web is becoming more visual (depending on which expert you talk to) it kinda makes sense.

    What doesn’t surprise me is that list and emotional headlines are the most shared. Buzzfeed may have caught onto this when they launched because they crank out some hilarious and somewhat shocking list posts.

    I guess it’s time to test some headlines with the EMV analyzer and your headlines plugin to see if I can increase my article shares.

  2. Winson Reply

    Thanks Garrett and Noah for this insightful post. I guess the biggest takeaway is the emv value of headlines. It’s always good to have a quick and easy way to gauge your headlines.

    Coschedule looks really cool too. I almost signed up before reading till the end of this post and miss out on the awesome offer.

  3. Daniel Sealock Reply

    The common words and phrases matrix is awesome. That is something that could come in handy down the road. To piggy back off Eric was alluding to, I was surprised to see the ‘shareability’ of Pinterest…..then I started to think how often my wife is on there and it started to click lol. I need to put more time into Facebook.

    I’m also trying to test headlines using unbounce.com for an app idea. This post has already helped. If anybody wants to check it out and give me feedback, I’m open to it. The url is http://www.tweader.co.

  4. Bryan Reply

    Just launched a new project with some help from AppSumo. Short term goal: sell something. Long term goal: build rich content using these headline tips to sustain the business. Thanks for the research, Garrett!

    1. Bryan Reply

      Wrote a headline with a 75% EMV Score with these tips! Let me repay you for my future success with this high-EMV headline for this post ; )

      You Need to Hear the Truth Nearly 1 Million Headlines Told Us (50% EMV – Spiritual)

  5. Shalin Reply

    I got a question for you. I tried few test titles with the Emotional Marketing Value Headline Analyzer and the results I found are bit confusing. Then I made up few titles like “How to 10 reasons you should try and here is why ” . Even though the title does not made any sense. EMV score showed between 33-44. I wounder why can it be like that. I really appreciate the research you have done and fact are very informative.

    1. Garrett Moon Reply

      From what I understand, EMV is measuring the ratio of emotional words/phrases v.s. non-emotional words/phrases. This would mean that word order is not all that important.

      This is also probably why exceeding an EMV of 40 or 50 doesn’t necessarily equal a better headline.

  6. Maia Reply

    Very nice. The influence of the social relevance of words on the number of views are a fact. I think that the customization of the same words and action verbs into every social network headline,will be very relevant.
    I compare it to when I analyze applications jobs to my companies, that the director of human resources sends me to evaluate. If the applicant has taken care to prepare and guide his cv to what we are looking for, his chances incrase in 70%.

  7. Tom Reply

    I think these trends look set to change in the next 12 months, given that web users will become tired/wary of the usuals (emotional bait, list posts etc). It’s interesting to see the big difference between “how” and “why” in titles – speaks volumes about what people want.

    1. Garrett Moon Reply

      Yeah, there are a certain group of headlines that are more ‘sensational’ than emotional, and I think you’re right – those posts will loose efficiencies in the future. Emotional headlines, however, are a fairly standard aspect of journalism and will probably always be fairly effective.

      In the end, though, content quality still wins all things.

      Thanks for the comment – good thoughts to add here.

    2. Nathan Love Reply

      Great point Tom. I think we need to be aware of the fickle nature of users, and the want to avoid being tricked. As Garrett suggests, content quality still trumps all so if you’ve the body to back up the head(ing), you’re all good.

  8. Chris Backe Reply

    The Pinterest percentage was really surprising – but I would’ve figured the text of a pin wouldn’t have mattered much given that it’s such a visual platform. Perhaps the excellent photo is the pre-requisite to get someone’s attention and the text is what pushes someone over the top.

  9. Jonathan John Reply

    Absolutely fantastic post. The amount of research in this post is fascinating! Thanks a bunch, Garrett. I’ll be using that headline analyzer from now onwards in my blog posts.


  10. Swathi Reply

    Amazing post! I’m surprised that Pinterest has the most shares. Lesson learned – use Pinterest more often. Of course history has shown that humans are more visual and that definitely explains it.

  11. Karen Reply

    As a food blogger, I’m rather surprised to see that food-related words are common in highly shared headlines in Google+, Facebook and LinkedIn(!), and not just Pinterest. Would be useful for crafting my next headlines.

  12. ace bhattacharjya Reply

    Loved this post- especially since we’re doing a lot of content work right now. I’m wondering whether anyone has created a told that does some cohort analysis between the EMV headline creator AND seomoz. Could be pretty damn cool- otherwise, you’re competing against buzzfeed and losing.

  13. Nathanial Reply

    Wow! What a great insight into what’s happening right now. Surely more evidence of the ‘snooze you loose’ train that is departing all marketing stations. Appears you need to engage an ‘A-Team’ of social marketing specialists to keep up, or be one of the few who jumps solo into the abyss, only to pop a parachute nobodies ever seen before!

  14. MikeA Reply

    Great stuff and thought-provoking. We are just about to buy Twitter + FB ads for a mini test campaign and decided to check the EMV of our headline “your way to great visibility; surprisingly easy” got EMV of 57%. Good validation of our copy writing. Thanks!

  15. Bryan Zimmerman Reply

    Great article here, but that was the worst downloaded cheat sheet ever. I know you were using it as a email squeeze for coschedule, and I am happy to get on their list, but when you call it a cheatsheet and it’s only a list of the 10 things you mentioned in the article (just the numbered points) I don’t consider that a cheatsheet.

    Just so I don’t get accused of complaining without offering solutions, here is my opinion on how you could have improved the cheatsheet:
    -add links to the emv analyzer
    -add a blurb about how you can integrate these points with coschedule
    -show specific in the real world examples of each point

    That could have made the cheatsheet so much more valuable.

    I will be using the emv analyzer. that’s huge. And I definitely will be trying to figure out how to use pinterest, however some of the markets I am in, it’s hard. I always try to remember to go where the hungry crowd is.


  16. roger Williams Reply

    Great post, Garrett! Have you looked into which gender shares more? Also, it would be good to see the CTR in the body of the content. I’d be intersted in knowing what advertisers and publishers say the actual value of a share is worth.

    1. Garrett Moon Reply

      Our dataset didn’t have any gender information, but that would certainly be interesting.

      Regarding the value of a share, that is a really hard calculation. The demographics of a share are all over the board, so you have no way of validating or maintaining the quality the traffic. It’s a loosing effort, and probably not that important of a calculation. We measure site visits and lead conversions. Shares simply put more people at the top of that funnel. That’s always a good thing.

  17. Susan Lee Reply

    I’m so glad I took the time to read this blog this morning. It really shed some light on the types of words to use and the types of headlines to use in order to get shares. I will definitely be using the strategies you’ve outlined here. Thanks Noah + Garrett!

  18. Mike Bailey Reply

    Fantastic post and insights. Of course if the content is poor people wont share, but you have to get them to give you their eyeballs first and the headlines do this. Then your content has to deliver and you have to have something valuable at the backend to make it all worthwhile, don’t do the hard work and get it all shared if the traffic ends up a cul-de-sac!

  19. Marie Williams Reply

    Wow, this is really good stuff. I too am blown away by the Pinterest stats, I would never have thought it was so powerful. I was clearly wrong. I plan on experimenting with the headline analyzer too, thanks for a great post full of amazing insight and new ideas to try out.

  20. larry Reply

    Did you analyze the Volume, meaning total exposures/views? A killer headline on a small blog will naturally have less shares than an average headline in a big reach blog. There must be a tipping point where size and quality combine to make a bigger difference.

  21. Henry Brown Reply

    Thanks! The Cheat Sheet, the Headline Analyzer, and the introduction to CoSchedule were worth way more than the price I paid for this post. Oh, wait, this info was FREE! Noah, you’re amazing!

  22. Jeff Finley Reply

    Wow, this was really helpful! I tried the EMV analyzer on one of my posts “My Quantum Star Healing Experience” and it gave it an 80%! Really? It seems like it should have received a much lower score based on the advice given here. I’d love to use CoSchedule though it seems neat.

    1. Garrett Moon Reply

      Well, EMV is in-part a ratio of emotional words to non-emotional words. They actually imply right on the site that with shorter headlines, the score will be less accurate. I would assume that is what you are experiencing. “Healing Experience” is also highly emotional so that is influencing the score.

  23. Shea Reply

    Great post, Garrett, with lots of juicy tidbits, but those Pinterest numbers have got to be off. Way off. 90% of ALL shares across ALL types of post? No chance. I think you might be out by an order of magnitude! The latest ShareThis post had Pinterest’s share of shares at 9%. Facebook: 64%.


    I take what you’re saying about specific content doing well on Pinterest but that doesn’t seem to be reflected in your study (unless I’m missing something). But all content? Doesn’t even begin to compare to Facebook as a sharing platform. So something is amiss somewhere!

    1. Garrett Moon Reply

      Yep, 90% is what OUR data showed. Of course, there is some audience influence on that number, so it may not be true for every site.

      That said, our data was based on raw numbers from Pinterest itself. This means that we accounted for both Shares and Re-Pins, and I really think that made a huge impact on the numbers. The thing that surprised me most was the sheer volume of shares on Pinterest. It was massive. A “viral” post on Twitter/Facebook may received thousands of shares. Whereas a viral post on Pinterest easily received 10 or 20 times that in many cases.

      Anyhow, what can I say? Pinterest is huge.

      I am currently digging into our Pinterest data a bit more to see how bloggers can better use it for traffic. Hopefully there is an infographic in the works.

  24. Sherryl Anderson Reply

    I love, love, love this post. Copywriting has been an interest of mine for some time now and it brings into light a lot of great insight. My top two insights were the emotional value of a headline and targeting for your right audience. I just started reading Gary Vaynerchuk’s Jab, Jab, Jab, Right Hook: How to Tell Your Story in a Noisy Social World and he talks about adjusting your content to match the social media platform your posting on.

    When marketing just used to be about tv ads, and all sorts of media buying, marketers would adjust their content based on the medium. Why not continue that on social media?

    A lot of times it’s tough to come up with great topics and headlines and this just made it easier. Thanks for writing this!

  25. Eddy Baller Reply

    Love learning these details. It makes sense to target your audience, but I never thought about changing headlines to target specific social networks.

    What would be the best way to do this considering that the article will have one headline (on the site)?


  26. Joel Mackey Reply

    I know a lot of people attribute the rise of list posting to BuzzFeed but I really think that was a product of the olden days when Digg, Mixx, Reddit & StumbleUpon were the core drivers of social traffic. The Digg Effect often was driven by “list posts” and emotional headlines certainly won over in that area as well. Controversial headlines have always seem to win out as well. I love the data, it always helps to confirm the intuition that runs through the industry. For those that say “lists and emotional” headlines will die out, I call fooeey, as that stuff has been working since the dawn of marketing. 🙂 EMV calculator was a new one to me and I’m looking forward to running my own analysis using BuzzSumo’s tool to see if the averages stay consistent. Thanks for great content that constantly inspires me!

    1. Garrett Moon Reply

      Great point!

      It can be exceptionally difficult to do this, especially if you are using a long-tail keywords. I usually try to combine both the keyword and EMV score in the headline, but there are some instances where it is not possible. In those situations, I usually go with the more emotional headline (the better headline) and put the “SEO” headline in the meta title tag. These days, Google is more about matching up subject matter than keywords anyhow, so this is pretty reasonable method.

      Another option is to use the SEO headline for the post itself and then use the more emotional headline for social sharing/promotion.

      Great thoughts, thanks for reading 🙂

  27. Mike Reply

    Man what a great article! I was blown away by the Pinterest stats. I personally never use it and struggle with how to create great content for that medium. I’m not sure if my audience is on there, but it’s certainly worthy of more attention by my team.

    Thanks for aggregating all this data!

  28. Will Reply

    Couple thoughts:

    I wonder if the EMV score of an email subject line would have similar impact on open rates? I would assume so but not sure.

    I wonder also how effective EMV is for predicting whether a headline for an ad (whether print or online) would be effective (ie get the viewer interested enough to keep reading).

    I know these are similar to social media headlines, but also a bit different. Would be interesting to see any differences there may be.

    1. Garrett Moon Reply

      Both of these are great thoughts Will. I would love to hear about the results if you ever decided to try them out.

      My guess is that emotional copy would always be beneficial in some way.

  29. Fran Reply

    Many thanks to you both! I played with the EMV tool before I got to the bottom of the post, then I went back and read the entire article again. This was wonderful information!

    For you….Do You Want Great Ideas FREE? Be Smart, Subscribe Today! 80% (I just can’t seem to get past 80. Now I’m on a mission!)

  30. Dennis Reply

    I just started last week with Coschedule Garrett. Seeing this just converted me to a subscriber.

    I loved your whole research. The data will be part of my content for years to come.

    Oh, and that Google+ mention about being home-oriented? I was noticing it for the past couple weeks but I couldn’t prove it. You just saved me a ton of time.

    Now, if only I could get that mega trial upgrade or lifetime plan 😀

  31. JR Fent Reply

    This is one of the most interesting posts I’ve read in a long time. I’m totally fascinated by the stats regarding ‘shares’ to words used in a headline. I also wonder, how much the quality of the content in the post comes into play – or do people share in social media without judging the content. Good stuff Noah. I will owe you a taco.

    1. Garrett Moon Reply

      Thanks man, glad you enjoyed the data.

      I think content quality is still hugely important. Many people do share without reading (take a look at the Buffer Daily app) but not as many as you might think. At the end of the day, quality still wins – particularly because of SEO and Google’s drive towards quality search results.

  32. Steve Reply

    Great information. Some of this seems almost too good to be true. Quantifying the information and then figuring out a way to replicate it, test it, and put it to practical use is a brilliant idea.

  33. Shannon Reply

    It isn’t all that surprising that list posts are tops! Take a look at any magazine cover; most use the list headline to get us to grab it off the rack and buy. (They spend thousands on market research to make sure those headlines convert.) I suggest start watching the magazines your target market reads and emulate the headlines on the cover.

  34. Gary M Reply

    Wow, this is some seriously powerful research. Hadn’t heard of the EMV tool before and will definitely start using it for testing headline variants. CoSchedule seems like an awesome tool, so hopefully I get a chance to take it for a spin!

  35. Sharon Dubois Reply

    Personally, I shy away from list posts as I have usually found only fluff content. This is why I’ve been reluctant to use them in my business. I now see how I can use them, with valuable content, to gain traction and exposure. Thanks guys!

  36. Salman Reply

    This is pretty interesting! I definitely need to give pinterest a closer look but Facebook is my target at the moment. Email marketing is also on top of my list so these keywords are good to know. Personally, I know I tend to open list emails more than others but I’ve also noticed a lot of emails are list-based so I hope that doesn’t weaken their open rates over time.

    1. Garrett Moon Reply

      Thanks Salman! I would definitely get super serious about email. Facebook as an organic marketing tool seems to be declining rapidly. Email converts better, and as the marketer you hold more control over the outcome. Good luck!

  37. James Broderick Reply

    Okay, this is one of those posts that goes my Pocket and is looked at, shared, and even cited for being freakin’ awesome in the quality of content that it shares with its readers. If this post doesn’t get a hell of a lot of attention from those of us looking to make a serious impact with our own content, I really don’t know what will. Thank you. Seriously. Headlines have been told over to us — and over again — that they’re single-handedly the most important part of your blog post. Even if you do write epic content, if the post doesn’t grab and hold attention with EMOTION, why the hell would anyone click and share? Thanks again, man. Great research.

  38. Jay Ru Reply

    Hey, just wondering I hear a lot of people flocking over to instagram. What are your stats on that? Please do share. I wonder if instagram would be about the same as pinterest?

  39. Sid Reply

    So I’ve been doing it right all along, yay!
    Love the opt-in box for the downloadable checklist. I actually use Sumome on my site and I didn’t realise it had this feature too. Going to try it out right away.
    Is there a way to make the button look more like a feature box, or to just use a clickable image instead? Seems like it might get more attention that way.

  40. Jessica Prescott Reply

    Fascinating! Also, it looks like, in your first graphic, perhaps a typo: the green dot in the legend should, I think, say “1,000 shares” not “10,000 shares”…right? Cheers 🙂

  41. Nathan Reply

    About to launch my competition, couldn’t have really come at a better time. Now I can better create headlines that gets more shares, and since the competition is about getting shares, it’s pretty crucial to success.

    Thanks to this, my headlines should improve yet again.

    Thanks Noah and Garrett

  42. Ming Jong Tey Reply

    Hi Garrett,

    Thanks for this awesome analysis. I haven’t tried the Emotional Marketing Value Headline Analyzer and I think it is an awesome tool to help to craft out emotional headlines.

    “Imagine to double the no. of shares of your existing posts by simply tweaking the headline with that awesome free tool!” –> EMV = 28.57% lol…

    Time to start tweaking the existing headlines of my posts 🙂


  43. Natasha Reply

    So glad that I ran across this post! I’m especially thankful for the newfound knowledge of the EMV Headliner Analyzer offered by the AMI; thank you so much Garrett!

  44. Anne Reply

    I love Pinterest so this is great news! How did Instagram fare? I’m surprised that lists did so well, I see them everywhere so I would have thought people were sick of them. Great article, thanks for the share!

  45. Oeli Reply

    Three things why this article blows my mind away and will yours too!!

    1.) Tacoos will not affect your analytic skills – otherwise Mr. Kagan woudn´t do this great article
    2.) Pinterest is core of sharing because you pin it – but thought no one ever beats FB
    3.) “Emotion Free Techniques” can be dangerous since we have learned from this article that emotions matters;-)

  46. Paul Montreal Reply

    I’m often a little skeptical of any study that essentially covers all industries. The quicker you can find that same data for your specific niche the sooner you’ll stop making “peanut butter and chocolate chip cookie infographics” to try and make your steel manufacturing company go viral on pintrest. And I see people trying to do that all the time. But beyond that, I love the emotion-o-meter. I’m going to be playing with that all night. I was talking with some people about a Victoria’s Secret video yesterday. Almost perfect marketing! Hitting the viewer with every emotion possible. I think the common theme that and this is – Make em feel something. Anything. And hopefully in our content and not just the headline. Thanks for sharing your data, great work.

  47. alejandro Reply

    The emotional content of the headlines stills strikes me, is the promese of a engaging and deep entry that could benefit the reader in some emotional way! and that’s really useful

  48. Britt Malka Reply

    Excellent post!

    I’m not surprised by Pinterest’s share of the shares. It makes it very important to add pictures to every blog post, something I’m guilty of forgetting.

    The headline analyzer… I keep getting reminders for it, and I keep forgetting to use it. Thanks for another reminder and for showing how effective it is. Now I’ve added it to my start-up pages.

    I’m going to test how well it will work with emails.

  49. computer repair in schaumburg Reply

    Definitely believe that which you said. Your favorite reason seemed to be on the web the simplest thing
    to be aware of. I say to you, I certainly get irked
    while people think about worries that they plainly don’t know about.
    You managed to hit the nail upon the top and also defined out the whole thing
    without having side-effects , people can take a
    signal. Will probably be back to get more. Thanks

  50. Mike Rheaume Reply

    Hi Garrett – fantastic post! I found this pretty illuminating and useful as we hone our own content strategy. We will be tapping into this as reference for sure. Also curious to check out CoSchedule.

    Thanks again,

  51. Nate Reply

    Takeaway #6 was astounding! Pinterest seems to be neglected by the majority of Social media campaigns or is merely an after thought. Surely Pinterest demands more focus, to fully utilized this ‘untapped’ user base.

    Thanks for the post.

  52. Jasmine Reply

    Hi Garrett, you mentioned that list posts work best on Twitter, FB, and G+. What kinds of headlines work best on Pinterest?

    1. Garrett Moon Reply

      Hey Jasmine, great question. One hint on this can be seen in some of the most popular words from Pinterest headlines. We actually have a Pinterest marketing cheat sheet in the works over on the CoSchedule blog. Here is a sneak peak: https://www.evernote.com/shard/s2/sh/f6f1d1fe-2daf-4fbb-880f-64335638ffb8/556d960d50abb7c8d78879c4f34a789f

      Make sure you get on our email list to be notified when the post goes live. I think it could really help! http://coschedule.com/blog/

      1. Paul Chakursky Reply

        Thank you Garrett for the valuable info! Pinterest is apparently for posts that are intended to go viral

  53. Jason Byer Reply

    I think the Emotional Value Calculator is a fun idea but for social posts but might be limited in use when looking at the SEO benefit of post titles. However, “small cute fuzzy babys” does provide a score of 100!

  54. Karen Hickey Reply

    11 Ways to Improve Your Headline Writing for Better Results. I appreciated the insight and the info is valuable. Any insight into how this varies if you are writing for B2B? I would think that the Lists and Emotional aspect still applies (I know that “Top 7 Lists”, etc. do well from experience), but it would be interesting to see if there are other distinctions.

  55. Robin Reply

    Fabulous post, thank you! The EMV tool is a huge takeaway for me. I’ve always been taught that a compelling headline was the most important part of a post or subject line for an e-mail, but had no way to quantify it.

    To return the favor, to maximize your shares on Pinterest and drive traffic to your website here are 2 things you must do: Add word graphics to the photo you post that will then link viewers back to a specific post. To do that, once you pin your photo, a window will pop up that says “See it now”. Click that button (it disappears fast, so you have to do it immediately) and in the window, type in the URL of the specific post that the photo with word graphics relates to. Then when anyone clicks on the photo, they will get taken straight to that post. You can see an example of that here with my Pin that links to my post: How Much Does An Interior Designer Cost? http://pinterest.com/pin/85709199133650878/

  56. Julian R Reply

    I think your methodology has been a bit flawed here.

    Your assumption is that headlines are the dominate factor in social sharing, when in most cases for consumers it is the content that encourages a share.

    Headlines lead to clicks not necessarily shares (although there remains small but poisonous segment of views that share things without first reading/ viewing them)

  57. Ed Bisquera Reply

    That is just amazing that Pinterest garners the highest number of shares. Was there a breakdown as to the types of content and EMV of headlines for Pinterest shares?

    Amazing data and takeaways, thanks for the insight into the power of headlines; in many ways this can be applied to Email and Offline marketing too, as the headline is a key factor in getting interest from a target audience, IMHO.

    And thanks for the heads about the headline analyzer – great tool! 🙂

  58. Jane Reply

    I see the terms ‘chicken’ and ‘apple’ listed for LinkedIn on takeaway 3. Both those terms surprised me; can you help us to understand that better please?

    It would be nice to see some examples of the actual headline mentioned in the article. Even a smattering of examples would give some sense of actual headlines.

  59. Dave Linabury Reply

    Some really fascinating insights in this post. Pinterest taking 90%? Who knew? One thing to be wary of in making correlation posts is that you are basing data solely on headlines when other factors affect the sharing/virality of the post, such as author popularity, time of day, day of week posted, topic, media coverage, whether or not the post had an appealing image, etc. Not saying the headlines don’t work, but that there may be more to the data story than just the headline?

  60. ptwylie Reply

    Garrett great post you really nailed it! While subject lines, headings are a huge driver of email openings, blog post shares you need to have killer content. Lists are as you outlined are huge I wonder though if the size of list matters? Like “5 Best Ways You can get Laid” vs. “165 Best Ways you can get Laid”

  61. Robyn Rivers Reply

    This is really interesting info. We are constructing a blog and would love to hit the road running. Thanks for the insights and facts.

  62. Emmett Reply

    Cool article, but shouldn’t you use this advice in the title of this post?

    “We Analyzed Nearly 1 Million Headlines. Here’s What We Learned” could be something like “7 Free Ways to Make Your Blog Win the Most Shares”

  63. Sherryl Perry Reply

    I find this research very valuable. Having access to this database is a real asset to your company and I for one am grateful that you’re sharing this information with us.

    Last week, I featured your article on the EMV tool with my blog readers and I’m working on a new article now. Rest assured, I’ll be linking to this post. Thanks! I’m so glad I discovered your blog last week (and now I’ve discovered this site).

  64. Todd Bartlett Reply

    We all knew the headline is a key factor in sharing and optins but i have never seen till now it actually analyzed with a really good size data set. Some great take aways in here and thanks for the EVM tool. Might be a great tool to test what you are are going to say to your wife or husband when they are upset with you. 🙂

  65. scott brockamp Reply

    This is great information. In a world of social share to increase business, it is great to see solid data.

  66. gobr Reply

    Would be interesting to know from where these shares come from, I live in Brazil and don’t know if this will apply, will test in the future.

  67. Steve Pohlit Reply

    Thank you – very helpful information. My takeaway was from the content and the comments. I am reminded of the message everything changes.

  68. Leah Drayer Reply

    Thank you Garrett for the informative article!

    This will help us out so much when we plan out our blog schedule and start to write our new blog posts in 2015.

    All of the data you used to back up your points is very appreciated and I will defiantly use the EMV analyzer tool.

    Thank again!

    -Leah Drayer

  69. Lolly Spindler Reply

    Hey there dorks!
    I LOVE this piece of research.
    Quick copy editing catch: Under Takeaway #3, it reads “People really like content that helps they do something awesome…”
    I think you meant to say “…that helps them.”
    Lolly Spindler

  70. Harrison Alley Reply

    This is awesome. I wish they would combine these findings with OK Dork’s analysis of what makes for a viral article to create “the ultimate guide” to writing a viral post

  71. Uttoran Sen Reply

    Excellent research Garrett,

    Staying close to Jonmorrow and others in the copyblogger team, I can tell that your research is 100% accurate. List posts are well received on the social media. Almost any post that has gone viral due to our efforts were all list based posts.

    It is always about “What is in it for me?” and that is exactly an article starting with – “you” explains. That is how a headline connects with a potential audience. That is how you generate empathy with your reader. Which follow up in Social shares, comments and mentions from those that consume the content.

    Thank you for this report, am going to apply quite a lot of this in my future articles,

    Uttoran Sen,

  72. Adam Reply

    I’d be interested in how audience size affects the ultimate number of shares for a given post. I would assume that a larger Follower/Liker group will yield a greater number of shares for a well written headline. Is there a correlation between audience size and successful share volume?

    For instance, can we observe that a post with a 35% EMV will average 1,000 shares among an audience of 100,000? This would give us a 100:1 ratio at 35%. Could I translate that to an audience of 1,000 people using the same ratio, and know that if I write headlines with EMVs averaging 35%, I will average 10 shares from them?

  73. KL Ellis Reply

    You drove me to near madness with your 10 takeaways on EMV.
    Short and dramatic images seem to take the cake, and shares.
    As we all know, Playboy was not successful because of the articles,
    it was the ahem, images.
    So to capture someones attention we need to learn the emotion
    and apply it in a ‘Playboy’ way for our niche.

  74. Leonardo Wild Reply

    Hi, I have been testing the EMV analysis website and found it very helpful, but it can also be misleading: use a series of power words and voila, you can get a high score without making any sense. The pragmatics is missing, as in most translation machines. The question is, though: What will happen if more and more people use (overuse) these high-scoring EMV words and the networks get flooded with such language, making them less meaningful? Like films that require more action, more bullets, more special effects (until the story is lost) to entertain us … or catch our attention?

    1. Nathan Reply

      Great comparison Leonardo. I often think about what we are trying to achieve: catching the attention of someone and then… who cares? Or making someone think. So much knowledge these days is surface.

      1. Leonardo Wild Reply

        Thanks, Nathan. That’s why, in order not to leave something as pure surface—or as they say in the space industry: “All boosters no payload”—I try to add an insight that could prove useful beyond the time it takes to read what I wrote.

  75. Sri Reply

    EMV score tool is really awesome. I checked all the old content & there is direct correlation between the EMv score and shares. We figured out we need to rework on our headlines. We will make changes and see what is the impact.
    Thanks for awesome post.

    Interestingly EMV score for most of the buzzfeed posts is above 30 🙂

  76. Maggie Reply

    I really had no idea that Pinterest was so big! The catch however with this is how to get it to work in different markets and of course to get it to work well, not just as a presence. It would have been great to see what your research would uncover with YouTube taken into consideration.

  77. Will Reply

    I’m hoping somebody could tell me if the headline if AMI’s website is down for good or is it just for maintenance. I’m hooked on their analyzer and I have no way to contact them with the site being unavailable to me. – DESPARATE

  78. Liza Reply

    I enjoyed your article and found it very informative. You obviously did a lot of research to put this together.
    I’m not surprised about Pinterest seeing that it’s mostly visual content and a picture says a 1000 words! I loved how you applied your research to this article as well – writing it as a list, with imagery, and with giveaways! ha! Smart! A fine example of putting information to use! I’m going to apply what I’ve learned here to my work for sure! And – I’ll pass it along as well! Thanks!

  79. Debbie Reply

    Thanks so much for this! I work for an SEO company and our main clients are in the real estate niche. I made up a random headline and put it in the EMV analyzer and actually scored 40.77…if anyone wants to hire me to make up your headlines, give me a shout :))

  80. Jaime Moss Reply

    It goes to show that people are generally not only more social than we let on BUT also more emotional. The steel faced expression of yester-year has definitely been booted into touch. Great article !