Lillian Podlog

A Look at How Three Brands Are Driving Social Engagement

By: Lillian Podlog | September 19, 2016 | 

What We Can Learn About Driving Social Engagement

Earlier this year, Fractl partnered with BuzzSumo to look at one million of the most shared articles between December 2015 and May 2016.

Research revealed that Facebook is capturing more than 90 percent of all social shares, up from 82 percent in 2014.

While Facebook has been monopolizing social engagement, we’ve seen a parallel increase in the number of websites with content that’s receiving a high level of social engagement.

As Facebook continues to alter its News Feed algorithm to discourage clickbait and encourage greater publisher diversity, we are likely to see fewer social media juggernauts and a wider pool of sites that are getting thousands of content-based interactions.

Now is as good a time as any to dust off your social media strategy and see if you can increase social engagement on Facebook.

To help you do so, we looked at how Upworthy and National Geographic consistently create shareworthy content.

We also took a look at NowThis, which doesn’t appear on our list of top sites, but has been racking up Facebook engagement through a different method.

What We Can Learn About Driving Social Engagement


Across 1,085 articles, the average share count was 15,263.

In 2014, Upworthy was the number seven site for the total number of shares.

In 2016, however, Upworthy did not even crack the top 20.

Although its total shares have dropped since 2014, Upworthy still has the highest average share count across all of the analyzed articles.

What accounts for their success to this point, then, and what lessons can we learn from its decline?

Focus On Positive Emotions

Upworthy is extremely effective at creating emotional content.

For instance, their most-shared article of the past six months was “Dad and Daughter Relationships As Explained by Ten Paintings.”

The article features the heartwarming artwork of Ukrainian painter, Soosh.

Upworthy doesn’t shy away from covering difficult topics, either, but always focuses on the hopeful aspects of each story.

This article, “102 Days After His Wife’s Death, Patton Oswalt Describes Grief As Only He Can” addresses the loss of a loved one, but is included in a section titled “Inspiration” (which is consistent with the article’s tone).

Of their top 100 articles between December 2015 and May 2016, every single article had a positive sentiment.

Positive emotions are prime content for virality.

What We Can Learn About Driving Social Engagement

Although their model of curating emotionally powerful content coupled with curiosity-gap headlines worked in 2014, it isn’t a lasting model.

Between the Facebook algorithm changes and the internet’s clickbait fatigue, Upworthy experienced a sharp drop in organic traffic last year.

What We Can Learn About Driving Social EngagementEarlier this year, Upworthy reorganized to focus more of the company’s energy on original video content.

Although its traffic has been increasing since January, the jury is out on whether or not the new strategy will be effective.

Be Careful of Data Dependency

This second point might not be a popular one, but one of the main lessons we can learn from the difficulties Upworthy had in the past year is to not rely too heavily on data.

The site’s signature headlines were the result of extensive testing.

Even though Upworthy has purportedly moved away from clickbait headlines, the company will have a hard time disassociating itself from content that over-promises, but under-delivers.

Headline testing proved that curiosity gap–style headlines were effective at attracting readers.

However, data didn’t predict that these types of headlines would become widely viewed as “insufferable” and “manipulative.”

Data only shows us what we decide to test.

Additionally, human intuition can be better at spotting the difference between an effective tactic and a gimmick, rather than an algorithm.

Although Upworthy may be able to regain the internet’s trust, it will be an uphill battle.

National Geographic

Across 2,884 articles, the average share count was 10,229.

National Geographic was popular before the internet, and although its initial progress may have been slow and its business model is still adapting, the organization has been highly effective at gaining social traction with its content.

The publication has the second-highest average share count of our sample—our analysis looked at social engagement across Facebook, Twitter, LinkedIn, Pinterest, and Google+.

A study by Shareablee—which looks at social engagement across Facebook, Twitter, Instagram, and YouTube—ranked National Geographic as the top social brand in the United States.

The company is popular on Snapchat as well, and its overall organic traffic has been steadily increasing since 2010.

What We Can Learn About Driving Social Engagement

Create High-Quality Content and Adapt to New Platforms

National Geographic has been doing so well because it creates great content—and has always created great content.

Its success throughout the years is a reminder of the durability of quality.

In particular, the publication is known for its photographs, which spread easily across different social platforms.

However, great content goes nowhere without effective promotions, and National Geographic has been careful to adapt to changing technologies and audience expectations.

For instance, it curates platform-specific experiences for its followers, and the brand is not afraid to gamble on new technologies.

Feed Your Audience’s Curiosity

National Geographic and Upworthy both know the power of audience curiosity (although, admittedly, they harness it in different ways).

Although some of its most shared pieces of content are photographs, this informational page about pufferfish has been shared more than 155,000 times.

Another top National Geographic post (with more than 90,400 shares) lets readers scroll through data about different types of sharks.

Part of their brand is that it develops trustworthy, scientific content that’s also accessible to the average person.

The publication’s headlines, therefore, promise an exploration of a certain topic (e.g., sharks) in a way anyone can understand.


NowThis doesn’t appear on our list as a top shared site because it distributes 100 percent of its content across social media and publishing platforms, but it has been making waves.

In fact, it was the most watched news publisher on Facebook from June 2015 to June 2016.

In July 2016 alone, the publishers’ content generated more than one billion views on Facebook.

Although their 100 percent distributed content model is not for every publisher or every brand (in fact, monetization has been difficult for the network), we can still learn from the company’s social engagement strategies.

Develop a Robust Distribution Network and Tailor Your Content to Each Channel

Athan Stephanopoulos, the SVP of  strategy and partnerships for NowThis, said:

We don’t create one video and then cut that video down to 15 seconds so that it can run on Instagram, and cut it down to six seconds so it can run on Vine. We will actually tell the story in a different way that is visually conducive to the platform.

NowThis, like any publisher that is effective on social media, doesn’t just cross-post its Instagram videos onto Facebook.

The company would rather tell the same story in completely different ways for each network.

Like BuzzFeed, the NowThis distribution network is not limited to social platforms.

In fact, NowThis publishes its content on BuzzFeed, AOL, Yahoo, and MSN.

It’s not enough to create a robust distribution network for your content, it’s also important to optimize for every social platform and distribution partner.

Reduce Friction

After NowThis had stopped publishing content on its homepage in February 2015, the audience grew 1,500 percent!

Rather than using social media to drive users to content, NowThis makes social media the final destination.

Although traditional brands and publishers do want users to ultimately visit their site, it is important to post both social media content that directs users to your website, as well as, content that a user can consume directly on the platform.

NowThis is the king in frictionless content consumption, and we can learn from its example.

Learn Social Engagement From Others and Listen to Your Audience

Every publication has a unique social media footprint, and the tactics that work for one audience may fall flat with another.

While it is important to pick up tricks from socially successful content creators, the key is to listen to your audience’s interests and preferences.

image credit: shutterstock. graph credits: Lillian Podlog and Fractl

About Lillian Podlog

Lillian Podlog is a Junior Creative Strategist at Fractl, a content marketing agency that specializes in the science behind viral content. She is also the proud (and protective) mother of two guinea pigs.

  • Corina Manea

    Wow! 90% of social shares captured by Facebook. It’s crazy!

    I love how National Geographic creates content, as well as their social media strategy.

    Thank you for a great post, Lillian. Good job on the research.

    • Hi Corina,

      I’m glad you enjoyed the post. 🙂

      Facebook is definitely doing something right! (From their perspective at least…)

      And I love the content that NatGeo puts out; I’m happy they’ve managed to be so effective across multiple platforms, and they aren’t afraid to get their hands dirty with new technology.

    • Not surprised at all. I would of guessed it before reading this. And based on user numbers as a share/user Twitter I bet is better. They just lack the user volume.

  • Thanks for sharing this! Good info to take back to my own drawing board.

    • I’m happy you enjoyed the post.

      Putting new information into effect is always the most challenging and exciting part. 🙂

  • I just want to bring levity to your data. You are down on the 3rd floor vs the Corner Office on the top floor.

    Getting 10,000 shares for an article is nice. But you leave out how many reads an article is getting. Shares mean nothing if someone doesn’t see it and then take action (read, reshare etc).

    Your numbers are global like the web. So having 10k shares of a NatGeo article from a pool of potential 1bil + readers do the math. Why are shares so low? If an article had just 1 million reads then only 1% shared it. Why so low?

    You also leave out another side of this. Which is you omit total reads, website visits etc. My take is based on how long social media and open graph has been here it has been a big failure and underperformed because it has been hyped as “We all want to share everything all the time’.

    I also would want to know how many articles were published? Is the top 1 million 50% or only 3% and if only 3% why did so many fail to be shared at volume?

    Lastly based on the networks if you were to ask me blindly what platform is getting the most article shares I would say: Well Facebook then Twitter because they are basically the only platforms people share news on.

    I do research on Facebook engagement. I was actually blown away at the news brand pages. In the last 4 years the average Fan of a Facebook page has seen engagement drop from 1x per year to on average now only 1x per 2 years. Meaning just to LIKE one item if you call that engagement (I do not btw) for major brands the pages are dead. Currently the average Fan of the NY Times engages (if including article LIKES) 1x per 69 days. Vs Coca Cola which is at 1x per 14,391 days. (running 7 day number btw)

    So I could say FB is successful for news sharing, but dead for most non-news major brands. But in the big picture…if I only engage on FB with the NY Times 1x per 69 days is that good? Not sure it is. All those article posts and just to get a measly LIKE it takes 69 days on Average?

    Just wanted to get you the 10000 ft view of things.

    • So I went to Nat Geo. Most article posts get less than 1k in shares and the engagement rate currently is 1x per 316 days per average fan. My studies have shown that 80% of social media activity is generated by 20% of the users.

      Anyway to determine what % of NatGeo’s fans are actually active? Or figure out uniques? Meaning if you chose 5000 articles and studied the shares how many individuals were involved? That would be great data. Do you have 10mil participating or just 20,000 out of 42 million.

      • Hi Howie,

        Thanks for the thoughtful reply. I definitely agree that we also need to think about % engagement. That is a more difficult question. The cost of liking a page on Facebook is fairly low. If a user doesn’t engage with content on that particular page (or similar content), then that page’s updates will simply not show up for the user.

        You could definitely do research on % active fans on Facebook, but it would be more difficult to obtain that information. Rather than relying on raw engagement counts, you would have to access and record the list of users that like particular updates and compare with other posts.

        But then we have yet another issue, we know that Facebook does not show every user every update. We would have to figure out how many people saw an update and choose not to engage on Facebook.

        The top-level view is difficult to access sometimes. 🙂