Case Study: Upworthy’s Most Viral Video Of All-Time

Time to read: 4.5 minutes

In my post entitled “The Little Sentence That Changed My Life,” I discussed how reframing your emotions as “data” undergirded the editorial system during my tenure at Upworthy — one of the world’s fastest growing viral sites of all time. I stressed the combination of personal emotional data (aka, your gut) plus communal analytic data (aka, what’s working with your audience according to clicks and views) to optimize content for viral success. This is a follow-up case study to expound on how that looks and feels in the real world. (Psst — I’d love to hear your reactions to this case study in the comment section below.)


In 2013, Upworthy made a 22-minute video of a teen musician, Zach Sobiech, losing his battle to a rare form of cancer go viral. This video wasn’t supposed to go viral. It had all the “signs” stacked against it. It was 22 min long. There were no celebrities featured. It wasn’t funny; in fact, it was quite sobering. It had already been distributed by and, and only had about 500K pageviews between them. While 500K views is nothing to sneeze at, it was a far cry from being truly viral for sites with such massive distribution reach. It wasn’t timely or trending, being almost 2 weeks after the video had dropped on Fox and People.

Yet despite all of that, after it was packaged and distributed on Upworthy, the video got 15 Mil pageviews, added 300K subscribers to our email list in three days, and raised hundreds of thousands of dollars for osteosarcoma research. That’s not even the best part — the best part is that after we posted the video, Zach’s song, Clouds, skyrocketed to the #1 position on iTunes, the first time in history that happened for an independent artist.

So how’d we do it?


Upworthy is known for testing multiple variations of content packaging (headline + image + supporting text) for each piece of content we curate and post. And this one was no exception. We tested upwards of 110+ options for this post.*

This was an example of an original control package (the way it looked on other sites):Zach package 1


With Months Left To Live, 17-Year-Old Zach Sobiech Says Goodbye Through Songs (VIDEO)

Which — let’s be honest — really isn’t that exciting.







Upworthy started its testing with this package:

Zach package 2


This Kid Just Died. What He Left Behind Is Amazing.

Which gained a 3% increase in clicks.









We then tested this package:

Zach package 3


This Kid Just Died. What He Left Behind Is Wonderful.

Which gained a 27% increase in clicks.









After multiple variations of thumbnails and headlines, we settled on this package:

Zach package 4


This Kid Just Died. What He Left Behind Is Wondtacular.

Which gained a 69% increase in clicks and became the winning package!!!








*In a following post, I will outline the principles and techniques of divergent thinking which were used to generate these headline options.

A funny kind of meta thing happens when media goes viral: the media starts doing media stories about the media that went viral. When that happened with Zach’s documentary, we told the story with numbers. We emphasized our data-driven culture and our proprietary testing tools. We reported the hundreds of thousands of dollars donated.

But there was more going on than just the numbers in this case study. There was a story behind the story — one that never got told to the press. That’s the story I want to tell you now.


Adam Mordecai is the writer on staff who found and packaged the piece. And (surprise, surprise) Adam’s own father had passed away from cancer a few years prior to this event. Here are Adam’s own words about his experience with this piece.

Me: Adam, how did you find the video?

Adam: “I had written up Zach’s previous video, and it got a whopping 10k views. A fan wrote me to let me know that Zach had passed so I Googled more videos of his and fell upon SoulPancake’s documentary. Then I hit play. Then I started crying and didn’t stop after it was over and started writing headlines and having flashbacks to my dad dying from pancreatic cancer and what one goes through when that happens. I wasn’t sure if I was biased but the whole thing seemed timely and wonderful.”

Me: Answer honestly — did you sense you had a mega hit before publishing it?

Adam: “On the second viewing I was still crying. So I had a sense it would be a hit.  When I started testing it, and it broke 10k views I started getting giddy. Spent the night having dad flashbacks while people tweeted at me about the loved ones they lost. Cried a bunch more. When it was at 40k, I almost crapped my pants. And stopped crying long enough to realize shit was getting crazy.”

Adam Mordecai spent 17 hours straight testing this piece of content and Tweeting at fans who were picking it up. SEVENTEEN HOURS. Notice how the emotional decision-making and analytic decision-making was all jumbled together, playing off each other, providing the creative tension I’ve seen replicated in many viral success stories.

If you remember the assertion from my previous post that society places a higher value on logical thinking than emotional thinking, it’s not hard to see why Adam’s personal story isn’t the one that got reported in the news.

As the Managing Editor of a respected media site, I couldn’t very well answer the question “To what do you attribute to this success?” with something like, “Well, the correlation between tears cried and pageviews seems particularly acute in this case study.”

Or could I?

After years of watching things go viral, I am convinced that there is more going on here than just numbers. Certainly, our testing tools had a lot to do with the video’s success…but that was no larger part of the story than Adam’s emotionally driven decision to spend 17 hours testing this one piece of content. He just wouldn’t let it go. In three years, I have scarcely seen a piece go viral without full consideration of the range of human emotion it can produce, which is nothing a machine can measure. If I had to guess, I’d say both the analytic components and the emotional components equalled a full 50% of the success of this piece.


My biggest take-a-way from this case study is that successful social media stories don’t start with the keyboard. They start with the heart.

We learn through this story that social media is about the social — not just the media. What Fox and People did wrong with this piece was they stopped short at reporting the information. Upworthy didn’t just tell you what happened, we told you why it mattered.

Of course technology can be used to support human objectives. Machine learning and data knows better than any human whether “amazing” or “wonderful” or “wondtacular” will be clickier — not a subjective editor. This is the kind of information that can only be gained by testing and observing in real time in an organic environment. And thank goodness for headline testing!

But there’s another kind of information that can only be gained by being human, by being self-aware, by tapping into your emotions. Sure the testing tools knew that “wondtacular” clicked best, but the computer can’t understand the context of why. Only people can know what it’s like to hold the hand of a loved one in a cold hospital room as his life slips ever so slowly away from him… Adam knew what his father’s face looked like, what the hospital smelled like, what that heartbreak feels like. And he was able to use that advanced ability for emotional synthesis to transform it into a compelling frame for others who can empathize with that feeling too. It’s easy to see how accurate it is to say something beautiful left behind out of all that sorrow is nothing short of “wondtacular.” And that connection, that simple human connection which happens billions of times each day across the globe, is stronger than any data, rational logic or test will ever be. It’s the kind of strength that produces hundreds of thousands of shares and puts a small documentary into the stratosphere of rock stars.

‘Til next time.
– S


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: