Many people know Upworthy as the originators of click-worthy headlines, some might say clickbait even.
And sure, headlines are a big part of what Upworthy does, but in a very strategic sense. I admire them for thinking outside the content box in a lot of ways, challenging themselves to write 25 headlines per blog post, coming up with cool SlideShares like these:
I had the pleasure of coming across even more of their cool tactics in Sean Ellis and Morgan Brown’s Startup Growth Engines.
Here are some snippets:
While editor-at-large Adam Mordecai cites their very low success rate in making things go massively viral as proof that virality is unpredictable, this flow chart produced by the company displays a level of growth-oriented thinking that isn’t common among other media outlets.
Upworthy understands what it takes to go viral, and then goes out to test and find the best combination of elements to drive that virality. Their primary signals are both shares per view and clicks per share.
To calculate shares per view, take the total number of social media shares and divide by unique visits to the content. Do the same for clicks/share, replacing uniques with clicks.
Facebook A/B tests for headlines
Here’s how they do it: First, they pick two promising headlines for the same content and create a bit.ly url for each—one with url?r=A and one with B. Next, they find two cities with similar demographics and populations amongst their Facebook fans and share one bit.ly with each city. They set a timer and wait for the clicks to roll in. When the time is up, they add a “+” to the end of the bit.ly and compare stats. The title with the most clicks is the winner. And they don’t just test clicks—they also compare shares per view to see which headlines results in the most reshares.