Want Social Shares? Then A/B Test Your Headlines
Over the years, I’ve hammered home the importance of using data to drive your content marketing, as opposed to just writing and distributing based off of gut feelings or vanity metrics. Well, we now have some exciting new research that shows just how powerful multivariate testing can be when producing a story.
Double your Sharing
According to new data from Priceonomics, simply A/B testing an article’s headline can almost double how frequently it gets shared on social media.
There’s even more good news. A/B testing headlines leads to:
- 34% more pageviews (that’s the mean improvement; 28% more if measured by median.)
- Up to 587% more social shares (or 75% more, if looking at median and not mean.)
- As much as a 6% increase in press mention. (Again, that’s the median; it’s -2.7% by mean, but I suspect that last metric might be within the margin of error for the experiment on the whole.)
Those are pretty astounding results, and are even stronger than the improvements usually earned through A/B testing an email. With emails, you’re lucky to see a 5% improvement in open rate, with commiserate changes to clicks at best. Here, the authors did something as simple as writing a second headline and letting a test run for a few hours, and got wildly improved results for their content.
Multivariate Testing – Not Just for Email
Too many people think that A/B testing is just for email marketing. But the data is conclusive – an analytical approach to your content marketing will dramatically improve performance.
Think about how much time your team spent on writing a single article. Brainstorming, sourcing quotes, finding images, writing it, and more… That’s not just hours, but real money too.
So, while it might take a tiny bit more work to test versions of your content before sending it to the world, it’s a fraction of what’s already been done. And it can make all the difference between a story that flops and content that goes viral.
How to Best Adopt This Strategy
Like any new finding in marketing, this technique can be used for good or for evil. Some might be tempted to test a handful of different clickbait-style titles for any given article.
But, it could also save you (and your writing team) from the embarrassment of dozens of similarly structured, vapid blog post titles.
Let’s say you have a new article out, discussing five new trends within your field and what makes one of those even more influential than the rest.
You could call it “Five new trends shaking up marketing. You won’t believe number four!” Or maybe “Your competitors all adopted this new growth strategy. What happened next will you make you cry.”
Those would be the pure clickbait options. And maybe they’d get you a few more click-throughs on Facebook. But would those same viewers want to share the article? Or would they be upset when they actually read your dry analysis and find that no, they weren’t brought to tears.
Alternatively, you could give it a to-the-point title: “Businesses drive growth through new techniques, including mobile-first” or maybe “how the shift to mobile is driving sales.” Maybe fewer people would click, but they’d be much more satisfied when they did, as the story matched the article.
No matter which of these example titles you choose, you’d still just be guessing. You didn’t guess when you wrote the article, you did actual research and spent the time to get it write. So why guess about the title, when just a little bit more work, and running a real test, can get you such stronger results?
Don’t fear data, embrace it!