I want to share a tool I love because it helps me write better headlines—the CoSchedule Headline Analyzer.
CoSchedule uses data compiled from millions of headlines on the web to look at three factors:
1. Headline types that get people to pay attention. These are things like lists, how-tos, and questions.
2. Word types that resonate with audiences. CoSchedule splits them up into common, uncommon, emotional, and power words.
3. And character length, the number of characters in your headline. Length can matter if you’re trying to get noticed on social media or in search engines.
The CoSchedule magic happens when you type your headline into the box and look at what it tells you.
Do this for enough headlines, and you’ll of headline “grades,” scored out of 100, so you can compare each one you’ve done and see which is “best.”
Now I keep using air-quotes there because the scores are somewhat arbitrary. Is one headline really a 77 while another is a 56? I don’t know.
But if you treat them as relative numbers—f you say “CoSchedule’s data seems to suggest that headline A is going to perform better than headline B”—then you can get a lot of value out of the tool.
I like to use it to challenge myself to keep writing new headlines.
“Is this the best I can do, or am I settling for something? Well, my headline-only scored a 65, so what if I tweaked this and this? Ahhhh, there it is.” I don’t stop until I can get at least a green light in the 70s. (For the record, I have NEVER seen an 80, so I don’t even know if it’s possible. )
The big test, of course, is how your headlines perform in the real world—if a CoSchedule 58 outperforms a 72 for you, then you’d be dumb not to use the 58.
But all things being equal, try to push yourself to get higher and higher scores. The CoSchedule Headline Analyzer might just help you break through what you thought was possible—and become an incredible headline writer.
Once caveat though—the major limitation of the tool is its bias toward short, scrappy headlines.
In other words, the CoSchedule Headline Analyzer will NOT help you with the Craig Ballantyne-style headlines I showed in my last post.
To figure out whether those are any good, you’ll have to put them through a different set of tests, which I’ll talk about next time.