We all want our headlines to get clicked, and we all want them to get shared. Here is the thing, it turns out that what readers share, and what they click, may be two different things.
What are those differences, and how can we reconcile them? Let’s find out by looking at some empirically tested results.
What Gets Clicked
Don’t Be Pushy Or Tricky, Be Clear And Emphasize "The Human"
In one major study, Outbrain and Hubspot teamed up to analyze more than 3.3 million paid headlines across over 100,000 publishers. They looked at several variables to see what impact they had on both click-through rates and post-click engagement.
What hurt clicks? The answers might surprise you:
- Mentioning the viewer.
- Using words like “best”.
- Words that imply some kind of shortcut, like “easy” or “trick”.
- Implying urgency, e.g. “now”.
It is funny, as these are the words marketers thought helped with views and created a dying urge in the reader to click, but reality strikes as - “Avoid them”.
- “Who,” not “why”.
- Bracketed elaboration, like [Study] or [Video].
And really, that is about it when it came to the factors that they looked at.
The overall lesson here?
It looks like we should be emphasizing the human factor, making it clear what readers can expect to find on the other side of the click, avoid telling users who they are and how they should feel, and not promising the cure for cancer once they get there.
The results were very similar for post-click engagement. Words that demanded urgency or made strong promises seemed to put readers on guard, making them less likely to click-through to other pages on the site or spend time on the article after the click. The one exception was the word “amazing,” the one and only superlative that seemed to improve post-click engagement, despite reducing click-through rates.
Chicago Tribune’s Kurt Gessler recently shared 13 before-after examples of headlines that emphasize how important it is to be clear in headlines. Each of the examples he shared resulted in at least a doubling of the click-through rate. Here are a few examples:
- “Chicago’s bridges are falling down,” changed to “Some of the Chicago area’s most-traveled bridges are ‘structurally deficient’, Study Says”
- “Frigid temps no deterrent to braver Bears fans,” changed to “High of zero. Wind chill of 20 below. Sunday matchup could be the coldest Bears game ever in Chicago.”
- “Lawmakers question pricing of Northbrook company’s $89,000 muscular dystrophy drug” changed to “Northbrook company’s $89,000 muscular dystrophy drug is sold for $1,000 in other countries”
That last example really hammers home just how important context is, given that the headlines were already very detailed and clear.
Tell, Don’t Ask
A survey by Engaging News Project showed headlines to 2,000 participants. They paired the fake headlines with well-known brands and asked the participants how they felt about the headlines, whether they would read them, etc.
They created three types of headlines:
- Traditional: These are designed to be short and clear, with an air of certainty.
- Forward-reference headlines: These have a more “click-bait” slant, introducing more uncertainty, e.g. “How [X] Are Using [Y] To [Z].”
- Question headlines: These are just what they sound like, asking a question in the headline, and allegedly answering it in the post.
First and foremost, they found that people responded more negatively to question headlines than traditional headlines. Forward-reference headlines performed the same as traditional ones. Overall, the participants felt worse about the question headlines and expected the content on the other side of the click to be less useful.
Second, the study found that topic mattered. While the question headlines never performed better than the traditional or forward-referencing headlines, they performed worse for some topics than others.
Finally, they found that brand was heavily important. Brands perceived as politically neutral performed better with question headlines than brands perceived as politically biased. This tells us something important about branding, but it doesn’t change the overall conclusion: you should probably avoid question headlines.
Clarity Goes For Pop-Ups Too
Sumo looked at over 150,000 opt-ins on some of their pop-up forms and found this: straightforward headlines got more clicks and conversions than “creative” headlines.
In their post on the topic, Sean Bestor claims that you should be creative with your headlines when it comes to articles, but that the results say something very different for pop-ups. It all depends on what you mean by creative, but looking at the headlines they tested at Sumo, the results are very much in line with the studies mentioned above.
Two of the four tests they ran were straightforward headlines run against question headlines. If you were paying attention to the earlier studies, it shouldn’t be surprising that the straightforward headlines worked better.
Here were the other two:
- “Don’t let your old articles die” vs. “The blueprint behind republishing content”
- “We’ve never given this away” vs. “Get our internal website optimization spreadsheet”
As you might have guessed, the second headline in each case did better.
All of this fits together robustly. Readers want a clear promise of what to expect on the other side of a click. Headlines designed to pique their curiosity by obfuscating information don’t work. The one exception we have seen is the “forward-referencing” headline. But while those obfuscate some information, they are very clear about things like who is involved, what the effect is, and so on. They just avoid giving away the punchline.
What Gets Shared
From Clarity To Gossip
Buzzsumo ran a massive study that looked at 100 million headlines to see which ones got shared the most on Facebook and Twitter.
Here is where things start to get interesting, because what gets shared can be in direct conflict with what seems to get clicked on, based on the studies mentioned above.
Twitter and Facebook also had surprisingly different results in some cases. Let's look at each of them.
Which Headlines Do Well on Facebook
The most shared phrase on Facebook had an average of more than twice as many shares as the second most shared phrase (which, for the curious, was "this is why").
The top phrase? “…will make you…”
In other words, we’re talking about headlines that look like “24 Pictures That Will Make You Feel Better About The World” or “13 Travel Tips That Will Make You Feel Smart.”
On Facebook at least, people seem to be looking for headlines that promise a personal impact, at least when it comes to what they will share.
It probably isn’t surprising to anybody that list posts did tend to do well. The number “10” in particular outperformed the others.
Emotional phrases also consistently performed better than non-emotional ones on Facebook. Examples include “tears of joy,” “is too cute,” and “give you goosebumps.”
Phrases associated with gossip also seemed to do well, which isn’t necessarily surprising for social platforms. Here we are talking about phrases like “twitter reacts to,” “are freaking out,” and “what happened next.”
Relatively long headlines seemed to do best, centering somewhere around 15 words. This, at least, seems to be in agreement with what works for click-through rates, since it is associated with clarity and specificity.
Which Headlines Do Well on Twitter
Phrases on Twitter were surprisingly different. While “…will make you…” performed fairly well, “this is what” was the top phrase. This may reflect a larger trend on the network because emotional phrases overall did not do particularly well while emphasizing newness was much more important.
A study by CoSchedule published at OkDork gave similar conclusions. List posts, posts that mentioned the reader, and posts that made promises to readers seemed to do better than others. The study also found that Pinterest and Facebook completely dominated re-shares.
What should we do with this information?
Since it seems to be in direct conflict with what gets clicked-through, the actionable advice seems to be a toss-up since it is hard to tell whether the increased sharing will outweigh the reduced clicks.
Well, there are a few things to keep in mind.
First off, since these studies focused on the most used phrases, rather than testing different versions of the same headline, it could be that the results lead us astray. After all, it very well may be the case that these phrases are just commonly used by the most re-shared brands, since “everybody knows” that is how you are “supposed” to write headlines, rather than that the phrases themselves are actually encouraging shares. For all we know, these headlines would have been shared more often if the phrases were dropped.
Second, this emphasizes the importance of platform-specificity. Clarity and directness seem to be the best approach for your title tags and on-site headlines, but it is worth experimenting with other formats when you post on social platforms.
(Well…it’s important to experiment on all platforms, including your own, but you get what I mean.)
Where Do I Stand?
My experience points at a stance somewhere in the middle, but I lean towards clarity over obfuscation.
In particular, those “forward-referencing” headlines seem to get the job done best, balancing curiosity and creativity with clarity. In general, I would follow the advice against words that trigger skepticism like “trick” and “easy,” or that push the reader with urgency.
On the other hand, headlines that promise something specific for the reader, like accomplishing a task in a specific number of minutes, seem to leverage those social impulses without compromising the clickability and trustworthiness.
Does it sound trite for me to finish this off by saying that you need to experiment to find out what works best for you? I hope not because the reality is that every market is different, each audience has different preferences and, crucially, different levels of skepticism.
Still, it helps to understand the differences between what gets clicked and what gets shared as you build the strategy that best suits your brand.