SECTION 1: Why Am I Writing About How To Write Killer Headlines Using Numbers?
Numbers are absolutely everywhere: ads, tv shows, the lottery and every other article. You can’t seem to escape them. Numbers are also an essential part of the media industry. They’re a bit like brain candy – something to snack on and remember for later.
Have you wondered why some articles are shared more? Why some headlines seem to be more clickable than others? Well I have and I also want to know why.
In this article I’ll be focusing on a single element of virality – numbers in headlines. Furthermore, I’ll be going even deeper by only researching whether odd/uneven numbers make the article more attractive to visitors. Not in terms of clickability, but from a “I’d share this” standpoint.
I’ve been fascinated by numbers all my life. Even though I spend most of my day writing and playing with words, there has always been a side of me interested in mathematics, the relationship between numbers and percentages. So this article will both satisfy my desire to delve deeper into my past and (hopefully) provide useful information to make your titles more attractive and clickable.
SECTION 2: Methodology and Tools
I care deeply about my time. I also worry about not wasting yours. So I’m combining my own original research with studies that have already analyzed hundreds of thousands of articles.
I apologize in advance for my errors in judgement – you’ll see comparisons between different types of websites and an analysis over unequal time periods. But worse of all, the foundation of this study is built upon the relationship between 2 elements:
- Article Headline
- Article Virality/Engagement
Virality, in this case, is defined as the number of shares (facebook shares, twitter tweets, linkedin shares, google +1s) for one single article.
Engagement is measured as the number of comments (positive, negative or neutral) for one single article.
Extracting comment numbers was done using Import.io. NOTE: This section will be updated soon.
The problem with this relationship is that people don’t share a headline. They share the entire article. They also rarely comment on headlines alone. I would’ve loved to have measured titles vs click through rates or headlines vs pageviews (like this post Paul Shapiro wrote on Noah Kagan’s blog).
SECTION 3: How Did I Choose the Websites? + See My First Results
Let’s call this article Version 0.7. It’s here, it’s organized, but it’s a mere fragment of what it might become in the future. I’ve started with a few websites I frequently see in my Facebook Newsfeed. It’s not the most scientific of reasons, but you’ve got to start somewhere, right?
I’ve chosen 5 websites to begin with and I’ll expand the list based on business categories, relevance to the study and your suggestions (feel free to leave a comment at the end of this article).
Here are the 5 initial websites included in my study:
- Canva Blog – Design School – DESIGN
- The Verge – TECHNOLOGY
- Viralnova – VIRAL
- Quicksprout Blog – MARKETING
- Huffington Post – NEWS
There was quite a bit of debate around what websites to choose. For example, I had Upworthy, Buzzfeed and Viralnova in the “Viral” category. But then Upworthy started creating its own content and explained the concept to newbies in a Slideshare presentation, so it doesn’t really feel viral-y right now. Buzzfeed is close, but thanks to its sleek content created on the website and YouTube channel, it mostly feels like teenage entertainment, so I cut it from this category just to be on the safe side.
So, without further introductions, let’s start with the Canva Blog.
According to the results from the Canva Blog:
- If you’re thinking about adding a number in your headline – you should.
- Your number in the headline should be divisible by 5.
- You can list a lot of items, if you can find them. Designers tend to spend more time analyzing, browsing, to find inspiration.
Let’s move on to our next website – The Verge.
This is where I felt the study wasn’t meant for all types of websites. If the Canva Blog content creators could name their articles whatever they wanted, The Verge writers were at a disadvantage: How could you review the Apple Watch, without having the words “Apple Watch” in the title?
Most of the first 165 headlines had even numbers in them. Moreso, numbers bigger than 4. Of course, numbers have a different meaning in the technology world. 50 magazine cover designs seems like a lot. But a 50 GB SSD? It’s barely big enough to install Windows on it. Not only that, but a 50 GB SSD is not noteworthy, in 2015.
I felt bad for not really being able to properly analyze their headlines, but such is life. Perhaps we’ll have more luck with the next website – Viralnova.
Because 165 just didn’t seem that big of an article sample for such a large site, I upped my game. It’s go big or go home. I analyzed 10X as many – 1.650 articles. And since some tasks weren’t automated, I did spend a few hours copying, pasting, comparing and cursing at the limitations Excel has.
According to the results from Viralnova:
- Headlines with numbers are shared more, only by a slight margin.
- Even numbers did fantastic – 34 for some reason was a clear winner.
- You should test the whole spectrum of numbers in your headlines – surprises might hide in plain sight.
Moving forward to our next website – The Huffington Post.
I’m not really sure what to think of these results – on one hand it proves that Huffington Post isn’t trying to be the Buzzfeed of news. On the other hand, most of the other headlines had a Viralnova-y feel to them. The large number of shares says they must be doing something great.
According to the results from The Huffington Post:
- Numbers help headlines. Even a small increase for each article could mean a 50.000 share difference for a site this size.
- The number 10 proved popular again, even with the sample size of 165.
- News stories always benefit from accuracy and precision, no matter how clickbaity their titles are.
I’ve left the most interesting analysis for last: Neil Patel’s blog over at Quicksprout:
Once again, numbers did better in headlines. Neil certainly is one to experiment, there are 27 different numbers in the 59 headline articles that contained them. Perhaps he’s using The Skyscraper Method to elevate his posts to the top of Google search results, by crafting much better articles than his competitors. Perhaps he’s testing different things and reiterating. Or both.
According to the results from the Quicksprout Blog:
- You should use the numbers 5, 7 and 10 in your headlines.
- If you decide to use larger numbers (22, 37, 40), your article should be pretty epic to match the headline.
- There isn’t that much difference betwen 5, 6 and 7. It does come down to the actual content of your article.
SECTION 4: Why Do Numbers Work?
There are 5 main reasons numbers in article headlines (and digital marketing in general) work so well:
- Curiosity – Turns out, when people are interested in the question, they tend to be more interested in the result. They get even more curious, when they get the result wrong. For example, “7 New Uses For Your Old Mop” should work better than “New Uses For Your Old Mop”. (Source: Cal Tech Study)
- Number fixation – We need to know how much a product costs and 1.97 or 1.99 is seen as more valuable (more bang for the buck) than 2.00. Not only that, but because odd numbers are seen as strange, they stand out more. (Source: Journal of Consumer Research)
- Symbolic reasoning – 7 is considered lucky and magical, 13 is usually unlucky and 10 is the Pythagorean symbol of perfection or completeness. (Source: Encyclopaedia Brittanica)
- Real sense of choice – Two options is binary, one or the other. But add another one and you’ve got three options. Now you can surely choose what you want.
- Time fixation – People feel anxious and stressed when they don’t know how long something will last. Having a timeframe (i.e. “The Best Productivity You Can Implement In 7 Minutes”) sets up an expectation and makes the task easier. (Source: Columbia University Study)
SECTION 5: Who Else Has Written on the Topic of Numbers in Headlines?
It turns out – a lot of people!
How to Write Magnetic Headlines
30+ Ultimate Headline Formulas for Tweets, Posts, Articles, and Emails
8 Winning Headline Strategies and the Psychology Behind Them
A Scientific Guide to Writing Great Headlines on Twitter, Facebook, and Your Blog
Neville Medhora and DataStories:
A perfect post recipe to get shares: How we reverse-engineered OkDork posts and found out what matters
5 Characteristics Of High Converting Headlines
SECTION 6: What Have I Learned? What Are the Ways to Write Great Headlines?[bctt tweet=”Is there a recipe for creating killer titles that get more shares?”]
There isn’t a clear recipe. Not only that, but there are A LOT of different factors at play – if the article had paid promotion, what day of the week it was launched, how many times it was shared by the authors and so on.
Each blog/site has its audience. So just because the Canva Blog has a lot of “50” in their headlines, that doesn’t mean it’s a magic number for them. It could also mean that their audience has become accustomed to the numbering format.
Viralnova tests out a lot of different titles – the ones I’ve analyzed might not be the original ones. I want to believe that there’s research behind the numbers in the articles. The same adaptation thinking can be applied here, in relation to Viralnova’s readers. In today’s digital landscape, a great headline can make you stand out from the other 2+ Million blog posts written every day.
The Huffington Post wasn’t really the best choice for the News category. But it was definitely fun combing through their headlines, looking at patterns and similarities with other types of websites. Their team sees the importance of great headlines and how article traffic can vary up to 500% based on them.
Quicksprout wasn’t such a surprise – Neil Patel is the master of tweaking and optimizing. He understands what a headline does and that 80% of readers never make it past the headline. For some reason his other blog – NeilPatel.com/blog – only showed a few results in Buzzsumo, so I left that analysis for another time.
Overall, I do want to analyze more articles and websites and find a way to properly dissect a larger number of headlines at a time. Since this is only the first incarnation of Part 1, expect updates in the form of:
- More data than 165 headlines/website
- Better/interactive graphics
- Data covering the relation between headlines and comments
Part 2 will be a further examination on numbers and headlines. I’ll be focusing on 5 more categories of websites: Entertainment, Movies, Fashion, Productivity. The last category is still a surprise and you’ll get a chance to find out what it is, when “Why Is 7 Better Than 6? How To Write Effective Blog Headlines – Part 2” launches early November 2015.