There is a formula. I resisted believing this for a long time because it felt reductive, like reducing writing to a math problem. But after looking at the data, the pattern is hard to argue with.
Number + Adjective + Keyword + Promise. "7 Common JSON Mistakes That Break Production Apps" follows this formula. "JSON Mistakes" does not. The first version has specificity, a number, and an implied benefit. The second is a label, not a headline.
This does not mean every headline needs to follow the formula mechanically. But understanding why certain structures perform better gives you a starting point that you can then make your own.
What the data actually says
CoSchedule analyzed over a million headlines and found consistent patterns in what gets clicked. Headlines with numbers outperformed those without by a significant margin. Headlines between 6 and 13 words hit the sweet spot for both search engines and social shares. Headlines that made a specific promise, either teaching something or revealing something, outperformed vague ones.
BuzzSumo ran a separate analysis of 100 million article headlines and found that headlines beginning with "How to" consistently ranked among the top performers. "X things" and "X reasons" formats also performed well. The word "why" in a headline correlated with higher engagement than "what" or "how."
None of this should be surprising. Specific headlines set clear expectations. Readers click when they know what they are going to get.
The odd number effect
A study published by Conductor found that odd numbers in headlines generated 20% more clicks than even numbers. "7 Ways" beats "8 Ways." "5 Reasons" beats "6 Reasons."
The prevailing theory is that odd numbers feel more authentic and less manufactured. A list of 10 feels like the author rounded up to hit a clean number. A list of 7 feels like the author included exactly what was relevant and stopped.
I have no idea if this theory is correct, but the data is consistent across multiple studies. I default to odd numbers in my headlines now. It costs nothing and the data supports it.
Power words and emotional weight
Certain words consistently increase click-through rates. "Mistake" performs well because it triggers loss aversion. People are more motivated to avoid mistakes than to achieve gains. "Secret" works because it implies exclusive knowledge. "Simple" works because it promises low effort.
The Advertising Research Foundation categorized headline words into emotional buckets. Words that trigger curiosity ("surprising," "unexpected," "strange") outperform neutral descriptors. Words that trigger urgency ("now," "today," "before") outperform timeless phrasing.
But there is a line between emotional resonance and manipulation. "This one weird trick" is a power phrase that works, which is why spam advertisers use it relentlessly. Using it labels you as spam. The goal is to use emotional language that accurately represents your content, not to bait people into clicking on something that does not deliver.
The curiosity gap
The curiosity gap is the space between what the reader knows and what the headline implies they could know. Upworthy built an entire media company on this technique before it became synonymous with clickbait.
A functional curiosity gap: "Why Senior Developers Avoid These 3 JavaScript Patterns." The reader wonders which patterns and why seniors avoid them. The headline promises specific, actionable information.
A manipulative curiosity gap: "You Won't Believe What This Developer Did With JavaScript." The reader has no idea what the content is. The headline promises nothing specific. This is clickbait.
The difference is whether the headline gives enough information for the reader to make an informed decision about clicking. Good headlines create curiosity while being honest about the content. Bad headlines create curiosity by withholding all information.
Title case vs sentence case
This is more contentious than you would expect. Title case capitalizes the first letter of major words: "How to Write Better Headlines for Technical Articles." Sentence case capitalizes only the first word: "How to write better headlines for technical articles."
Data from email marketing A/B tests generally shows sentence case performing slightly better in casual contexts. It reads as more natural and less promotional. Title case performs better in formal contexts and on platforms where it is the norm, like major news publications.
On Dev.to, I use sentence case for most articles because the platform has a casual tone. On LinkedIn or a company blog, title case might be more appropriate. Match the convention of the platform.
A/B testing your own headlines
If you publish on a platform that supports it, test two headlines for the same article. Publish with one headline, note your view count after 48 hours, change the headline, and compare. This is not scientifically rigorous, but it gives you directional data.
I keep a spreadsheet of my headlines and their performance. Over time, patterns emerge. My audience responds to numbers and specific technology mentions. They ignore superlatives like "best" and "ultimate."
The real test
A headline works when it accurately sets expectations and the content delivers. A brilliant headline on a mediocre article produces high bounce rates and low engagement. A mediocre headline on a brilliant article means fewer people find it.
The goal is alignment between the promise and the delivery. "5 PostgreSQL Query Patterns That Halved Our Response Time" works if the article actually contains 5 patterns and they actually improved response times. The headline is specific, numbered, includes a keyword, and makes a concrete promise. That is the formula at work.
Where people go wrong is writing the headline first and then stretching the content to fit. Write the content first. Then describe what you wrote in the most specific, honest, and compelling way you can. The formula is a tool for description, not a template for fabrication.
I'm Michael Lip. I build free developer tools at zovo.one. 350+ tools, all private, all free.
Top comments (0)