Clickbait: The Headline Wins, the Text Loses, the Site Pays
Half past ten at night. The coffee next to me has gone cold, the screen too bright. In front of me a headline I wrote myself: "You won't believe what happened next." I don't believe it. I clicked anyway. Nothing ever happened.
What clickbait actually is
Nothing complicated: a headline promises something, the text doesn't deliver it. The technical term is the curiosity gap – you say enough to itch, and too little to satisfy. "This one thing ruins your ranking." Which thing? Click. This isn't a new trick, tabloids have done it at every newsstand for a hundred years. The only difference: the newsstand never measured how long you held the paper before putting it back. The internet measures.
The usual suspects
- "You won't believe what happened next."
- "Number 7 shocked me."
- "Experts hate this one trick."
- "This changes everything."
You recognize them by the fact that they still mean nothing after you've read the article. Number 7 shocked no one. Number 7 was a bullet point.
Why it works anyway
Because that's how we're wired. Incomplete information creates tension, and tension wants to be resolved. That's not a character flaw, that's wiring. That's why clickbait works even on people who know exactly what it is – I know it, and I click anyway. At half past ten at night, everyone clicks. And that's the trap: the trick works, but only once per person.
Thesis: clickbait is just good marketing
A punchy headline is, in the end, just advertising for your own text. Every tabloid did this long before the internet existed. If you don't stand out, you don't get read.
Antithesis: what it costs
When someone lands on a page from Google, stays briefly, and goes back to the search results to click the next one, Google gets a fairly clear signal of how satisfied that person was. This is called pogo-sticking. Google regularly pushes back on the idea that there's a simple dwell-time ranking factor you can look up in a table – that's probably accurate and still misleading, because evaluating helpful content ultimately comes down to the same question: did this text end the search or not? A headline can get you the click. Whether you get to keep it is decided by the second paragraph. And anyone disappointed twice on your site clicks the result below yours the third time, even if you rank higher.
Clickbait vs. honest: a direct comparison
| Clickbait | Sharp, but honest |
|---|---|
| "This one thing ruins your ranking" | "Why six of our own pages compete for the same keyword" |
| "You won't believe what we measured" | "11,000 impressions, zero clicks – the breakdown" |
| "Experts hate this trick" | "The ten-minute title test" |
The right column isn't the boring option. It's the option that doesn't leave you feeling conned after reading it. The test is simple and sometimes stings: read your headline, then read your first paragraph. If the headline promises more than the paragraph delivers, change the headline – or rewrite the paragraph.
Conclusion: four things I do instead
The answer goes first. Anyone asking a question should find it answered in the first paragraph; the rest is for people who want more detail.
Numbers over adjectives. "Amazing results" is empty air. "From position 38 to 12 in eleven weeks" is a headline that claims something checkable.
Check what actually happens. Google Search Console shows, for every query, how many people saw your title and how many clicked. Many impressions, few clicks: the headline is the problem. Many clicks, fast bounces: the text is the problem.
Write the headline last. You can't honestly summarize something that doesn't exist yet.
The meta layer: I had ChatGPT rate my own article
Before the original article went live, a language model reviewed it critically, not to optimize or rewrite it, just to assess it: content, entertainment value, SEO, GEO. Result: idea/experiment 9/10, entertainment value 8/10, originality 9/10, subject-matter content 7.5/10, SEO potential 8/10, GEO/LLM potential 8.5/10, credibility 7/10 – positive overall, but not perfect. Reviewers especially liked the combination of definitions, examples, a comparison table and an FAQ, which makes the text easy for AI systems to parse too. Criticized: overly absolute claims about ranking factors and a source reference that stayed too generic. That's an interesting limit of AI content right there: a text can sound convincing and still not be editorially finished.
The short version
Clickbait isn't evil. It's just expensive, and you only notice the bill later, when the numbers stop growing and no one can say exactly why. The honest headline gets fewer clicks on day one. It gets more on day thirty. Write the headline your text can actually deliver on. Then go to sleep.
Sources
- Rankmio: Clickbait: The Headline Wins, the Text Loses, the Site Pays (original article, the basis for this piece, including the full ChatGPT review)
- Google Search Central – on evaluating content by quality, not how it was produced
FAQ
- Is clickbait bad for SEO?
- Indirectly, yes. Users who bounce back to search disappointed signal that their question wasn't answered. There's no single "dwell time ranking factor", but systems for evaluating helpful content work on exactly this question.
- What is the curiosity gap?
- The deliberately open part of a piece of information. The headline hints at a result without naming it. That gap creates tension many people resolve by clicking.
- Where's the line between a good headline and clickbait?
- At delivery. Sharpening, sparking curiosity, and making a promise are all fine. It becomes clickbait when the text doesn't keep that promise.
- How do I know if my titles are working?
- Through click-through rate in Google Search Console. Many impressions with few clicks points to a weak title. Many clicks with fast bounces points to a text that doesn't deliver.