Every guide tells you the first line matters. Most then give you templates, which is how you end up with feeds full of "Here's what nobody tells you about..." and "I made a huge mistake so you don't have to."
Those work in the narrow sense that they get attention, and fail in the broad sense that the reader feels manipulated and remembers the feeling rather than the content.
What a hook is actually doing
One job: making the second line worth reading.
That is it. It does not need to shock, tease, or promise. It needs to create a small gap between what the reader knows and what they want to know.
The most reliable way to create that gap is a specific fact with an implication you have not stated yet.
The fix was one line. Finding it took four hours.
The gap is obvious. Why four hours? The reader keeps going, and nothing was exaggerated to get them there.
Generate several, keep one
The practical technique that works better than any template: write five openings before choosing.
The first one you write is almost always the most obvious framing, which is why so many posts open with setup. The fourth or fifth tends to be the one that starts closer to the interesting part.
Score them on whether a stranger would read the next line. Not whether they are clever.
Openings that work, as shapes
Not templates to copy, but patterns worth noticing:
The inverted expectation. "Building the app was not the hard part."
The unglamorous origin. "This started as a spreadsheet."
The subtraction. "I deleted half the MVP before launching."
The misplaced blame. "The bug was not in the AI. It was in the data."
The admission. "I almost did not build this."
What they share: each states something specific and true, and each leaves a question open. None of them promise value in the abstract.
Openings to avoid
"Excited to announce." Signals a press release. People scroll on reflex.
"Here's something crazy." You have asserted it is interesting instead of demonstrating it.
"N lessons every founder needs to know." Listicle framing invites a listicle, which is rarely what you actually have.
"Let me tell you a story." Just tell it.
Questions as openers. "Ever wondered why X?" mostly gets "no" and a scroll.
Write it last
The hook should be written after the post, not before.
You do not know what the most interesting part is until you have written the thing. Writing the opening first means committing to a framing before you know what you are framing, and then bending the content to match it.
Write the post. Find the best sentence in it. That is usually your opening, and the post often reads better with it removed from where it was.
The test
Read only the first line. Would you read the second?
Then the harder one: does the post deliver what the first line implied?
A hook that promises more than the post contains is clickbait regardless of how it is phrased. That gap is what readers are actually reacting to when they say something felt like clickbait, and it is entirely within your control.