The Psychology of Killer Hooks & Addictive Storytelling

Here are two openings for the same article.

In today's rapidly evolving technological landscape, artificial intelligence is transforming how organisations operate.

Two agents are given the same task. Both return the same, correct answer. Only one of them is safe to deploy.

The first is not wrong. Every word of it is defensible. It is also invisible: the eye slides off it, because nothing in it could surprise anyone.

The second makes a promise. It states a fact, then contradicts the obvious conclusion from that fact, and the gap between the two is a question the reader now wants answered. (It is the actual opening of this site's architecture article, rewritten from a draft that began on a far more general sentence.)

This article is about the difference between those two openings: why some ideas hold attention while others are ignored, and how to design for the first without sliding into manipulation.

Reading is a series of decisions to continue

Nobody commits to an article. They commit to a sentence, and then — if it earned it — to the next one.

That reframes the first line's job. It does not need to summarise. It needs to answer one question: why should I give you my next five seconds?

An opening built from context ("In today's…") answers: no reason yet. An opening that creates a question answers: because you now want to know something. Everything else in this article is a way of producing the second answer, over and over, until the piece is finished.

Mechanism 1: The information gap

The most useful single model of curiosity comes from the behavioural economist George Loewenstein. His 1994 review proposed that curiosity arises when attention is drawn to a gap between what we know and what we want to know. Two details of that theory matter to anyone who writes.

First, the gap needs existing knowledge. You cannot be curious about something you know nothing about. The reader has to know enough to see the missing piece.

Second, the gap has to feel closable. A mystery with no plausible answer produces indifference, not curiosity.

Compare:

• "The company failed because of poor decisions." — no gap; the loop is already closed.
• "The company had every resource it needed. So why did it collapse in eighteen months?" — a specific gap, with enough information that the reader starts guessing.

A piece that holds attention keeps opening and closing gaps like this:

1. A question
2. A partial answer
3. A sharper question
4. An unexpected development
5. A new question
6. Resolution

These are open loops. The craft is giving the reader exactly enough to become curious about what is missing — not so little that they are lost, not so much that nothing is left to discover.

Mechanism 2: Prediction and its violation

A widely used family of models in cognitive science treats perception as prediction: the mind continuously forecasts what comes next and pays particular attention where the forecast fails. You do not need to accept the whole theory to use its practical consequence.

If the reader can predict your next sentence, it carries no information for them. If nothing is predictable, they cannot build a model of where you are going, and they leave. Compelling writing lives between the two. It lets the reader form an expectation — I know where this is going — and then bends it — …but not quite.

That is what a pattern interrupt is: a sentence that contradicts the reader's default.

• Expected: "Successful people wake up early."
• Interrupted: "Waking up earlier doesn't make people more productive. It exposes why they weren't."

The interruption does not need to be dramatic. It needs to create a friction the reader wants to resolve. What changed? Why did the assumption fail?

Mechanism 3: Stakes and specificity

Information tells the reader something. Stakes make them care about it.

"The company was running out of money" is information. "They had enough cash for exactly three weeks" is a clock. The second version gives the reader a consequence and a deadline, and every sentence after it is read against that clock.

Specificity does related work. "It was a difficult day" asks the reader to take your word for it. "At 4:17 p.m. the dashboard showed three critical alerts at once" puts them in the room. Show the consequence rather than naming the emotion: not "he was nervous", but "he reread the same sentence three times before pressing Send."

Chip and Dan Heath's Made to Stick makes concreteness one of its six principles for ideas that survive, and it is the cheapest of the six to apply: replace the abstraction with the instance.

Mechanism 4: Tension — desire against obstacle

Beneath almost every story that holds attention is one structure: someone wants something, and something is in the way.

Without an obstacle there is nothing to wonder about. Tension is the distance between where the protagonist is and where they need to be. In nonfiction, the protagonist is often an idea, a system, or the reader themselves.

A craft rule captures the mechanics neatly. Popularised by the South Park creators Trey Parker and Matt Stone, it says: if the beats of your story are joined by "and then", you have a chronology. Join them with "but" and "therefore" and you have a story. Each beat should change the situation that the next one inherits.

Controlled revelation

The most common structural mistake is to reveal everything at once — the data dump. A piece that works decides deliberately what the reader should know now, what later, what they should merely suspect, and what stays hidden until it can land.

Think of it as a sequence of doors, each more interesting than the last, until the final one connects the pieces. Every answer should make a better question possible.

This is also why strong nonfiction is often organised around one central question that each section answers a little more deeply. The question is the thread; the sections are the doors.

Rhythm is part of the argument

Short sentences land. Longer sentences, with room to qualify and connect, create space for reflection. Paragraph breaks give the eye somewhere to rest.

A one-line paragraph creates emphasis.

Like that.

Writing that alternates set-up, tension and release reads as movement. Three long explanatory paragraphs in a row read as homework, however good the content inside them.

The architecture of a piece that holds attention

The payoff of nonfiction is the same as the payoff of fiction: something changes. The reader started believing X and finishes understanding Y. An article that only delivers information has skipped the last row.

Five hook patterns — and how each goes wrong

The principle beneath all five is the same: create a gap between what the reader expects and what the piece promises to reveal — and then reveal it.

Where this meets engineering

These mechanisms are not confined to essays. Most technical communication fails for the same reasons weak openings do.

Onboarding and product design. A first screen has the same job as a first sentence: answer why should I give you the next five seconds? An empty state that says "No projects yet" is a closed loop. One that shows what the first project will look like opens one.

Design documents and architecture reviews. Most design docs begin with background. The reader wants the decision and the tension behind it: what are we choosing, what does it cost, and what breaks if we are wrong? Lead with the failure the architecture prevents, and the diagram that follows has a reason to exist.

READMEs and developer experience. A README's first screen should say what the project does and let the reader prove it with one command. Every paragraph before that is a paragraph the reader must take on faith.

AI interfaces. Here the payoff is trust. Streaming the first words quickly answers is anything happening? — but the more important rule is honesty about state. An interface that presents a confident answer while its model is actually unavailable has broken the promise the interface made.

Presenting complex systems. The same arc works for an architecture walkthrough. Open with a concrete failure (the hook). Ask why the obvious fix does not work (the question). Show what it costs (the stakes). Let the architecture arrive as the answer (the revelation). An audience that has felt the problem does not need persuading that the solution matters.

Persuasion versus manipulation

"Addictive storytelling" should make any engineer a little uneasy, and the unease is useful. The same mechanics that sustain attention can exploit it.

The difference is what happens after the gap opens:

• Earned: curiosity → discovery → satisfaction.
• Exploited: curiosity → delay → disappointment.

Clickbait is a curiosity gap with nothing behind it. An endless feed is an open loop engineered never to close. Both work, briefly, and both spend the reader's trust to buy a click.

A practical test: would the reader endorse the technique if they could see it? A reader who notices that an article opened with a sharp question and then answered it feels respected. A reader who notices that the answer was withheld for eleven paragraphs to keep them on the page feels used. Write for the first reader.

The goal is not to trap anyone. It is to make continuing worthwhile.

The editing pass

One question does most of the work: why is this sentence here? A sentence should do at least one job — create curiosity, raise tension, deliver information, establish context, change a perspective, advance the argument, or set up the next idea. If it does none of these, cut it.

Before publishing anything that asks for attention, ask:

1. Does the opening give a reason to continue?
2. Does every section add something the previous one did not?
3. Is something at stake?
4. Does the piece open questions — and close them?
5. Will the reader see the subject differently at the end?

Key takeaways

• Readers commit one sentence at a time. The first sentence's job is a promise, not a summary.
• Curiosity needs a gap the reader can see — and believe can be closed.
• Stakes and specificity turn information into something a reader cares about.
• Structure is controlled revelation: each answer should make a better question possible.
• The same mechanics govern onboarding, design docs, READMEs, AI interfaces and architecture reviews.
• The line between persuasion and manipulation is whether the promise is kept.

Further reading

• George Loewenstein, "The Psychology of Curiosity: A Review and Reinterpretation," Psychological Bulletin 116(1), 1994 — the information-gap theory of curiosity.
• Chip Heath and Dan Heath, Made to Stick: Why Some Ideas Survive and Others Die (2007) — concreteness, unexpectedness and the curiosity gap in practice.