What Makes Content Addictive?

Sagar Udasi, 21^st December, 2025

I've spent around a decade consuming educational content online and learning from it. Over the years, I did observe a pattern in videos that had 10 million+ videos, compared to the ones with 50k views despite being technically better in all ways. The difference isn't production quality or even teaching ability. It's the structure. The content is engineered in a way that each minute creates a stronger reason to watch the next one. The viewers feel they are approaching insight, not consuming information. It's optimally designed to fancy with the brain's neurochemistry of addiction.

In it's core essense, addictive content creates a specific neurochemical loop: curiosity spike \(\rightarrow\) partial satisfaction \(\rightarrow\) new curisoity spike. It's not about keeping people in suspense, that's just annoying. It's about making the resolution of one question, and immediately reveal a more interesting question.

The Structure of Addictive Content: Long-form (> 10 mins)

At a high level, long-form videos work when they stimulate the experience of discovery. Not the result of the discovery, but the process of discovery. Most educational content fails because it transmits conclusions. Great content re-enacts the cognitive struggle that led to the conclusion. This experience can be engineered in a video in four phases. Treat these phases like the typical parts of an engagaing video; basically it's anatomy.

Phase 1: The Attention Phase

Curiosity is different from interest. It's a psychological tension caused by inconsistency. The brain pays attention when expectations are violated, predictions fail, a pattern appears without an explanation, etc. So, attention is engineered by creating inconsistency. In pop-culture, this inconsistency is also known as the hook. Interestingly, these hooks can be categorized in seven categories by the kind of norm they violate.

Now, of course, we won't use examples mentioned above directly as a hook.

The attention hook must pose a question or an idea that the video explicitly promises to resolve or clarify. If resolution isn't promised, curiosity decays. Also, the framing matters.

• "What is special relativity?" is dull.
• "What did Einstein say in his special theory of relativity?" is informative.
• "Why can't you go faster than light?" is the one we need!

Let's assume we successfully chose the right hook style based on the content, the subject, the intent, and the audience; and we did grab the attention of the learner in the first thirty seconds to a minute.

Phase 2: The Commitment Phase

Even if the question/hook is interesting, why should I care about it as a viewer? To retain the viewer, we need to achieve two things: (a) make viewer commit to a position, and (b) prove their current mental model is insufficient.

The entire point of making viewers watch an educational video is to deliver something they don't know. However, most videos either get too aggressive in making their viewers realize their current understanding is inadequate, or they remain too gentle making viewers feel the content is boring and is waste of their time. The most common way to make someone feel dumb is to speak in a very high-level convoluted jargonish language which is not beginner-friendly, and professor carrying a visible persona that they know the subject pretty well and viewers are novices. On the other hand, the most common way to make lecture boring is to avoid the intricacies in depth which does justice to the subject, and giving it a very superficial treatment. One has to achieve the above-mentioned two items very strategically.

Derek Muller (Veritasium) interviews people on the street. Most people give answers that will match viewers'. When you have a say, an opinion, an idea which you want to get tested willingly, you picking up a side, etc., it means you're committed to a position. You're now invested.

He then shows how our common intuition fails and demonstrates the wrong answer. Since, this is a collective failure of us and everyone who were interviewed, that failure doesn't feel personal. It naturally and tacitly proves that our current mental model is insufficient, and we feel the desire to need a better one.

Another interesting thing which Veritasium ensures is that this failure is local, not total. He raises a very powerful question in the start, and then segues to an experiment they conducted. So even when the audience fails, they failed in that gamified experiment. They didn't explicitly failed in answering the core question. Partial success \(\rightarrow\) stronger desire for refinement.

Assuming, you successfully engineering the video and bought user's commitment. You retain the user now for the next 3-5 minutes.

Phase 3: The Content Delivery Phase

So far, the video is interesting, not addictive (yet). This is the phase where magic happens and the secret is nested curiosity loop. The brain stays engaged when each step resolves something, but opens a larger or deeper question.

Consider this phase as a series of segments, and each segment follows this template: Start with basics to answer the question posed in the hook \(\rightarrow\) Provide a partial resolution \(\rightarrow\) Reveal the limitations \(\rightarrow\) Escalate to a deeper question \(\rightarrow\) develop to the resolution for this newly posed question and repeat the loop.

Textbook example of this strategy comes from Veritasium's video on How Electricity Actually Works¹. He starts by asking the very basic question, how electricity flows. The intuitive answer is via the flow of electrons in the metal. He then breaks this intuition by an experiment where the wire is long enough that it should take electrons sufficient time to reach to the light bulb. But the bulb turns on immediately! So, surely it's not the travelling electrons that is causing the flow of electricity. He presses again, "Okay, so electricity doesn't flow through the wire the way you thought... but then how does it actually work?" This time he reveals an idea, or simulation, or experiment to show that electricity flows as electro-magnetic waves. Hence, it also travels at the speed of light and the bulb lights up immediately despite a very long wire. But then he raises another fundamentally deep question, "Right, so it's in the electromagnetic field... but wait, if that's true, then why do we even need wires?" Each answer is satisfying enough that you feel progress, but incomplete enough that stopping would mean living with half-knowledge.

This pattern is so important that I call this as The Law of Just-Enough Satisfaction. Too much resolution \(\rightarrow\) viewer leaves. Too little \(\rightarrow\) viewer gives up. The sweet spot is, I understand more than before, but stopping now would be irrational. This works psychologically because of Zeigarnik effect that says unfinished tasks stick in mind for longer that the tasks that were finished. People tend to forget the tasks they finished, but remember the ones that are yet to be done. Also, this can even work due to effort justification, a variant of sunk-cost fallacy (I've come this far, let's finish the rest of the video too).

Some basic tactics to follow could be:

• Answer the how, not the why.
• Resolve mechanics before meaning.
• Delay naming the concept.
• Introduce formalism only when begged for.

In 3Blue1Brown videos, you see eigenvectors long before you hear the term. By the time notation arrives, it feels inevitable.

It depends on the content topic, the audience, and the number of segments (number of questions in depth in loop) that we can come up with, that will eventually decide for how long can I retain the user. Ideally, with 2-3 deep questions, one can retain the user from 10-12 minutes to even 30-35 minutes based on the quality of execution.

Phase 4: The Upgrade Phase

Finally, a lot matters on how you end the video content. The best videos make you feel like you've been upgraded, not just informed. You're not the same person who started watching. The goal is to provide this satisfaction. Hence, the name is intellectual upgrade phase.

The trick here is simple. Great videos end by connecting ideas, showing reuse of the same idea in other applications, opening new doors, or setting up the base/curiosity for the next concept that follows. Great videos end not by closing the topic, but by expanding it. Never fully close the loop.

The Structure of Addictive Content: Short-form (< 100 seconds)

Short-form is harder because we have to create the complete addiction loop in under 100 seconds. The structure is necessarily different.

Phase 1: The Hook

We have to create a hook in 5 seconds! Hence, you don't try to hook with a question, but with some bold statement or violation. Something that is impossible, contradictory, paraodixcal or extreme. We only have 5 seconds before they swipe.

For example, "This shape has infinite perimeter but finite area. How?!"

Phase 2: The Commitment

Make them commit to a prediction. Pause and think: which shape is bigger? or "Before I show you why, guess what happens." Do not make these guesses cognitively heavy. If one needs to think hard to come up with a guess, they will simply swipe.

You have to buy the commitment in the next 10 seconds.

Phase 3: The Delivery & Reveal

We don't have time to break the assumptions, and develop our intuition to the answer. So, we must reveal the answer immediately and finish our nested curiosity loop as quick as possible. For example, "It works because of X... but that means Y, which seems impossible." Then resolve the mystery Y and reveal yet another deeper pattern.

Visuals make a great deal in short-form content. The faster the screen changes, the more it flickers with content moving, the more engaging it is.

Phase 4: The Satisfying End

End with either a connection ("this is why [familiar thing] works") or an implication ("this means we can [surprising application]"). Other ideas to come up with satisfying ends can also be explored based on the subject matter.

The goal is to make them immediately want to watch the next video.

Summary: Seven Principles of Addictive Educational Content

1. Start with a Hook. Not Information. Hooks that violate something deep demand resolution. "Here's how integration works" is informative. "Why does integrating this function give us \(\pi\) ?" is a hook.

2. Make Them Commit Before You Reveal. When people commit to a prediction, they become invested in the outcome. Their ego is now involved. They'll keep watching to see if they were right.

3. Reward Quickly, Completely, Yet Partially. Each mini-resolution should feel satisfying, but it should immediately reveal a new question. The satisfaction says "you're making progress." The new question says "but you're not done."

4. Respect the Curiosity Curve. Curiosity follows a curve: it peaks at moderate uncertainty. If something is too easy, there's no curiosity. Too hard, curiosity collapses into confusion. The art is keeping people in the zone where they're slightly ahead of their current understanding.

5. Build Understanding, Don't Transfer It. The most addictive educational content makes people feel like they're discovering, not being taught. Use the Socratic method compressed: pose the question, show why naive approaches fail, give just enough hint, let them feel the "aha," confirm they were right.

6. Visual Representation Must Do Work. Visuals aren't decoration. If you can't explain why the visual representation makes something clearer than words, cut it.

7. Create Intellectually Expanding Cliffhanger Exits. End when understanding is 85% complete, not 100%. That 15% gap is what makes them click the next video. But the 85% must feel solid. Also, the exit of the video should conceptually expand to new ideas and applications.