Hooked by Nir Eyal: The Hook Model, a Worked Example, and Its Limits
8:14 a.m. A notification appears. You tap it. There is a message, a comment, and something else you did not expect to see. You scroll. Before closing the app, you reply to someone and update a preference. Later that afternoon, nobody sends a notification. You open the app anyway. That progression is almost a miniature version of Nir Eyal's Hooked.
The book examines how certain products move from being something users consciously remember to use into something connected to a routine, need, or emotional state. Eyal calls the mechanism the Hook Model: Trigger → Action → Variable Reward → Investment The framework is simple enough to memorize. Human behavior is not.
First, Something Starts the Behavior
A trigger can be visible. An email. A notification. An icon. A reminder. A recommendation. Those are external triggers. The product is effectively saying: "Come back." But the more interesting stage arrives when the reminder moves inside the user. Boredom. Loneliness. Uncertainty. Curiosity. A recurring task. Now the person does not need the product to call them. The situation itself has become associated with the product. Bored? Open the app. Unsure what to do next?
Check the tool. Need to feel connected? Open the platform. That is a much stronger relationship than a notification alone.
Then the Product Must Make the Next Move Easy
A trigger means little if the user encounters friction immediately afterward. Five menus. A confusing interface. Too many questions. A long setup. Eyal's Action stage is about making the desired behavior simple enough to happen. This does not mean every product should become simplistic. It means unnecessary difficulty creates an opportunity for the user to stop. If someone already wants the outcome, the product should not make them fight for it.
The same principle appears from the individual side in Atomic Habits, where James Clear explores how cues, friction, and rewards can make repeated behavior easier or harder to sustain.
The Reward Works Because You Do Not Know Exactly What Comes Next
Open a feed. Maybe nothing interesting appears. Maybe there is something hilarious. Maybe a useful article. Maybe a message you were hoping for. That uncertainty changes the experience. Eyal divides variable rewards into several broad forms. There are social rewards: approval, comments, recognition. There are rewards of discovery: information, opportunities, something worth finding. And there are rewards connected to the self: progress, completion, mastery. The useful idea is not that randomness magically makes people addicted.
That is too simplistic. Anticipation can make repeated engagement more compelling when there is a meaningful possibility of receiving something the user values. The reward still has to matter.
Then the User Puts Something In
A profile. Preferences. Photos. Followers. Progress. Time. Content. Effort. That is Investment. The product now contains more of the user's history or personal configuration than it did before. A language app knows where the learner stopped. A social platform contains relationships and conversations. A productivity system contains projects and workflows. Returning becomes more useful because previous use improved the next experience. The cycle can then begin again.
The same mechanism can support very different outcomes. A learning app can help someone practice daily. A fitness product can make healthy behavior easier to repeat. A productivity tool can help users organize recurring work. The same principles can also be used to maximize screen time simply because attention creates revenue. That creates an important distinction. Habit is not automatically value.
A person returning fifty times a day does not prove the product is helping them. Engagement is a behavioral metric. Benefit is a human outcome. A responsible product creator has to ask both.
What Product Builders Should Actually Look For
The wrong starting question is: "How can we make people open this every day?" The better starting point is: "What recurring problem would make repeated use genuinely useful?" A product with no recurring value cannot be rescued forever by clever notifications. The hook works best when it attaches itself to something worth returning for. Then designers can examine the rest: What triggers the need? How much friction stands between the user and value?
What reward makes the experience worthwhile? What can the user invest that improves the next visit? That is far healthier than treating engagement as a mysterious number on an analytics dashboard.
Once You See the Loop, You See Your Own Behavior Differently
This is where Hooked becomes useful even for people who never design products. The next time you open an app automatically, you can examine the sequence. What happened just before I opened it? Was I bored? Did I receive a reminder? What did I expect to find? What have I already built inside this product that keeps me returning? The model turns an automatic behavior into something visible.
That also connects with Thinking, Fast and Slow, which examines how much of human judgment and behavior can happen automatically before deliberate thought steps in. And visibility creates choice. A product designer can use the Hook Model to build better recurring experiences. A user can use the same model to notice when a recurring experience has begun using them. The framework itself is neutral. What matters is the direction of the habit it helps create.
A Worked Example: A Reading App That Respects the Reader
This is an illustrative design exercise created for this article, not a case study from the book or a tested product. Imagine a reader who saves book notes but rarely returns to them. The useful outcome is recalling an idea when it matters, rather than opening an app as often as possible.
The weakest link might be the notes themselves. If they are vague, a better notification will not make reviewing them worthwhile. Before adding reminders, a designer could help the reader capture a specific idea, its context, and one situation in which it could be used.
Start with a small, voluntary trial and a question: can readers retrieve and explain a useful note more easily after a week? Record completed reviews, but also ask whether a note helped with a real decision. A higher open rate with no improvement in recall would be a reason to revise the design, not declare success.
Give readers an obvious way to pause reminders, export notes, and leave. Avoid punishing a missed day. A reading tool should support learning even if success eventually means spending less time in the tool.
Where the Hook Model Is Less Useful
Not every valuable service needs a habit. An emergency repair service or an occasional document-signing tool may serve users best through reliability when needed. Forcing daily engagement into those situations adds friction. The model also describes a possible behavioral loop; it does not by itself prove that a product will retain users or improve their lives.
Sources and Further Reading
For the author's explanation and supporting resources, see Nir Eyal's Hooked resources and his Hooked workbook. The reading-app exercise and evaluation questions above are BookGistX commentary, rather than claims of results reported by Eyal.