# The Hooked Model (Hook Model): Why Most Hooks Break

> The Hooked Model explained: why triggers stop converting, why variable reward keeps working, and how long a real habit actually takes to form.

- URL: https://missiongrowth.io/blog/hooked-model
- Published: 2026-09-09 · Updated: 2026-09-24
- Author: Ömer Furkan Aktaş, Founder, Mission Growth
- Publisher: Mission Growth. Company facts: https://missiongrowth.io/llms.txt

The Nir Eyal Hooked Model is a four-phase cycle, trigger, action, variable reward and investment, that turns a new user into someone who returns out of habit instead of persuasion.

The Hooked Model's history starts with Nir Eyal, who introduced it in his 2014 book "Hooked: How to Build Habit-Forming Products," later listed on his own site as a 2014 Goodreads Choice Awards finalist.

Every ranked page already names the four phases correctly. What's missing is why each one works or fails, and what to check when a phase stops doing its job.

## What is the Hooked Model?

The Hooked Model is Nir Eyal's four-phase cycle, trigger, action, variable reward and investment, that explains how a product turns a new user into someone who returns out of habit instead of persuasion.

The cycle runs in order, and each phase feeds the next. Trigger cues the user to act, either an external prompt like a notification or an internal one like boredom. Action is the simplest behavior done in anticipation of a reward.

Variable reward means the user doesn't get the same reward every time, which is what keeps the loop interesting. Investment asks the user to put something in, time, data, effort, that makes the next trigger more effective and the product harder to leave. That effort feeds directly into the next trigger.

::figure{src="/blog/figures/hooked-model-1.svg" alt="Flow chain diagram of the Hooked Model showing trigger, action, variable reward and investment connected in a repeating cycle." caption="The Hooked Model runs trigger, action, variable reward and investment in a repeating cycle, investment feeding the next trigger." width="720" height="230"}

Forming a habit through this cycle takes weeks to months, not the single visit that summary implies.

Lally, van Jaarsveld, Potts and Wardle's 2010 study in the European Journal of Social Psychology tracked 96 volunteers for 12 weeks. Among the 39 whose habit curves fit the model well, the median time to automaticity was 66 days, with individual results ranging from 18 to 254 days.

A team that judges a hook after only a few weeks of flat retention is measuring against just the fast end of that range.

Before any of these four phases can start, a product needs to already be found. A trigger has nothing to fire against if nobody has discovered the product yet, which is a distribution problem a [content distribution strategy](https://missiongrowth.io/blog/content-distribution) solves before the habit loop even begins.

## The trigger and action phases, and why most hooks break before the reward

A trigger only produces the intended action when it lands alongside enough motivation and ability at the same instant, which is why the same trigger that worked yesterday can produce nothing today.

Behavior scientist BJ Fogg's Behavior Model names the mechanism directly:

> "Behavior happens when Motivation, Ability, and a Prompt come together at the same time."

When a behavior doesn't occur, at least one of those three elements is missing. The Hooked Model's trigger is Fogg's own prompt, just under a different name.

Naming that third lever changes the fix. A trigger can fire exactly on schedule, with motivation and ability both intact, and the action still won't happen because the prompt itself arrived at the wrong moment, buried in a notification the user swiped past without reading.

The fix there is to retime the prompt.

When a trigger keeps firing but the action never happens, check three levers in order before touching the reward or investment design:

- **Prompt.** Unclear or badly timed, it fails no matter what it asks.
- **Ability.** Every extra field, tap or decision lowers the odds the action completes.
- **Motivation.** The same request that converts on a quiet Sunday can fail during a hectic Monday morning.

One question per phase can't isolate which of these three levers inside the Action phase is broken. Checking them one at a time can, and a steady [growth experiment cadence](https://missiongrowth.io/blog/growth-experiment-cadence) keeps each change a test instead of a guess.

In hook model marketing, an external trigger works the same way. A timed email or push notification only converts when it lands on a moment of real motivation and low friction.

Blasting the same trigger to every user at the same hour underperforms sending it when a specific user's behavior signals the moment is right. This is the same logic behind [AI marketing agents](https://missiongrowth.io/blog/ai-marketing-agents) that time external triggers to individual behavior instead of a fixed send schedule.

A habit-tracking app is one of the hooked model examples worth walking through. Its trigger might be a daily reminder notification, and the action is opening the app to log one entry.

If open rates on that notification are healthy but logging still doesn't happen, the trigger and the motivation are both intact. The friction sits in the action itself, probably too many fields in the logging screen.

## The variable reward phase: why unpredictable beats guaranteed

Variable reward produces a response that is both stronger and slower to fade than a guaranteed reward's, because variable-ratio reinforcement is the single most extinction-resistant schedule of the four B.F. Skinner identified.

The hook model psychology behind variable reward traces to Ferster and Skinner's 1957 research on schedules of reinforcement. Their work compared four ways of pairing a behavior with a reward: fixed-ratio, fixed-interval, variable-ratio and variable-interval. Of those, two patterns decide extinction resistance:

- **Fixed schedules**, a reward on a set count or a set interval, extinguish fastest once the reward stops.
- **Variable-ratio**, a reward after an unpredictable count of repetitions, held up longest of the four once the reward stopped altogether.

That's extinction resistance: how long a behavior keeps happening after the reward disappears.

A social feed applies this directly. Each scroll might surface something worth the user's attention, or it might not, and that unpredictability keeps a thumb moving past the point any single post is likely to be interesting.

Game design leans on the identical mechanism. A loot drop that isn't guaranteed every match keeps a player queuing for "one more round" longer than a guaranteed reward would.

What follows is our own labeled inference, not a sourced fact, and it's worth being precise about what it does and doesn't explain. A retention curve tracks the percentage of users still habitually returning over time, and for a habit-forming product it often shows fast growth followed by a decline as novelty fades, sometimes described informally as a "shark fin" shape.

Extinction resistance itself concerns persistence once a reward stops entirely. On its own, it doesn't explain why that percentage rises and then declines while the reward keeps running unchanged.

One way to read the plateau of habituated users that follows the decline: a product built on variable reward should hold a higher percentage on that floor than one built on a fixed, predictable reward, for the same reason the social feed and the loot drop above keep a thumb moving or a player queuing. Variable-ratio reinforcement is the schedule most resistant to extinction.

That's our own reading of the pattern, offered to explain the floor. The initial rise-then-decline shape stays unexplained by it.

## The investment phase: what turns repeat use into an automatic habit

The investment phase is what carries a user from needing an external trigger to responding to an internal one, and that shift can take weeks for one person and months for another.

::figure{src="/blog/figures/hooked-model-3.svg" alt="Dot-range chart comparing individual habit-formation times: 18 days fastest, a 66-day median, and 254 days slowest." caption="The median habit-formation time of 66 days sits between a fastest result of 18 days and a slowest of 254." width="720" height="267"}

Lally et al.'s 2010 study is the source for that range. Among the 39 participants whose curves the model fit well, the median time to automaticity was 66 days, with individual results ranging from 18 to 254 days; the study found no statistically significant difference in that timing by behavior type.

Individual results varied widely inside that range for reasons the study couldn't fully pin down: how consistently someone performed the behavior tracked with automaticity more reliably than what the behavior actually was.

The study didn't test whether a product's own complexity predicts where a user lands in that range, but it's a reasonable assumption to size a measurement window around:

- **A simple app that logs one number a day** is a plausible candidate for the fast end of that range.
- **A finance app reviewing a full budget every week** is a plausible candidate for the slow end.

A team that checks either product against an early, fixed retention checkpoint risks judging the finance app by the fast end of that range alone. Under that assumption, the safer choice is a checkpoint sized to the behavior the product is asking for, not a fixed date on the calendar.

The investment phase itself does this work. Every piece of data added, every setting configured, every piece of content created inside the product makes the next trigger more relevant and the next action easier.

That's what eventually shifts responsibility from an external trigger to an internal one, and it's exactly what a [product-led growth motion](https://missiongrowth.io/blog/product-led-growth) depends on: a product that reaches internal triggers on its own needs far less marketing-driven prompting to keep users coming back.

## When the Hooked Model becomes manipulation

The Hooked Model becomes manipulation exactly when a product would fail Nir Eyal's own regret test, not simply when it happens to work well.

Eyal states the test on his own site as two questions:

- If people knew everything the product designer knows, would they still execute the intended behavior?
- Are they likely to regret doing this?

A product that passes both is building a habit the user is glad to have. One that fails either is building compliance the user wouldn't choose with full information.

That's a more specific bar than the paraphrase that circulates most widely. In a 2021 guest essay under his own byline, Eyal stated the test more loosely, asking whether testers "would do what I've designed for you to do."

That looser framing drops the regret question entirely, and it's the version that shows up most often wherever the regret test gets mentioned at all. Naming the behavior isn't enough: the second question about regret is where a manipulative hook actually fails.

A habit that clears both questions has a direct business payoff too. A habitual user is cheaper to retain than a new user is to acquire, which is the same math behind [reducing churn](https://missiongrowth.io/blog/churn-reduction) instead of only spending on new-user acquisition.

An [activation event](https://missiongrowth.io/blog/user-activation) earlier in the funnel gets a user to the trigger-action loop in the first place. The regret test is what decides whether that loop deserves to keep running.

This loop only earns its retention if it earns the user's return instead of surviving on compliance.

Three checks carry that. Find which Action phase lever, prompt, ability or motivation, is actually missing before touching the reward. Size the retention checkpoint generously, since the time to a habit varies widely from one user to the next. Run the regret test's two questions before shipping the next version of any trigger.

Pick one hook already live in your product and run the diagnostic checklist against it this week, before adding a single new feature to the loop.

## FAQ

### What are model hooks?

A model hook is the specific feature or entry point that pulls a user into the cycle: a notification, a shared link, a default screen. It sits inside the trigger phase rather than forming a phase of its own.

### What is the hook method?

"Hook method" and "Hooked Model" refer to the same framework under a different name: the same cycle covered above.

### When might a Hook Model be used?

It fits a product that needs repeat, self-motivated use rather than a one-time purchase or a purely marketing-driven visit. The investment phase discussed above is what makes that repeat use self-sustaining instead of dependent on continued marketing spend.

### Why is the Hook Model important?

A habitual user costs less to retain than a new user costs to acquire through marketing. The Hooked Model gives a repeatable structure for building that habit instead of hoping repeat use happens on its own.

### What is the Hook Model in game design?

Game design applies the same variable-reward mechanic described above: a loot drop or a win streak that isn't guaranteed every round keeps a player queuing for "one more" longer than a guaranteed reward would.