Cue, Routine, Reward: Why Website Blockers Don't Break the Habit Underneath Them

Almost everyone who has tried a hard website blocker has also, at some point, uninstalled it — or found the one loophole that lets the old routine keep running. That’s not really a willpower story. It’s what happens when a tool removes the door but leaves the part of your brain that automatically reaches for the doorknob completely untouched. Understanding why requires looking at what a habit actually is, mechanically, rather than treating “doomscrolling” as a single bad decision made over and over.

Habits aren’t decisions — they’re loops

Psychologist Wendy Wood, whose research at USC has spent decades studying how habits form and persist, estimates that something like 40 to 45 percent of daily behavior is performed habitually — in the same physical or digital context, with little active deliberation involved. Wood’s work, along with collaborators David Neal and Jeffrey Quinn, describes habits as running on a three-part loop: a cue (a time of day, a notification, the specific boredom of a loading screen), a routine (open the app, start scrolling), and a reward (novelty, social validation, distraction from whatever you were just doing). Repeat that loop enough times in a stable context and the behavior stops being a choice at all — it becomes something closer to a reflex, triggered by the cue before the “should I do this” question ever reaches conscious thought.

This is the part hard blockers misunderstand. A blocklist targets the routine — it makes step two of the loop impossible on one specific site. But the cue is still firing, and the craving that cue produces doesn’t know the site is blocked. So the loop looks for the nearest available substitute: a different app, a different tab, a different site that scratches the same itch. Block one and the cue routes around it, because the cue was never actually about that particular website. It was about whatever moment of boredom, stress, or transition triggered the loop in the first place.

Where the Fogg model fills in the gap

Stanford researcher BJ Fogg’s Behavior Model, first formalized in a widely cited 2009 paper, offers a complementary way to see the same problem. Fogg’s shorthand is B = MAP: a Behavior happens when Motivation, Ability, and a Prompt converge at the same moment. Miss any one of the three and the behavior doesn’t occur.

A hard blocker attacks Ability — it tries to make the behavior physically impossible. That sounds like the strongest lever, and for some people, briefly, it is. But Ability is also the easiest of the three to route around: a new tab, a different device, a five-second uninstall. Motivation and Prompt are untouched, so the moment Ability dips even slightly — a lapse, a workaround, a bad day — the other two are still fully loaded and the behavior fires immediately. Fogg’s own framing of persuasive design leans instead on shaping the moment around the prompt: making the unwanted path slightly harder and the alternative path easy enough to take instead, rather than trying to seal a door shut completely.

Why friction works where denial doesn’t

Wood’s more recent writing on behavior change points to three real levers for breaking an unwanted habit loop: changing the reward the loop delivers, disrupting the context so the original cue stops firing as reliably, and adding friction that makes the habitual response slightly harder to complete on autopilot. That third lever is the interesting one, because it doesn’t require removing the option at all — it just requires inserting a beat between the cue and the routine, long enough for the slower, deliberate part of your thinking to catch up with the fast, automatic part that’s already halfway into the scroll.

That’s a different theory of the problem than a blocklist. A blocker bets that if the door is locked, the craving gives up. Friction bets something more modest and, per the research, more durable: that most people don’t actually want to be mid-scroll fifty minutes later, and if you can just get them to notice the moment — a full stop, however brief, however small — a meaningful share of the time they’ll take the exit instead of pushing through it. The interruption doesn’t need to be punishing to work. It mostly needs to exist, break the automaticity, and get out of the way once attention has actually landed somewhere else.

This lines up with why the category of tool that just interrupts rather than blocks tends to have better staying power than a hard denial does. It also lines up with the discipline-versus-design distinction we’ve written about before: the fix that actually works isn’t a stronger wall or more resolve in the moment — it’s something external that reliably breaks the loop’s third beat before the routine finishes running.

What this means if you’re trying to change a habit, not just block a site

A few things follow directly from treating this as a loop instead of a single bad habit-shaped monolith:

• Identify the actual cue, not just the site. If it’s boredom at 3 p.m., blocking the site doesn’t touch 3 p.m. boredom — it just relocates where it goes looking for relief.
• Don’t expect denial alone to survive contact with a strong cue. Ability is the easiest lever to route around; Motivation and Prompt are doing most of the real work in the loop.
• Insert a genuine pause, not a wall. A short, finite interruption gives the slower, deliberate part of your brain a chance to weigh in before the routine finishes on autopilot — that’s the mechanism, not just a nicer-feeling version of blocking.
• Expect the loop to weaken gradually, not disappear on day one. Habit research treats these loops as things that fade with disrupted repetition, not switches that flip off the first time you interrupt them.

None of this means blockers are useless for everyone — for some people, at some moments, total denial is exactly the blunt tool that’s needed. But for the loop as habit researchers actually describe it, the more durable fix tends to be the one that meets the cue with a pause instead of meeting the site with a lock.