Why people drop out of quit-porn programmes

Most people who start an online self-help programme for porn use don’t finish it. In the only randomised trial of one, 89.4% of the intervention arm dropped out, and the large effects it reported were measured on the third who stayed. For anyone choosing a tool, the dropout rate is the more useful finding.

What did the Hands-off trial find?

Large effects among those who finished, with no therapist involved.

Bőthe and colleagues published Hands-off in the Journal of Behavioral Addictions in 2021, open access. It had two arms, 264 people and a six-week web-based self-help programme. On problematic pornography use, d = 1.32. On how often people used it, d = 1.65. Self-perceived addiction d = 0.85, craving d = 0.40, confidence in avoiding it d = 0.87.

As of 2026 it is the only randomised controlled trial of a web-based self-help tool for problematic pornography use, and the evidence that a self-directed tool can move this at all. Most citations stop at these figures.

How many people dropped out?

89.4% of the intervention arm, against 44.7% in the control arm. Overall follow-up was 34.5%. So the effect sizes were computed on roughly a third of the people who started, specifically those who stayed to the end of six weeks.

The authors call it a feasibility trial and "only a first step". Quoting d = 1.32 without the dropout misrepresents the paper, and it is done constantly.

What does the dropout rate show?

It describes the people a tool like this is for.

The participants volunteered for a research study about their own porn use, which already signals high intent, and got a structured six-week programme. Nine in ten in the intervention arm stopped showing up. The shortage was not information or a plan. It was a plan that holds up on an ordinary weeknight.

Outside trials it looks the same. Apps come off the phone within a week. The accountability partner gets three messages and then nothing. The 30-day programme is abandoned on day 4, and the person blames themselves, when the researched version kept about one in ten. More on the early days in why you keep relapsing on day two or three.

What should a tool for this do differently?

Put less weight on programme content and more on the daily loop. If nine in ten won’t finish six weeks, a tool has to work on an ordinary night, in the ninety seconds someone will give it. That means three things.

Help during an urge is one tap away. No module, no lesson. Anything that requires reading is hard to use at 11pm.

A slip doesn’t delete the record. Making the fourth day feel like starting over is a reliable way to produce a dropout. Gardner, Lally and Wardle, in the British Journal of General Practice in 2012, found that missing an occasional opportunity "did not seriously impair the habit formation process" and that gains resumed. That research is about building simple daily routines rather than stopping something, and "did not seriously impair" is not "does not matter". It is still the best available evidence that one bad night is one bad night. A counter that resets to zero asserts the opposite, without a source.

The daily action costs almost nothing. Harkin and colleagues, pooling 138 studies and nearly 20,000 people in Psychological Bulletin in 2016, found progress monitoring promoted goal attainment at d+ = 0.40, and worked better when progress was physically recorded rather than kept in your head. That is general health behaviour, not porn. Marking a day takes about two seconds, the kind of action that survives a 90% dropout environment.

What this doesn’t show

It doesn’t show that programmes are worthless, or that this app works.

Hands-off is one feasibility trial with a small sample, and its authors say so. No such trial has been run on Monke Reset or any app in this category, and nothing here is treatment. If this is causing serious distress, a professional is a better call than an app. Is porn addiction real explains the diagnosis that does exist. A self-directed tool is the only approach with any evidence attached, that evidence is thin, and the dropout figure is the part most worth building around.

The short version

The most-quoted number from this trial is d = 1.32. The more useful one is 89.4%.

If you’ve quit four apps, that matches the trial’s result. It points toward a tool built around the ninety seconds people give it, not a course most won’t finish. What those seconds should contain is in the first sixty seconds of an urge, and what research says about the weeks after is in what happens when you quit porn.