Why Every Quit-Smoking Success Rate Is Smaller Than It Looks

A programme that helps 25 people out of every 100 who sign up can advertise a 60% success rate without anyone lying once. Not a trick, not fine print. Just a choice about who gets counted, made four times in a row.

If you're about to pay for something that promises to get you off cigarettes, this is the most useful thing to understand first — and it applies to us as much as to anyone else.

The ladder

Take a hundred people signing up for any quit programme. These numbers are illustrative, but the shape is what real ones look like.

A hundred enrol. Somewhere between forty and sixty finish it — dropping out partway is the single most common outcome in this field, for every programme ever studied. Say fifty finish.

Of those fifty, thirty answer the follow-up survey a month later. People who ignore that email are not a random sample; the ones who quit successfully are noticeably more likely to reply.

Of those thirty, eighteen say they aren't smoking.

Now pick your fraction. Eighteen of thirty is 60%. Eighteen of fifty is 36%. Eighteen of a hundred is 18%. All three are arithmetically true, none of them is a lie, and only the last one answers the question you're actually asking, which is what happens if I sign up.

That last approach has a name — intention to treat. Count everyone who enrolled, and treat anyone you lose as still smoking. It's the standard proper cessation trials use, and it exists precisely because dropping out of a quit programme usually means it didn't work.

Why the flattering number is the default

Not because everyone in this industry is dishonest. Because the flattering number is easier to produce and easy to defend in a footnote.

Anyone who finishes a paid programme, then opens a survey email, then fills it in, is the most motivated slice of a group that already chose to pay. Measuring them tells you something real about them. It tells you much less about you.

This will apply to Nicoxit too. Our completion-based number will look better than our enrolment-based number, exactly as it does for everyone. We plan to publish both and say which is which.

Three things a number has to tell you

If any one of these is missing, treat the figure as decoration rather than information.

Who was counted. Everyone who signed up, or only the finishers? This is the 60-versus-18 gap, and it's usually invisible in the marketing.

When it was measured. Thirty days after a course ends is a short window, and relapse is heavily back-loaded. Cessation research generally reports at six or twelve months for that reason. A thirty-day figure and a twelve-month figure are not comparable, and the shorter one always looks better.

How it was checked. Self-reported, or verified? A well-run trial of a smoking-cessation programme published in Addiction in 2020 measured carbon monoxide in participants' breath at 26 weeks, analysed by intention to treat across all 620 people randomised, and reported 19.4%. That's what a fully specified number looks like: a denominator, a time point, and a breath test. It is also smaller than most numbers you'll see advertised, and it means far more.

A note on badges

"Clinically proven" is not a regulated phrase in India or most other markets. It doesn't require a randomised trial, a control group, or a particular result. Usually it means a study exists — which is worth something, but not what most readers assume it means.

WHO validation is a real process, and a good signal about how carefully an app was designed. It's a review of an app's scientific grounding, not a measurement of how many people quit.

Neither is a reason to distrust a product. Both are a reason to read past the badge to the number underneath, and then to read past the number to its denominator.

What we'll publish, and what we won't

There is no success rate on this page, or anywhere else on this site. Inventing one before we have it would be worth nothing to you.

Here's the commitment we can actually keep. When we report numbers, we'll say how many people started, how many finished, how many we heard back from, and over what period — so you can compute the unflattering version yourself rather than taking ours. Where a figure is soft, self-reported or based on a small sample, we'll say so beside it. Some of what we publish will be less impressive than what you can read elsewhere. That's the trade we're making.

What we're building in the meantime

31 modules. 7 days. Then it's done. About 15 minutes a day. A course, not a tracker, covering cigarettes, vapes, hookah and IQOS.

You keep smoking through it. The first instruction is: don't quit yet. The week goes on examining what you believe the cigarette does for you, while those beliefs are still live enough to test. We don't replace the nicotine — we take apart the reason you reach for it.

Grounded in MBCT, ACT and implementation intentions, with the research cited openly. Day 1 free, tools free, community free. No streak that punishes you for slipping.

Ask who got counted. The answer tells you more than the number does.

Want a read on what's actually holding you? The 3-minute Quit Profile quiz gives you one, and your plan arrives the morning we launch.

Related reading:

• The Best Quit-Smoking Apps in India (2026)
• Allen Carr's Easyway: Does Quitting "Without Willpower" Actually Work?
• Do Nicotine Patches Actually Work?

References

1. Frings, D., Albery, I. P., Moss, A. C., Brunger, H., Burghelea, M., White, S., & Wood, K. V. (2020). Comparison of Allen Carr's Easyway programme with a specialist behavioural and pharmacological smoking cessation support service: a randomized controlled trial. Addiction, 115(6), 977–985. https://doi.org/10.1111/add.14897
2. Whittaker, R., McRobbie, H., Bullen, C., Rodgers, A., Gu, Y., & Dobson, R. (2019). Mobile phone text messaging and app-based interventions for smoking cessation. Cochrane Database of Systematic Reviews, CD006611. https://doi.org/10.1002/14651858.CD006611.pub5