What 50,000 Users Taught Us About Breathing Practice

Updated June 2026

Most wellness content recycles the same five studies. This post does something different: it reports what we observed across more than 50,000 people who downloaded Vayu, including the numbers that do not flatter us.

We are publishing this because aggregate behavioral data on breathing practice barely exists in public. Researchers study small cohorts over weeks. App companies sit on usage data and publish marketing claims. The honest middle, what real people actually do with a breathing app over months, is missing. Here it is.

How many people stick with a breathing practice?

Far fewer than the wellness industry implies, and far more than cynics expect, depending entirely on which group you measure.

Here is our full retention picture, with nothing hidden:

These three numbers describe the same product. The difference is commitment level. A casual installer who downloaded the app after seeing a video keeps practicing about 6% of the time. Someone who joined a structured study with check-ins stayed at 64%. Someone who paid stays at 92.5%.

The lesson is not "our app retains badly" or "our app retains brilliantly." The lesson is that breathing practice behaves like exercise: intent predicts persistence. Free installs are curiosity. Payment and structure are commitment devices, and commitment devices work.

Industry comparisons usually cite consumer app retention averages in the low single digits at day 30, which makes our broad install cohort typical and our paying cohort unusual. We think the paying number is the one that matters for anyone deciding whether a breathing practice can become durable: when people opt in with real intent, it sticks.

How often do committed users actually practice?

Active Vayu users average about 2.2 sessions per week, and 25% of them complete 3 or more sessions per week.

That 25% figure deserves context. Industry engagement benchmarks for mindfulness apps put the share of users hitting 3+ sessions weekly around 8%. Our active users clear that at roughly three times the rate (Supabase session data, 4-week window, n=498 user-weeks).

Why the gap? Our working hypothesis, supported by what users tell us, is session length. A guided meditation asks for 10 to 20 minutes. A breathing session delivers a felt physiological shift in 3 to 5 minutes. Lower time cost, faster feedback, more repetition. If you want the mechanism behind that fast feedback, our explainer on the physiological sigh covers why certain breath patterns shift state within a minute or two.

What physiological changes show up in the data?

In our naturalistic study of 199 users, currently under peer review at SAGE Journals, regular practice tracked with three measurable changes:

Adherence inside the study was 71% at 3+ sessions per week, much higher than our general population, which is what "coached cohort" means in practice.

Two honest caveats. First, this is a naturalistic study, not a randomized controlled trial: participants chose to practice, so motivated people are overrepresented. Second, the paper is under review, not yet published, and we will update this post when it lands. If HRV is new to you, start with what HRV is and why it matters and then how to improve HRV with breathing exercises.

Which techniques do people return to?

The patterns people repeat are the short, structured ones with a clear finish line.

Across sessions, the durable favorites in our library map closely to the techniques with the strongest evidence base: exhale-weighted patterns like cyclic sighing, simple symmetric structures like box breathing, and pre-sleep patterns like 4-7-8 breathing. Open-ended unguided breathing, by contrast, rarely becomes a habit. People keep what has a shape.

This matches the broader behavior-change literature: friction kills habits, structure builds them. A technique with a defined start, a defined pattern, and a defined end is easier to repeat at 11pm on a Tuesday than an instruction to "breathe mindfully for a while."

What kills a breathing habit?

Day three to day seven. That is the cliff in our data, and almost everyone who churns goes over it there.

The first session is novelty. The second is confirmation. The drop happens when the novelty is spent and the habit loop is not yet formed. Users who cross roughly the two-week line look very different: their week-over-week persistence climbs sharply, which is consistent with the paying-subscriber retention figures above.

The practical takeaway if you are starting: do not aim for a 20-minute daily practice. Aim for three short sessions in your first week, attached to an existing routine (after coffee, before opening the laptop, lights out). Small and anchored beats ambitious and floating.

What we still do not know

Honest data posts admit their gaps, so here are ours:

• Causality. Our study is naturalistic. People who breathe regularly may differ from people who do not in ways we cannot fully control for.
• Long-horizon outcomes. We can see months of behavior, not years of health outcomes.
• Self-selection. Everyone in our data chose to download a breathwork app. Population-level claims need population-level samples.

We would rather publish numbers with caveats than claims without sources. That standard is also why you will not find a "breathwork cures anxiety" headline anywhere on this site; for what the clinical literature does support, see the broader evidence reviewed across our HRV and technique posts.

FAQ

How long does it take to build a breathing habit? In our data, users who keep practicing past the first two weeks show sharply better persistence afterward. Plan for two deliberate weeks; after that it gets easier.

What is a realistic breathing practice frequency? Active users average about 2.2 sessions per week. The study cohort showing HRV and stress improvements practiced 3+ times weekly. Three short sessions a week is a realistic, evidence-aligned target.

Does breathwork actually change anything measurable? In our naturalistic study (n=199, under peer review at SAGE Journals), regular practice tracked with a 28.6% HRV increase (p < 0.001), a 7.4 bpm drop in resting heart rate, and a 16% decrease in perceived stress (p < 0.05).

Do most people quit breathing apps? Most casual installers do, across every app category. The picture flips with commitment: 92.5% of Vayu's paying subscribers are still active at day 30.

What is the best breathing technique to start with? The ones people actually stick with are short and structured. Box breathing for focus, cyclic sighing for quick stress relief, 4-7-8 before bed.

Vayu is a breathwork app that pairs live biometric feedback from the watch you already own with adaptive breath guidance. More than 60,000 downloads to date. If you want your own numbers in next year's version of this post, try Vayu free.

Editor note for Dhruv (delete before publish): the brief asked for session timing patterns (time-of-day distribution). That data is not in FACTS.md and I will not invent it. If you want a timing section, Atlas can pull the hour-of-day histogram from ios_app_sessions and I will add it with real numbers. Also: the day-3-to-7 churn cliff and two-week persistence claims are directional from our retention curves; if you want them hard-numbered, same Atlas pull.