Structured self-experimentation, taken seriously
COHORT turns a health intervention into a structured experiment: it runs it against a baseline and measures the result, with its success threshold fixed before you started. Built by self-experimenters, for self-experimenters. How we grade a result is public — see our methodology — and so is our privacy engineering — see What we store.
Where we are: an early-stage beta
COHORT is an early-stage beta used by a small group of invited testers. The same things every member confirms before their first session hold for everyone, always:
This is not medical advice.
COHORT is a self-experimentation tool for informational and educational purposes only. Nothing in this product — including evidence summaries, experiment templates, cohort comparisons, or AI-generated content — is medical advice, diagnosis, or treatment. Always consult a qualified physician before starting, changing, or stopping any supplement, medication, diet, or health intervention.
You experiment at your own risk.
You design and run your own experiments. You are solely responsible for what you choose to do with your body and your data. COHORT and its operators accept no liability for any health outcomes, injuries, or damages arising from your use of this product.
This is an MVP.
Features may break, change, or disappear. Data may be lost. Evidence results may be incomplete or contain errors. Cohort comparisons are based on a very small user group and have no statistical validity at this stage.
Not for medical conditions.
COHORT is not intended for diagnosing, treating, or managing any disease. If you have a medical condition, are pregnant, or take prescription medication, do not run experiments without your doctor’s supervision.
Why it exists
Self-experimentation has a credibility problem — not because n=1 can’t work, but because most of it is done without a baseline, without a pre-registered definition of success, and without honest accounting of confounders. COHORT exists to make one person’s experiment as rigorous as it can honestly be.
Get in touch
Questions, feedback, or a hole in our reasoning? notifications@cohort.study — we genuinely want to hear it.