In August 2018 the FDA allowed Natural Cycles to be marketed as the first app for contraception in the US. It reviewed data from 15,570 women, with failure rates of 1.8% a year with perfect use and 6.5% with typical use.1 Earlier that year a Stockholm hospital had reported 37 unwanted pregnancies among women who said they used the app, and Sweden's Medical Products Agency opened an inquiry.2 In September it concluded that the pregnancies matched the app's stated typical-use effectiveness and asked the company to state the risk of pregnancy clearly in its instructions. Natural Cycles had already made the change.3
How predictable is a cycle, really?
Much of what femtech users misread starts with a number they were taught at school. In 2019 researchers from University College London and Natural Cycles analyzed 612,613 cycles from 124,648 app users. The average cycle was 29.3 days, and only 13% lasted exactly 28 days. Most of the difference in length came from when ovulation happened: the phase before ovulation averaged 16.9 days but ranged from 10 to 30. The authors concluded that the belief that ovulation consistently happens on day 14 is wrong.4 The study was funded by Natural Cycles, and the users were not fully representative of the population, but the scale is hard to dismiss.
A prediction built on data like that will move, and people who expected a calendar experience the movement as a fault. Someone qualified has to explain that the movement is normal and what it means for the person in front of them.
Where does the gap between perfect and typical use come from?
Typical-use failure in the first year, against perfect use
RAPS, reporting the FDA decision, 2018; CDC
Every method has a version of the gap. The CDC puts typical-use failure at 7% a year for the pill and 13% for male condoms, and between 2% and 23% for fertility awareness methods, depending on the method and how it is used.5 The spread comes from behavior: what people understand about fertile days, what they do when a reading is unclear and how consistently they follow the instructions. The cheapest way to narrow it is a clear explanation from someone qualified, repeated often enough to stick.
Who explains menopause, if doctors weren't taught it?
Menopause apps have grown into a gap that medicine left open. In a survey published in Mayo Clinic Proceedings in 2019, 177 residents in family medicine, internal medicine and gynecology at 20 US programs were asked about menopause care. Only 6.8% said they felt adequately prepared to manage it, 20.3% reported no menopause lectures during residency, and 93.8% said training in it was important.6 Women who can't get answers from their own doctor go looking elsewhere, and a brand that offers a named menopause specialist answering real questions is filling a need the health system has acknowledged.
That brings responsibility. The specialist has to be clear about what can be answered in a group session and what needs a personal consultation, and the brand has to resist turning the session into a sales pitch for its own product.
What makes people willing to ask?
Femtech also asks for more sensitive data than almost any other category. In 2021 the FTC settled with Flo Health, a period and fertility app used by more than 100 million people, over allegations that it shared users' health information with marketing and analytics companies including Facebook and Google despite promising privacy.7 In August 2025 a California jury found that Meta had violated the state's privacy law by collecting data from Flo users without consent.8
People will only ask about bleeding, sex, fertility or symptoms they find embarrassing in a setting they trust with the answer. Survey research has shown for decades that people disclose more when they answer privately than when they answer to a person.9 In a live session, anonymous questions are the private answer sheet.
- Make anonymous questions the default, and say so before anyone types.
- Record consent before anyone joins, with a plain explanation of what is recorded and who sees it.
- Pair a clinician with a moderator. The moderator answers practical questions privately so the clinician can focus on the ones the whole room needs.
- Explain limits as clearly as features: what a prediction means, what it doesn't, and when to see a doctor.
Adelo Insights · Health tech and the expert voice
