Health and wellness
Wellness is a retention problem. Health is a compliance problem. Most products are both.
We build health and wellness products that people keep using and that regulators, clinicians, and health systems will accept. That means habit design where it is a wellness product, and safety, privacy, and interoperability the moment it crosses into clinical.
The state of play
Money is back in digital health. Trust is the thing being priced.
Funding has rebounded, but buyers, payers, and regulators have all raised the bar at once. These are the numbers that decide which health products survive that shift.
~3%
of users are still there on day 30 for the average health and fitness app. Almost everything spent acquiring them is gone within a month.
Business of Apps benchmarks, 2026
80%
of digital health products carry no clinical evidence at all. In a market where payers and health systems now check, evidence has become the moat.
Peterson Health Technology Institute, 2023
$7.8m
was the FTC penalty against a therapy app for sharing user health data with advertisers. Wellness apps outside HIPAA are not outside enforcement.
US Federal Trade Commission, 2023
$14.2bn
went into US digital health in 2025, up 35 percent, with more than half of it to AI-enabled companies. Capital is back, but concentrated in fewer, better-proven products.
Rock Health year-end report, 2026
51.9% → 38.8%
clinician burnout after thirty days using an ambient AI scribe. This is where AI in health is delivering measurable results today.
JAMA Network Open, 2025
7
AI chatbot companies placed under formal FTC inquiry over mental health safety, alongside new state laws. Shipping an AI wellbeing feature casually is now a legal risk.
US Federal Trade Commission, September 2025
Health products people trust
and keep using.
Wellness and fitness products
Habit loops, streaks, and progress built in deliberately, because in a category where almost nobody is left after a month, retention design is the product rather than a polish layer.
Telehealth and care platforms
Consultations, scheduling, records, and messaging built to work for the least confident patient in the room, and to hold up under the clinical safety standards a health system will ask about.
Mental health and wellbeing
Products designed with crisis pathways, escalation to a human, and data handling that treats mental health information with the sensitivity it actually warrants.
Remote monitoring and wearables
Device and sensor data brought together into something a clinician can act on, with alerting that surfaces what matters instead of burying it in noise.
Clinical AI with guardrails
Documentation, triage support, and summarisation built with retrieval, human oversight, and evaluation, and scoped deliberately away from claims that would make it a regulated device.
Compliance and interoperability
HIPAA, UK GDPR and NHS clinical safety, and India’s DPDP and ABDM, plus the FHIR and records integration that decides whether a provider can adopt you at all.
The current landscape
Where health products fail, and where the work is.
Five ways a promising health product stalls, and what building for them properly actually looks like.
The friction
Users disappear inside a month
The average health app keeps around three percent of its users to day thirty. Acquisition spend evaporates before a habit has a chance to form.
What we build
Onboarding, habit design, and personalisation built in from the start, because in this category the retention curve is the business model rather than a metric to fix later.
The friction
Nobody checked which side of the line they were on
A feature that interprets symptoms or recommends treatment can make a product a regulated medical device. Founders often discover this after building, when it is expensive.
What we build
We establish early whether your product is wellness or clinical in each market you are targeting, and design to that answer, because it determines the evidence, the safety case, and the entire route to market.
The friction
Health data rules that carry real penalties
Wellness apps outside HIPAA still face FTC enforcement, UK GDPR, and India’s DPDP rules. Regulators have already fined popular health apps for exactly this.
What we build
Consent architecture, data minimisation, and audit trails designed in from the first commit rather than retrofitted, which is both cheaper and far less dangerous.
The friction
AI wellbeing features are now a liability
Mental health chatbots are under formal regulatory inquiry and active litigation. A model that cannot reliably recognise a crisis is not a feature, it is exposure.
What we build
AI scoped deliberately, with human oversight, crisis detection and escalation paths, and evaluation before launch, so the feature helps without pretending to be a clinician.
The friction
No evidence, no adoption
Four in five digital health products carry no clinical evidence, and independent evaluators now say so publicly. Payers and health systems increasingly will not buy without it.
What we build
The data infrastructure to generate real-world evidence built into the product, so you can demonstrate outcomes rather than assert them when a buyer asks.
Case study · Mental health and wellbeing
A mental health platform built for people who come back.
We built MyndStories a platform designed around the hardest metric in wellness, which is whether anyone returns. Considered onboarding, content journeys that adapt to where someone is, and privacy handled carefully because mental health data deserves more care than most products give it.
- 4.1xDay 30 retention vs category average
- +58%Returning sessions per user
- 7 wksFrom kickoff to public launch
How we help you win
Four places a product firm earns its keep.
Not a feature list. The four moves that, on the evidence above, decide whether a health product gets used, gets adopted, and gets to stay.
Make retention the product
When the category baseline is roughly three percent at day thirty, onboarding and habit design are not polish. We build the loops that turn a download into a routine, which is the only thing that makes the acquisition maths work.
Know which side of the line you are on
Wellness and regulated clinical products have completely different obligations. We settle that question early and design to the answer, so you are not rebuilding after a regulator or a health system asks.
Ship AI that is safe to put near a patient
Given the current scrutiny of health AI, the responsible build has guardrails, human oversight, and evaluation. We scope AI to where it genuinely helps and away from claims it cannot support.
Ship the last thirty percent
Security, consent, clinical safety, audit trails, and records integration are where health products stall between a working demo and something a provider will actually adopt. That stretch is what we specialise in.
Fit
Whether this is right for you.
A good fit if
- You have a health or wellness product and need it built to a standard clinicians will accept
- Retention is your real problem and you want it designed for rather than patched
- You are unsure whether your product counts as a regulated medical device
- You handle sensitive health data and want privacy designed in rather than bolted on
- You want an AI feature built with guardrails rather than a chatbot pointed at patients
Probably not a fit if
- You want the cheapest build in a category where safety and privacy carry real penalties
- You need clinical or regulatory sign-off rather than a product team, which is a different profession
- The plan is to make clinical claims without the evidence to support them
- Nobody on your side can decide what the product should and should not do medically
- Data protection is treated as paperwork rather than as a design constraint
Insights
- ArticleAI in Healthcare: From Diagnostics to Predictive CareAn overview of how AI-powered insights are enabling faster diagnoses, automated workflows, and data-driven decision-making for providers. Nov 29, 2025
- ArticleThe Rise of Wellness Tech: Why Engagement Matters More Than FeaturesExploring how personalized journeys, habit loops, and gamification are driving long-term user retention in wellness and fitness apps. Nov 29, 2025
- ArticleHow Telemedicine is Redefining Patient Care in 2025A deep dive into how virtual consultations, secure communication, and AI-driven workflows are transforming modern healthcare delivery. Nov 29, 2025
Common questions
Answered before you have to ask.
Is our product a regulated medical device?
It depends on what it claims to do, and it is the first thing worth settling. Broadly, tracking, coaching, and general wellbeing content sit outside device regulation, while anything that diagnoses, triages toward a diagnosis, or recommends specific treatment usually falls inside it in the US, UK, and India alike. We work this through with you early, and where the answer is genuinely uncertain we will say so and recommend you take regulatory advice rather than guess.
How do you handle patient and health data?
As a design constraint from the first commit. Depending on your market that means HIPAA in the US, UK GDPR and NHS clinical safety standards in the UK, or India’s DPDP rules, plus consent management, data minimisation, encryption, and audit trails throughout. Wellness apps that sit outside HIPAA still face regulators, and enforcement against popular health apps has already happened, so we do not treat that as a lower bar.
Our retention is terrible. Can that actually be fixed?
Usually improved substantially, yes, though nobody should promise you a number before seeing your data. The category baseline is brutal, which means most of the opportunity sits in the first week: onboarding that gets someone to a first real result quickly, a reason to return tomorrow, and personalisation that reflects what they actually did. We instrument the retention curve from day one so the work is measured rather than assumed.
We want an AI feature for mental health. Is that safe to ship?
It can be, with the right scope and real caution. What is not safe is a general chatbot pointed at vulnerable users, which is precisely what regulators and courts are now examining. We build these with clear boundaries on what the AI will and will not do, crisis detection with a genuine escalation path to a human, and evaluation before launch. If the honest answer for your use case is that AI should not be in that position at all, we will tell you that.
Can you integrate with health records and clinical systems?
Yes, and for anything provider-facing it is usually the thing that decides whether adoption is possible at all. That means FHIR and HL7 for records, single sign-on, and in the UK the clinical safety and assessment requirements the NHS applies to digital tools. We treat it as core rather than as a later phase.
Do we own everything, and what does it cost?
You own all of it: the repository, the hosting, the data, and every credential, in your name throughout. On cost, we price after understanding the scope rather than before, because a wellness app and a clinical platform are entirely different builds. A thirty minute call at no charge, then a written scope with the work broken down and a fixed number against it.
