For the last decade, consumer health technology has been organized around measurement: Steps. Calories. Rings closed. Resting heart rate. HRV. Sleep stages. Strain. Recovery. VO₂ max. Body temperature. Blood oxygen. Training load.
The problem is not that these metrics are useless, but rather that consumers are drowning in biometric fragments. The more advanced wearables have become, the more they create a cognitive burden to the user. It can be overwhelming for general users to make sense of all the data without a team of coaches, nutritionists or doctors available to help them interpret it.
Google’s newly announced health ecosystem is an attempt to change that. The Fitbit app is becoming the Google Health app. This will be paired with Google Health Coach, a Gemini-powered analysis layer. Fitbit Air is launching as a low-cost, screenless wearable designed for continuous passive sensing. Together, these products suggest a strategic shift: the future of consumer health may not be about who builds the most impressive watch, ring, or band, but who can turn continuous health data into useful, personalized, trustworthy guidance.
The hardware is just the sensor layer. The app, the AI, and the service are becoming the product. This will change our healthcare system; to what extent remains to be seen.
What Google Announced
Google is effectively reorganizing its consumer health strategy around three connected pieces: a rebranded app, an AI coach, and a new screenless tracker.
Starting May 19, 2026, the Fitbit app is becoming the Google Health app. Google describes the new app as a centralized place to track fitness, sleep, wellness trends, medical records, and data from connected apps. The redesigned app is organized around four major tabs: Today, Fitness, Sleep, and Health. It is meant to preserve Fitbit’s consumer tracking DNA while expanding into something broader: a personal health dashboard that can integrate medical records, third-party app data, and AI-generated insights.
The move marks a transition from Fitbit as a device company to Google Health as a software and services platform. Google also says the new app can connect with other health and fitness apps (via Health Connect, Apple Health, and Google Health APIs), sync medical records in the U.S., and bring more of a user’s health data into one place. Reportedly, health and wellness data will NOT be used for Google Ads.
The second piece is the wider launch of Google Health Coach. Google Health Coach is becoming available globally as part of Google Health Premium, which replaces Fitbit Premium. Google says premium plans start at $9.99 per month or $99 per year. The service will also be bundled into Google AI Pro and Ultra tiers. Google Health Coach is positioned as a 24/7 personal health and wellness advisor. It is designed to help with fitness plans, sleep improvement, recovery, nutrition, proactive insights, and personalized questions about a user’s own data. Google describes it as a coach that can connects all the dots and reduces cognitive load.
Finally, they also previewed Fitbit Air, which is described as its smallest and most affordable tracker. A screenless wearable designed for comfortable 24/7 health monitoring. It starts at $99.99, includes a three-month Google Health Premium trial, and is designed specifically to pair with the Google Health app. Its positioned for continuous and seamless tracking, sleep, and fitness insights, with a week-long battery life rather than smartwatch-style interactivity.
More About the AI
Google Health Coach is built with Gemini, but it should not be thought of as a generic chatbot dropped into a fitness app. Google suggests something more ambitious: a personal health AI system designed to reason across longitudinal wearable data, user goals, behavior patterns, and health context. The company’s research blog describes a system meant to synthesize health data, generate coaching insights, and adapt over time.
A general chatbot can answer, “How can I sleep better?” But a useful health coach should be able to answer a different question: “Why was my sleep worse this week, and what should I do differently tonight given my workouts, stress, resting heart rate, and usual sleep pattern?”
That requires several distinct capabilities:
- Longitudinal memory. A useful coach has to know your baseline. A resting heart rate of 62 may be normal for one person and elevated for another. An HRV of 35 may be concerning in one context and stable in another. A single sleep score is less meaningful than a pattern.
- Multimodal data interpretation. Health behavior is not captured by one sensor. Sleep interacts with training load. Training load interacts with stress. Stress interacts with nutrition, alcohol, illness, menstrual cycle, travel, and work schedule. A coach that only sees steps is not a health coach.
- Agentic reasoning. Google has described a multi-agent approach involving a conversational agent, a data science agent, and domain expert agents. In plain English, that means the system is not just generating text. It is meant to retrieve the right data, analyze it, and then generate a response informed by health and fitness expertise.
- Behavioral translation. The end product cannot simply be “your HRV is lower than usual.” The value is in turning that observation into a recommendation: take a lower-intensity day, move your workout earlier, prioritize sleep consistency, reduce alcohol, add recovery, or monitor for signs of illness.
This is why Google Health Coach is important. It is not only a new feature. It represents a shift from data presentation to data interpretation.
Google also emphasizes that the coach is informed by expert input and science-backed guidance. The consumer-facing Google Store page describes the coach as built with Gemini, able to remember goals, learn from progress, proactively check in, answer personalized health questions, create adaptive fitness plans, and provide detailed sleep insights.
For clinicians, that shift is worth watching closely. Patients may increasingly arrive not only with Apple Watch ECG tracings or sleep scores, but with AI-generated interpretations of their health trends. Some will be useful. Some will be overconfident. Some will blur the boundary between wellness coaching and medical guidance. The consumer health AI layer is coming either way. The question is whether it becomes trustworthy enough to help people — and constrained enough not to mislead them.
Competitive Landscape
WHOOP has been ahead of the game because it understood earlier than most that the future of wearables was not really about displaying more data but about turning continuous biometric tracking into a coaching relationship. That broader product philosophy appears to be where Google is now taking Fitbit. Their core stack is recovery scores, strain targets, and sleep coaching with proactive guidance built around interpretation, not just data display. They have adapted their own AI layers overtime (WHOOP Coach) and recently announced two new AI features. “My Memory,” which gives members control over their persistent coaching context, and “Proactive Check-Ins,” which surface personalized recommendations without requiring manual input.
Interestingly, on the heels of Google’s big announcement, WHOOP released it’s own big news. In a pivot away from AI, they will make available on-demand clinician video consultations for US users, starting with a comprehensive review of your continuous biometric data and, when available, blood work and medical history via HealthEx records integration. WHOOP is valued at $10.1 billion after closing a $575 million funding round in March.
Apple’s position is both dominant in hardware and conspicuously underdeveloped in AI health services. Apple controls roughly 60% of all smartwatch revenue according to IDC data, and yet the Health app still mostly gives access to raw data and leaves the interpretation to you. However, the company is working on a service internally codenamed “Mulberry” and publicly referred to as Health+, reportedly featuring an AI coach trained on data with input from staff physicians. Reports had it arriving with iOS 19.4 in spring or summer 2026 but this appears to have been delayed which may be a big benefit competitively for Google Health.
Hume occupies a more focused niche. Hume Health’s Band is positioned around longevity and metabolic health rather than fitness performance. Hume’s Body Pod is an at-home body composition device that tracks body fat, muscle mass, visceral fat, metabolic age, water composition, and many other body metrics. Hume frames the product around understanding whether weight change reflects fat loss, muscle preservation, hydration shifts. In the GLP-1 era, “weight” is no longer enough and consumers increasingly want to know whether they are losing fat or lean mass, improving metabolic health, or simply dropping water weight.
The broader field is crowded. Samsung, Oura, and Garmin have all added AI-generated coaching and observations to their apps. Microsoft launched Copilot Health in March, connecting wearable data and health records. OpenAI launched ChatGPT Health in January. Everyone is converging on the same product: an AI that synthesizes your full health picture and tells you what to do about it. The differentiators are going to be AI quality, depth of data integration, safety guardrails, hardware, and price.
The Big Shift: Promise and Peril
Hardware is the tool. Software is the product. AI is the service.
For years, the wearable market was hardware-led. Companies competed on industrial design, battery life, GPS accuracy, sensor quality, screen brightness, and form factor. Those features still matter, but they are becoming less differentiating. Most serious consumer devices can now track heart rate, sleep, activity, temperature trends, recovery, and exercise with enough fidelity to be useful. The harder problem is no longer measurement. It is meaning. Consumers do not want another chart showing that their HRV declined or their sleep score dropped. They want to know why it happened, whether it matters, and what they should do differently today. Google Health Coach, WHOOP Coach, Oura Advisor, Apple’s increasingly proactive health features, and Hume’s body-composition guidance are all different expressions of the same shift.
This also changes the business model. Hardware revenue is episodic; coaching revenue is recurring. Subscription services (Google Health Premium) are evidence that consumer health companies increasingly see software, analytics, and personalized guidance as a durable source of value. The device is more of a low-friction entry point into an ongoing health relationship.
This may be disruptive for traditional healthcare. Hospitals, EHR vendors, remote patient monitoring companies, and digital therapeutics firms have spent years trying to build patient engagement tools, but many remain trapped inside clinical workflows, reimbursement constraints, portals, and episodic care models. Consumer tech companies operate differently. They live in our daily habits: phones, watches, rings, apps, notifications, and subscriptions.
The boundary between wellness and medicine will continue to blur. Google can describe Health Coach as wellness. WHOOP can frame blood pressure features as insights rather than diagnosis. Oura can position its guidance around readiness and resilience. But consumers do not experience their bodies in regulatory categories. Sleep, stress, exercise, weight loss, medications, blood pressure, labs, arrhythmia alerts, and symptoms all belong to the same lived experience. As these platforms ingest more wearable signals, medical records, nutrition data, labs, and user-reported context, they will increasingly shape when people seek care, what questions they ask clinicians, and how they understand their own health risks.
The direction is unmistakable. We are moving from activity tracking to health interpretation, from wearables to health services, and from dashboards to coaches. This transformation is not being led by traditional medical device companies, hospitals or clinicians, but by big consumer tech. That creates both promise and danger. A good AI coach could help people detect patterns earlier, personalize behavior change, and make prevention more actionable. It could explain that rising resting heart rate, declining HRV, and poor sleep may reflect overtraining or early illness. It could help someone on a GLP-1 understand whether weight loss is coming at the expense of lean mass. But the same systems could also over-interpret noisy data, create anxiety, or misdiagnose potentially serious conditions. A misleading recommendation in this context is different from a bad product suggestion or a flawed travel itinerary, and it is something we will have to adapt to and learn how to regulate.
Follow us on Substack:
To ensure timely access to posts and future updates, be sure to subscribe to This week in Cardiovascular AI – an email newsletter for followers and subscribers that summarizes recent research in Cardiovascular AI.
Cardio Cypher 