By Dr. David Gorstein, Partner & Population Health Practice Leader
WHOOP and Oura command astronomical valuations on the promise that they’ll become the operating system for health and wellness data—integrated with healthcare in ways that are insightful and actionable.
From personal experience, I know that they can’t even make that true in the narrow, well‑defined use case of training data.
Warning shot across the bow (today’s press release from Training Peaks):
“Health metrics from Oura Ring now integrate directly into TrainingPeaks, giving you a full-circle view of your training and life. Connect your Oura Ring to see sleep, HRV, stress scores and more populate in your Health Insights dashboard.”
So far, most of these partnerships are effectively brand events: “X now integrates with Y” but integration depth is extremely thin. It’s just data appearing on another surface, vs a true synthesis layer that changes recommendations, training plans, or clinical decisions.
Training is hard enough. Tracking it shouldn’t be.
Right now there’s a structural gap in endurance + healthspan tech:
- Training platforms (e.g., TrainingPeaks) own the structured work: power files, TSS/IF, planned sessions, coach workflows.
- Wearables (e.g., WHOOP, Oura) own the continuous physiology: sleep, HRV, strain, illness, meds, travel, labs.
The real “operating system” would simply fuse those two. That’s where the value is.
If a device can’t even ingest power and basic workout metadata from where serious athletes and coaches already live, it’s not an OS. It’s a very good sensor feeding someone else’s OS.
That has two big consequences:
- Coaching: Without structured training data, any “AI coach” is guessing about intent. It sees the strain, but not the plan.
- Strategy: If TrainingPeaks (or someone else) pulls in wearables’ health data while wearables don’t pull in structured work, the platform becomes the hub and the devices become interchangeable inputs.
Taking it to the next level of insight for Whoop or Oura will require:
- Ingesting TrainingPeaks power + planned/completed workouts and fusing that with strain and healthspan.
- Understanding why serious coached athletes already treat TP as the OS and wearables as inputs.
- Defining what has to be true for “OS of performance and health” to be more than a pitch‑deck slogan.
To justify their valuations and investor expectations, WHOOP and Oura are already expanding beyond elite performance into healthspan, chronic conditions, preventive care, labs, and eventually deeper healthcare workflows, but If you can’t tie simple power‑based training load to autonomic response today, you’re not going to be the layer that turns blood + omic data into “next best action” tomorrow. You’ll be another point solution in someone else’s care stack.
Who will emerge?
Every company wants to be the system of record / OS, so everyone gatekeeps their data. Strava talked about being the OS for the athlete years ago (wanting to pull in sleep data from Oura/WHOOP), but no one let them so that vision never came about. Strava’s own API access is also heavily limited. WHOOP has historically been selective with partnerships because they believe data is their primary asset. All players want the context, but nobody really wants to give up control of the user graph or the data layer.
WHOOP is trying to get deeper insight and give better recommendations using their AI coach, which is actually pretty cool. To know more about the person: what they’re training for, what their plan is, what their health concerns are, etc, is derived from asking the user directly. It’s not optimal for companies to collect that information manually (AI chat) or build native workout features (ie, Strength trainer in WHOOP, but there are MUCH better workout apps)
That means the “full-context OS” probably does not happen through normal integrations. It likely only happens through acquisition, otherwise, the incentives are too misaligned. I think it’s not just a product strategy problem but also has to do with the current competitive dynamics in this market and the need to justify these crazy valuations. But so far the integrations and new insights remain half-baked.
If you’re building in this space and see it differently, I’d genuinely like to hear why.