What WHOOP’s Data Can Tell You—and What Will Ahmed’s Interview Leaves Unproven

1. Guest Background

Will Ahmed is the chief executive of WHOOP, the wearable technology company established in 2012 to help users improve performance and health. Its offering combines a wearable device with membership software for sleep, recovery, stress and behavioral tracking. In this approximately 56-minute All-In interview, Jason asks Ahmed to explain how that offering became a subscription business, what its growing collection of health data might enable, and where its ambitions encounter regulatory boundaries.

Ahmed supplies a relevant piece of his own background: while studying at Harvard, he read medical research and became interested in heart rate variability. He recalls seeing an opportunity to measure a signal continuously that existing research had made interesting to him. He presents that opportunity as part of WHOOP’s origin. The account helps explain his emphasis on longitudinal observation; it does not establish clinical credentials or validate the company’s subsequent health claims.

His experience with athletes also informs the discussion. Ahmed announces LeBron James as a global ambassador and says James had already used WHOOP for eleven years, having been among its first hundred wearers. He distinguishes that earlier relationship from later commercial arrangements involving investment and rights to an athlete’s name, image and likeness. For a device intended to be worn around the clock, he argues, authentic adoption mattered before paid promotion. Jason’s enthusiastic user perspective makes the interview concrete, while Ahmed’s answers remain the account of the executive building and selling the service.

2. What the Episode Covers

The business story begins with a mismatch between adoption and use. Ahmed recalls launching a consumer product in 2017 for a one-time payment of $500. Buyers wore it for a long time, but too few people bought it. WHOOP lowered the entry barrier through subscriptions, a change he associates with growth in 2018 and 2019. He puts the interview-time membership range at $15–$30 a month. His growth account is striking but requires careful wording: roughly twelve years to reach one million members, another eighteen months to reach two million, and about six months to reach three million. Four million is his expectation for year-end, rather than a confirmed current total.

The product organizes its information around sleep, recovery and strain. Sleep includes consistency, actual sleep versus time in bed, and restorative stages such as slow-wave and REM sleep. Recovery is expressed as a percentage with a red, yellow or green presentation. A separate Health Monitor exposes respiratory rate, resting heart rate, skin temperature, blood oxygen and HRV. Ahmed also lists ECG monitoring, AFib detection, Advanced Labs bloodwork, Healthspan and Blood Pressure Insights as capabilities added for members using hardware already in the market.

Jason’s demonstrations show how the service tries to move beyond presenting scores. He describes an AI coach using his Japan flight details to suggest sleeping during the first half of the journey, avoiding immediate sleep on arrival and adjusting caffeine timing. Ahmed confirms that WHOOP formed a team specifically for travel. Jason also describes requesting training around avoiding injury and skiing better. Journal supplies another layer of context: Ahmed says it supports more than a hundred behaviors, lifestyle choices, supplements, drugs and emotional states. He reports daily journaling by more than half of members, about eight app openings a day and AI interaction by 80% of the membership.

Medication tracking makes the planning workflow especially visible. Jason says he lost 43 pounds while using a GLP drug, moving from 213 to 170 pounds, and now uses a maintenance dose. He also says he was not fully using WHOOP during most of that journey. His displayed plan includes 9,800 steps, two strength activities a week, daily hydration, protein tracking and medication check-ins. The AI asked questions, and he confirmed the plan. A 300-day streak and top-one-percent rankings for some loaded walking activities demonstrate additional engagement mechanisms.

The conversation later broadens to positioning and access. Ahmed acknowledges Apple Watch’s success as a multipurpose smartwatch while arguing that screenless wearables are difficult to build; no comparative market evidence establishes WHOOP’s advantage. Athlete adoption also does not guarantee permission during competition. He says the NBA was still considering the issue and describes Australian Open players being asked to remove WHOOP despite prior ITF permission, followed by other majors expressing support for wearables.

3. Core Views: Reasoning, Examples, and Limits

Ahmed’s strongest business argument concerns incentives. Recurring payments give WHOOP a reason to keep improving the experience of people who already own the device. He connects that obligation to engagement, behavioral change and health improvement, and contrasts it with a hardware business’s incentive to generate another sale through another release. In his account, hardware should change when it offers a substantial improvement, while the existing membership keeps gaining capabilities. This is an explanation of how the model can encourage continued service. It is not evidence that a subscription necessarily produces better health or that every added feature delivers its intended benefit.

The corresponding health argument starts with a personal baseline. Ahmed describes HRV as variation in the intervals between successive heartbeats and says WHOOP uses it in stress monitoring and recovery scoring. His emphasis is on how an individual differs from their own established range. Jason illustrates the appeal: after two days of unusual respiratory rate and skin temperature readings, he later developed chest pain and sought emergency care. He says testing indicated lung inflammation rather than a heart attack. He also recalls abnormal readings prompting a COVID test. These stories make continuous observation meaningful without establishing that the wearable identified either condition. The readings, symptoms and subsequent medical investigation played distinct roles.

Ahmed’s cancer anecdotes make that distinction more consequential. He describes a baseball player whose sleeping heart rate rose from a usual level around 50 to 120–130 beats per minute before testing revealed stage-four Hodgkin lymphoma. Another longtime user’s HRV reportedly fell from around 100 to 50 before testicular cancer was discovered. These are accounts of unusual changes followed by diagnoses. The interview does not show how often similar changes occurred without those diseases, or how often affected people had no such warning. Without those denominators, the stories cannot establish screening sensitivity, false-positive rates or disease specificity. Their evidentiary value is in suggesting research questions and illustrating attention to change.

The data argument is more substantial when understood as a chain of inputs. Ahmed reports a fourteen-day battery, charging while worn, average use of 23.6 hours a day and more than fifty billion hours of continuous physiological data. Those are executive disclosures. The design choices explain how WHOOP seeks to reduce interruptions, but accumulated hours do not themselves demonstrate predictive accuracy. Ahmed then describes developing internal models with open-source models and pairing physiological signals with blood tests and medical records. Those additional sources could supply reference information about what a signal means. His discussion of inferring an out-of-range biomarker remains exploratory: cortisol is an example of a possible targeted testing prompt, not a demonstrated substitute for measuring cortisol. No prediction-performance figures are supplied.

The Journal demonstration reveals the same problem at an individual scale. Jason thinks roughly ten observations, including five yeses and five noes, can generate a behavioral insight. He shows a positive 7% association between warm baths and his recovery score; Ahmed shows a negative 9% association for his own alcohol use. Such feedback can motivate closer observation, but the percentages do not establish an intervention effect. The vitamin D exchange is the interview’s clearest corrective. When Jason sees a negative 3% result, Ahmed proposes travel as a possible confounder. The AI explicitly distinguishes correlation from causation and notes that vitamin D had not been logged during the recent fourteen-day period. A persuasive interface number still depends on whose observations it summarizes, when they occurred and what else was happening.

Blood pressure adds a regulatory dimension without resolving the measurement question. Ahmed says 70,000 volunteers uploaded daily readings over several weeks, allowing WHOOP to pair those readings with existing physiological data. He describes a goal of accuracy comparable to a medical device, but gives no error distribution or independent validation results. He separately explains WHOOP’s wellness rationale, recounts an FDA warning letter, and says subsequent guidance supported wellness applications for blood pressure while BPI remained available. His belief that WHOOP’s case and industry feedback helped produce that change is his interpretation. His statement that the ECG feature received 510(k) clearance describes a different pathway. Intended use, regulatory status and measurement performance require separate evidence.

Even the financial discussion benefits from this discipline. Jason’s roughly billion-dollar revenue calculation uses four million members and an assumed $250 annual payment; his reference to a roughly $10 billion valuation belongs to that discussion, rather than an Ahmed revenue disclosure. Ahmed’s promise that WHOOP sells memberships rather than user data is similarly important as a stated trust commitment, while the interview does not independently audit its implementation. Across these subjects, the analysis becomes stronger by preserving the source and scope of each claim.

4. Learning and Application

For a product team, the subscription lesson begins with diagnosis. WHOOP’s reported starting point was sustained use among buyers alongside inadequate sales. Before adopting a similar model, a team could examine whether an entry price is obstructing demand, whether existing users actually return, and what continuing value would justify continuing payment. Track engagement separately from the task users want completed. Ahmed’s incentive argument makes ongoing improvement a business responsibility, but neither app openings nor subscription growth alone proves that responsibility has been fulfilled. The account also cannot isolate pricing from every other influence on growth.

For users, the coach demonstrations suggest a practical way to request useful help: provide a goal and the circumstances that constrain it. Jason supplied flight timing and a skiing objective, giving the system something more concrete than a request to improve health. A proposed schedule can then be reviewed for feasibility and accepted deliberately. His GLP workflow included questions and personal confirmation. The transferable idea is that interaction process; his 9,800-step target and twice-weekly strength activities remain his displayed plan, rather than a standard for every person using the medication.

Behavioral insights can be treated as prompts for better observation. When a result resembles Jason’s warm-bath percentage, retain the person, outcome and observation period before interpreting it. Check whether travel, training changes or missing entries might explain part of the association. Choosing one manageable behavior to record consistently can make later interpretation easier than changing many habits simultaneously, although it takes time and still cannot eliminate confounding. The vitamin D exchange suggests a useful follow-up question for an AI coach: what alternative explanations fit this result? That question can improve scrutiny without turning a journal into a controlled experiment.

Medication-related observations need another boundary. Jason’s weight loss is a personal account, and he acknowledges limited WHOOP use during much of it. Ahmed’s claim that GLP-1 weight loss is typically half fat and half muscle is not accompanied by a study or conditions in the interview; creatine enters as anecdotal experience. The supported product application is contextual tracking and a reviewable behavior plan. The discussion does not establish supplement effects, justify medication adjustments or reveal an individual’s body-composition change. A 300-day streak documents repeated recording, not treatment effectiveness.

Teams evaluating a health algorithm can separate technical questions from use questions. Ask how reference readings were obtained, what errors occurred, which populations were tested and whether independent validation exists. Then ask what decision the output is intended to support and what regulatory pathway applies to that specific function. Ahmed’s 70,000-person study describes a recruitment and data-pairing resource, not answers to every performance question. His account of guidance changes should likewise not become a blanket conclusion about the regulatory status of all blood-pressure products. The ECG comparison reinforces the need to inspect functions individually.

The hardware discussion offers a related ownership decision. Jason wants WHOOP-branded cuffs and scales; Ahmed describes comparing building, partnering and acquiring, while using a relatively commoditized cuff for initial calibration and ultimately hoping to replace it. A team can ask which component it must own to achieve its central goal. External components may reduce the immediate scope of development, while creating integration and control tradeoffs. This strategic logic does not prove that the proposed cuffless measurement performs adequately.

The closing discussion of private-company transactions supplies a final verification habit. Ahmed recommends obtaining the target company CFO’s support and accepting opposition. Jason recalls a broker going silent after he requested the actual seller’s identity and offered to coordinate with the CFO. Before treating a quote as an actionable opportunity, establish who is selling, where the shares come from and whether company coordination is possible. CFO support does not settle every transaction condition, and silence alone does not prove fraud. The useful boundary is narrower: an apparently specific price does not establish that the advertised transaction exists or can be completed.

Source

• Original episode: WHOOP CEO, Will Ahmed: 50B Hours of Health Data, a 4 Million Member Data Moat, & the FDA Fight