Willow, the wearable breast pump maker, and Oura, the ring-based sleep tracker, maintain dominant positions in categories now crowded with cheaper imitators by controlling something knockoffs cannot replicate: longitudinal user health data, according to Modern Retail. Both companies report holding category share north of 40% despite waves of lower-priced hardware entering their markets over the past three years.

The mechanism is not patent protection or trade dress. It is the closed loop between hardware, anonymous aggregate user data, and algorithm refinement. Willow collects anonymized pump session data across hundreds of thousands of users to refine pressure curves and session recommendations. Oura uses three years of nocturnal biometric data from millions of users to train its readiness and sleep staging models. Each firmware update widens the performance gap because the competitor launching today starts with zero baseline data.

This works because health hardware categories reward accuracy over price. A user does not choose the cheapest sleep ring if it misclassifies REM stages or the cheapest breast pump if the suction profile causes discomfort. According to the executives interviewed by Modern Retail, retention in both companies exceeds 85% annually, a figure driven by the perception that the product improves over time as the dataset grows. The data moat compounds: more users generate better algorithms, which attract more users.

The steal for a small physical-product brand entering a nascent category is to instrument the product from day one for anonymous usage telemetry and to communicate that the product learns. Launch the hardware with basic feature parity but emphasize that version two, three, and four will improve based on real user patterns. Publish a simple dashboard or changelog showing how aggregate data informed the latest update. The cost line is modest: a basic cloud telemetry stack runs under $200/month for the first thousand units, and a quarterly email update explaining one data-informed improvement builds the moat narrative without additional ad spend.

This does not require venture funding or a PhD team. It requires deciding early that the product generates usage data, anonymizing it responsibly, and making one improvement every quarter that you can attribute to that dataset in plain language. A small brand selling a kitchen scale for macro tracking, a posture wearable, or a hydration bottle with sensor feedback can run the identical play: the hardware is table stakes, but the learning curve is proprietary.

The broader pattern is that in any category where the product touches the body or daily routine, the brand that closes the feedback loop first owns the category long after the patent expires and the price drops.