Oura's IPO: What a $15.6 Billion Smart Ring Valuation Says About the Health Data Business

On 21 September Oura began marketing 50 million shares at $40 to $44 each. At the top of the range the smart ring company would be worth about $15.6 billion on a fully diluted basis, trading on the Nasdaq as OURA, with the first trades expected the week of 28 September [1][2].

The easy version of this story is “a ring company is worth more than Garmin’s fitness division.” The more useful version is what public investors are actually being asked to buy. Four-fifths of Oura’s revenue is still a $400 piece of hardware. The valuation rests on the other fifth: a $5.99 subscription with an 89% gross margin that 85% of paying members are still paying for a year later [3][4]. The market is being asked to price interpretation of health data as a business in its own right. That question matters to every company building on wearable data, whether or not it ever buys the stock.

The terms, and who is selling

Sources: [1][2][5].

The split in the table is the first thing to read. At the $42 midpoint, Oura itself receives roughly $570 million before fees. Existing holders receive about $1.5 billion, and more if the underwriters exercise their option [2]. That comes on top of a pre-IPO buyback in which Oura repurchased about $1.09 billion of preferred stock from seed through Series C-1 investors, retiring roughly 17% of those rounds [6].

That is not a criticism. Oura does not need much new capital: it made about $71 million of operating income in the nine months to June and holds $372 million in cash [4]. But it frames the event correctly. This IPO is mostly a liquidity event for a decade of investors, and a price-discovery event for everyone else.

What the S-1 actually shows

The headline net loss of $924 million for the nine months to June is an accounting artefact. When Oura bought back preferred stock above its book value, about $985 million was recorded as a deemed dividend and charged against income available to common shareholders. Before that charge the company earned about $61 million, against $1.5 million a year earlier [4][6][7].

The operating numbers are the story:

Sources: [3][4][7][8].

Two readings follow. The first is that the subscription is a genuinely good business. An 89% margin product that nearly every buyer adopts and most keep paying for is rare in consumer hardware, and it is why the company can price the ring near cost-plus and still grow profitably [4]. We mapped where Oura’s $5.99 sits against every other health subscription in the 2026 price map: at the bottom of a $6 to $10 band that nearly everyone who interprets health data has converged on.

The second is that the subscription is still small. $240 million over nine months is about a fifth of revenue. Hardware grew 65% and carries a margin that has already fallen from 65% blended in fiscal 2024 to 55% [4]. If ring sales slow, as they eventually do for every device, the subscription has to carry a $15 billion valuation on its own.

The risks the filing names

Several are specific to health data and worth reading closely.

Accuracy is now a disclosed legal risk. The S-1 lists a class action alleging that Oura misrepresented its ability to detect sleep stages, and cites real or perceived inaccuracies in its data among the risks to the business [3][9]. We traced where the company’s sleep accuracy claim comes from earlier this year. Once a company is public, the gap between a marketing claim and a validation study becomes a securities disclosure question as well as a consumer one.

Hardware quality costs real money. Oura recorded $84.4 million of additional warranty expense in fiscal 2025 tied to battery issues, and accrued warranty liabilities rose to $132 million [4][6].

The patent war is two-sided. Oura has used its ring-construction patents against Ultrahuman, RingConn, Circular and, more recently, Samsung. Samsung responded with its own complaint at the International Trade Commission on 12 December 2025, seeking import restrictions on Oura rings [3][10].

Distribution is concentrated. Roughly half of hardware revenue flows through retail, and two customers account for 12% and 10% of revenue [4].

The AI layer is rented. Reporting on the filing notes Oura’s dependence on external model providers for its AI features [6]. For a company whose growth story is an AI advisor on top of the ring, that is a margin and control question, not a footnote.

The multiple, and the two IPOs that came before

At the midpoint, Oura would trade at roughly 9.7 times annualised sales and about 8 times the close to $2 billion it expects for fiscal 2026. Garmin trades nearer 7 to 7.5 times; Apple nearer 12 [11]. Priced as premium consumer hardware, 11 times trailing revenue looks rich. Priced as a high-margin, highly retentive health platform that happens to use a ring as its acquisition channel, it starts to make sense [4]. The IPO is a vote on which description is true.

Two predecessors are why the question matters.

Sources: [12][13][14].

Both were hardware-first businesses whose recurring revenue could not carry the valuation once unit growth slowed. Fitbit’s subscription was optional and late. Peloton’s was compulsory but tied to a $1,500 bike whose demand collapsed after 2021. Oura’s differences are real: the attach rate is 94%, the device is $350 to $500 and replaced every few years, and the company is already profitable. Its similarity is also real: most revenue still comes from selling the device.

The three numbers that will decide how it trades

1. Membership share of revenue. It needs to climb from 20% toward something that can carry the valuation without new rings. Membership grew nearly twice as fast as hardware in the nine months to June [7]. Watch whether that gap holds after the first price change.

2. Retention by cohort, not blended. 85% weighted-average twelve-month retention is excellent, and it improved from 81% in fiscal 2023 cohorts to about 87% in recent ones [4]. Public investors will want it by cohort, by channel and after price changes. It is also a measure of a self-selected population: people who spent $350 or more on a ring. That is why a health app should not use 85% as its own benchmark, a point we develop in wellness program KPIs.

3. Revenue from someone other than the wearer. The S-1 says the platform can “support significantly larger populations as we continue to expand access, build clinical evidence, and deepen integrations with health plans, employers, and care providers” [3]. Essence Healthcare already gives Oura rings to some Medicare Advantage members [15]. Enterprise revenue changes the ceiling on the business. It also changes the evidence bar, because a health plan buys outcomes, not engagement, as the insurance telematics record shows.

What a listed Oura changes for everyone else

Public benchmarks exist now. Attach rate, twelve-month retention, daily-to-monthly usage, opens per day and gross margin by segment are audited numbers for a consumer health subscription for the first time [3][4]. Product teams can stop guessing what “good” looks like for a paid health tier. They should also stop comparing themselves with it directly. Oura’s members are 72% women, about 63% above $100,000 in household income, and more than half manage a chronic condition [3]. That is the premium half of a market that is splitting in two.

Accuracy claims get more expensive. A public company with a listed class action on sleep staging will be more careful about what its marketing says, and competitors citing Oura’s numbers should be too. For any company whose product depends on a derived score, how that score is validated is becoming a disclosure question, not only a product one.

The platform will get more careful, not more open. Public companies protect what the market values. What the market is valuing here is the interpretation layer on Oura’s own data. Oura’s June 2026 API agreement already bars aggregators from passing its data to any AI model [16], a position we examined in can you feed wearable data to an AI. Expect the terms that govern third-party access to be defended, not loosened, once they sit under a stock price.

Interpretation is where the value is, and investors now agree. The central claim of the Oura IPO is that the recurring layer on top of the sensor is worth more than the sensor. That is the same claim every health app, coaching product and data platform makes. It will be tested in public for the first time this autumn.

Where we sit

Oura is one of the four cloud connections Sahha reads, alongside WHOOP, Garmin and Dexcom, so a stronger Oura is good for the users our customers serve, and its licence terms constrain what can be built on its data. We have no position in the offering. Our view is that the IPO is healthy for the category, because it forces the question the whole industry has avoided: whether people will keep paying to have their health data interpreted. Oura’s answer, for its own premium members, is yes. Everyone building for the other half of the market still has to earn theirs.

The short version

Oura is marketing 50 million shares at $40 to $44 for a valuation of about $15.6 billion, trading as OURA from the week of 28 September. About 73% of the shares are sold by existing holders, after a $1.09 billion pre-IPO buyback. The business is profitable before a one-off accounting charge, and 80% of its revenue is still hardware. The valuation rests on a membership with 89% gross margin, 94% attach and 85% twelve-month retention. Fitbit and Peloton show what happens when recurring revenue cannot carry a hardware valuation. For everyone building on health data, the listing publishes the first real benchmarks for a health subscription, raises the cost of accuracy claims, and makes the interpretation layer the thing public markets are pricing.