Hinge Health works — for the 4% who use it

I'm their power user with the hardest ROI to prove

I started a remote job a few years ago and noticed, gradually, that I was moving a lot less. Some aches followed. Nothing alarming, but enough to make me want to take action before it escalated. I had spent time earlier in my career working on a product strategy for a medical device targeting chronic lower back pain, which left me with a healthy respect for how quickly these things compound if ignored.

Around that time, my employer held their annual benefits explanation meeting where HR walks through what’s available, what’s new, what you should probably be using but aren’t. Hinge Health came up. I enrolled that week.

A year later, I am still using it. I complete my weekly goals almost every week. Not much else in the digital health space has stayed with me the way Hinge has. I thought about what makes Hinge different. The product is well-built with great UX. It makes users feel seen in a way that most digital health products don’t bother with, or tend to be very bad at. My PT was genuinely engaged, helpful, and accessible.

After a couple months of consistent use, the aches that brought me to Hinge in the first place had resolved. But I stayed. I found a use case that was never explicitly advertised: Hinge as a preventive layer, a way to offset long hours at a desk before I require clinical attention. On busy days when I skip my workouts entirely, I still get my dopamine hit from having moved and stretched for at least those ten minutes of my daily exercise playlist.

Hinge went public in that time. Digging into the S-1 made me realize I’m the outlier user, though. Maybe even the wrong kind of user for Hinge. Among a list of accomplishments, there’s a number that Hinge buries in the filing, one that points to a structural constraint.

What Hinge actually is

Hinge Health is a digital musculoskeletal (MSK) clinic sold B2B to self-insured employers and health plans. Members get a smartphone app with guided exercise therapy, a care team (physical therapists, health coaches, and sometimes physicians), and motion-tracking technology that provides real-time feedback on exercise form. Depending on enrollment tier and condition, members may also receive the Enso, an FDA-cleared transcutaneous electrical nerve stimulator for adjunctive pain relief.

In 2024, Hinge served 2,256 employer and health plan clients, covering 49% of Fortune 100 companies. The business model is per-engaged-member: employers pay only for employees who actually use the platform, not for everyone eligible. This is smart design. The company gets paid only when it delivers an experience. But it also sets up the dynamic I want to examine.

The clinical evidence is real

Before getting to the structural argument, the product case deserves its due. Hinge has invested more in clinical research than most digital health companies: a 68% reduction in pain among program completers, 56% fewer spinal fusion surgeries among participants versus matched non-participants, and a 42% reduction in new opioid prescriptions.

The Peterson Health Technology Institute evaluated eight virtual MSK platforms in 2024 and found that physical therapist-guided platforms like Hinge improved pain and function comparably to in-person physical therapy, with a net decrease in spending. That is an independent evaluator, not a company press release.

For the people who engage with Hinge, this is a product that works. There is no doubt that Hinge is effective. The question is who actually receives that effect and why the number is as small as it is.

The 3.9% number

In 2024, the S-1 reported that 15.7 million people were eligible for the platform through their employer or health plan. Their annual yield — the share of those employees who actually enrolled and used the platform — was 3.4%. As of Q4 2025, that number had grown to 3.9% of 20 million eligible people. Hinge frames this as a growth story, and by their own trajectory it is: 2.9% in 2022, 3.0% in 2023, 3.4% in 2024, 3.9% in 2025.

But compare it against other employer benefits. Traditional Employee Assistance Programs — long considered the utilization floor for employer benefits — see 2–5%, and Hinge is in that range. Employer wellness programs with financial incentives broadly see 20–40% participation. The gap between those benchmarks and 3.9% is meaningful, even accounting for the fact that MSK care is a more specific, episodic need than a general wellness program.

The standard read of low utilization in digital health is: awareness problem, activation problem, benefits communication problem. Those are real. But I think they explain only part of what’s happening at Hinge. The deeper constraint is in how the business model proves its own value.

The ROI model can only see who is already in pain

The primary activation tool is HingeConnect — a program that ingests employer claims, pharmacy, and prior authorization data to identify members already in an active MSK episode and reaches out to them directly. Targeted enrollment through this channel grew over 160% year-over-year in 2025 and is the primary driver of yield improvement. The logic is sound: find the person already in pain, reach them at a moment of health intent, make enrollment feel inevitable rather than optional.

But notice what HingeConnect requires: a claim. A prior authorization. A pharmacy record. Something that has already happened in the healthcare system. It is, by design, a tool for finding the symptomatic population — the employees who are already expensive or about to become so.

This is not a criticism of HingeConnect. It is a precise description of why the employer B2B ROI model works. The value story to a CFO is built on avoided cost: a prevented spinal fusion saves an employer $50,000 or more. That arithmetic is clean and attributable. The S-1 reports 117% net dollar retention, which means employers are expanding their contracts because the ROI is legible. It shows up in claims data. It is measurable within a budget cycle.

The population that cannot be reached this way is the one that never generates a claim in the first place. The employee who does her exercises consistently for two years and never develops a problem produces no counterfactual in the data. You cannot prove a surgery that didn’t happen to someone who was never heading toward one. Prevention, by definition, eliminates its own evidence.

This is the structural constraint. The employer ROI framework, built on claims reduction, cost avoidance, and annual budget cycles, is systematically blind to preventive value. Hinge didn’t design that framework. They’re operating rationally within it. But it means that the population HingeConnect can find and the employer can fund is bounded by who is already sick enough to show up in the data. The preventive majority — the employees who would benefit from MSK care before they become expensive — are invisible to the measurement system that justifies the spend.

I am in that invisible population. The product serves me well. Their business model has no clean way to prove my value to my employer.

A note on the competitive question

The urgency of this constraint increases when you look at the competitive landscape. The closest competitor, Sword Health, acquired Kaia Health in 2025 and has pushed toward a more AI-driven model with fewer PT touchpoints and more algorithmic guidance. The competitive pressure is essentially a question about where the clinical value actually comes from: the exercise program, or the human oversight of it? If AI-driven programs with lower cost structures can match outcomes, the margin profile gets harder to defend.

Hinge’s continued emphasis on its care team is both a differentiation claim and a bet on which side of that question the evidence lands. My own experience suggests the PT relationship does real work — not because the exercises alone wouldn’t help, but because having someone paying attention changes commitment to the program. That is one data point. If large employers start demanding outcomes data stratified by care model intensity, the market will get a cleaner answer. And the pressure to resolve that question becomes more acute if Hinge needs to expand its addressable population beyond the symptomatic cohort it currently serves well.

What it would take to change this

What looks like a behavior-driven ceiling is actually in the measurement infrastructure of employer health benefits. Closing it requires changing the proof of value, not just the activation tactics.

The most plausible path within the employer channel is a longer ROI window and a different metric set. Absenteeism, short-term disability onset, and productivity scores are all measurable through HR systems over a 2–3 year horizon, and they capture value that claims reduction misses. A member using Hinge preventively may not generate claims savings, but she may generate meaningfully fewer disability claims over three years. Some large sophisticated employers, particularly in tech and professional services where workforce productivity is more tractable to measure, could be sold on this framing. The pitch adds a second proof point alongside surgery avoidance: “your low-acuity users cost you less over time.” That requires longitudinal data Hinge is accumulating but has not yet made the centerpiece of its employer value story, at least not publicly.

The other avenue is the health plan channel. Plans have longer actuarial windows than employers and measure total cost of care across multi-year periods. If a plan structures a partnership around preventive enrollment — triggered by age, occupation, activity level, or early risk signals rather than existing claims — the ROI attribution shifts to actuarial math rather than a single budget year. Hinge already has relationships with all five major national health plans. Whether any of those partnerships are structured this way is not disclosed, but the infrastructure to pursue it exists. Some precedent exists elsewhere in the system: Medicare Advantage plans already offer preventive benefits that traditional insurance won’t touch, and value-based care contracts between plans and provider groups already use multi-year cost trend rather than single-year claims as the ROI metric. The commercial logic for longer attribution windows is already operating in adjacent markets.

Neither path is simple. Both require Hinge to develop a proof of value for a member type that the current model wasn’t built to see. That is a product marketing problem as much as it is a data or contracting problem — it requires articulating a new value proposition to a new buyer argument, not just collecting more evidence for the existing one.

Final Thoughts

Hinge Health is a well-built product with real clinical evidence behind it. The 117% net dollar retention and 98% client retention tell you employers are finding value in a form they can measure and expand. That is a genuine achievement in a category where most digital health companies struggle to prove ROI at all.

The 3.9% yield is not a failure of outreach. It is a rational outcome of a measurement system that can only fund what it can see. What it can see is the symptomatic population that shows up in claims. The preventive majority is real, clinically meaningful, and currently commercially unreachable through the existing employer ROI model.

I’ll keep using Hinge. And I’ll keep an eye on whether they can build a proof of value for the employee who never files a claim — and whether the employer, or the health plan, can be convinced to pay for it before that claim arrives.