Why do people quit the exercises that decide whether they walk normally again?
Adherence to home rehab is the single largest lever on a surgical knee outcome, and it is the one nobody is watching. The clinic sends a person home with a sheet of exercises and a follow up in six weeks. What happens in between is invisible to everyone, including the person doing it. Discovery started by treating that invisible middle as the whole problem.
A rep done with confidence and a rep abandoned in fear look identical on a chart. The difference is everything, and no product was measuring it.
Discovery framing note
Reconstructing the timeline from the kitchen table to the quiet quit.
Switch interviews rebuild the story backward from a decision. Here the decision was the one nobody announces: the day the exercises stop. People were asked to walk through the moment they were handed the plan, the first session alone, the first twinge that scared them, and the slow drift into doing less. The same shape kept surfacing, and it was emotional long before it was practical.
I could not tell if I was doing it right or hurting myself, so I did less of it. Less felt safer than wrong.
Participant, 6 weeks post ACL reconstructionThe app I tried counted my reps. It never once told me whether the reps were any good. I was alone with a number.
Participant, knee replacementEvery twinge felt like proof I had wrecked the surgery. By week three I was guarding the leg more than using it.
Participant, meniscus repairI did not want a coach in my face. I wanted to know I was safe and then be left alone to do the work.
Participant, 9 weeks post reconstructionThe handoff
A printed plan and a six week follow up. Confidence is high because a professional is still in the room.
The first solo session
The professional is gone. The first question arrives immediately: is this right, and is it safe.
The twinge
A normal recovery sensation gets read as damage. With no one to ask, fear fills the silence.
The quiet quit
Doing less feels safer than risking harm. The drift is gradual, unannounced, and it decides the outcome.
Four forces decide whether someone keeps going. Two pull toward progress, two pull back.
Mapping the forces made the design problem legible. The push of the situation and the pull of a better way were real, but they were being overwhelmed by the anxiety of being judged and the habit of avoidance. Range could not just add more pull. It had to remove the anxiety, because that is the force that was winning. Open each force to see what surfaced.
- I cannot see whether I am getting better, so I cannot tell if the work is worth it.
- Alone at home, I have no way to know whether a movement is safe or doing harm.
- The outcome that matters most to me is the one I have the least visibility into.
- Tell me only when I drift, and stay quiet when I am doing well.
- Give me proof the work is adding up, so the next solo session feels worth it.
- I want presence without pressure, a sense that I am safe without a sense that I am surveilled.
- What if it tells me I am failing on the days I already feel fragile.
- What if it flashes red and confirms the fear that I have hurt myself.
- What if a piece of software makes a medical judgment it has no business making.
- Skipping a session is easier than risking a movement that might hurt.
- I would rather not measure than be told the number went the wrong way.
- Avoidance feels like safety, even though it is the thing quietly ending my recovery.
The synthesis was a single sentence that every later decision answered to. Range is a doubt removal machine. If a feature removes doubt it earns its place. If it adds doubt, even a clever feature, it is cut.
One functional job, carrying an emotional job and a social job underneath it.
The functional job is what the person hires Range to do. The emotional and social jobs are why it has to be done a particular way. Range was designed to satisfy all three at once, because solving only the functional job is exactly the trap the rep counting apps fell into.
When I am recovering from knee surgery alone at home, I want to know that each rep is safe and that my range is adding up, so I can keep doing the work without fear of harming myself.
This is the job a rep counter cannot do. Counting reps proves activity. It does not answer the only two questions the person actually has: was that safe, and is it working. Range answers both, in the moment and over time.
When a normal twinge makes me afraid I have undone the surgery, I want to feel capable and unafraid, so I can stay in the recovery instead of guarding against it.
The emotional job is why the feedback economy exists. Silence on a good rep, one calm cue on a drift, warm color instead of alarm red. Every one of those choices is the emotional job refusing to confirm the fear that ends adherence.
When I see my therapist again, I want to arrive as the patient who did the work, so I can be met with progress instead of a lecture, and hand over a record in seconds.
The social job is why progress ends in a clean therapist summary, and why a persistent dip routes to a human rather than a verdict. The product is not the last word. It is the thing that hands a better conversation to the person who is.
Wide, then sharp. Twice.
Discovery opened wide into the lived experience, then narrowed to a single problem statement. Develop opened wide again into ways to coach without surveillance, then narrowed to three moments and one hard tradeoff. The diamonds are not decoration. They are the discipline of not narrowing too early.
The wall of doubt
Switch interviews and the forces of progress. The hard problem is emotional, not technical. Pose estimation was never the risk.
A doubt removal machine
One sentence to measure every decision against, and one boundary drawn early: Range coaches, it never diagnoses.
Coaching without surveillance
Wide exploration of feedback that does not pressure. Silence as feedback, change instead of score, color that refuses alarm.
Three moments, one tradeoff
Calibration, live coaching, and progress. Plus the senior decision to stop and hand back to a human at the safety boundary.
Every moment in the product is an answer to a force.
The synthesis is only worth something if you can trace it into the interface. Each of the three live moments exists because a specific force had to be removed or earned. Open any of them and the discovery is visible in the design.
Calibration
Being seen gently. The app finds you. The only thing it asks you to move is the phone, never your body.
Open the prototypeLive coaching
Restraint made visible. Silence on a good rep, one calm cue on a drift, a stop that protects rather than scolds.
Open the prototypeProgress
Proof on the days you doubt it. Change, never a score, and an honest dip the design meets with calm.
Open the prototypeThe hardest decision was the one that limited the product.
Remove doubt, or get out of the way.
Every screen earns its place by removing a specific doubt the research surfaced. Is this safe. Is it working. Am I still doing this for a reason. Anything that could not answer one of those was cut, no matter how clever it looked in isolation.
Range coaches. It never diagnoses.
The strongest pull in the room was to let the product judge. Discovery said the opposite. The anxiety that ended recovery was the fear of a machine making a medical call. So the moment something looks wrong, Range stops, logs it, and hands back to a human. Less product, on purpose, because the boundary is the trust.