How Insights works
Insights turns the sessions you've logged into a picture of how your dog is doing over time. This page explains exactly how those numbers are worked out, what they mean, and — just as importantly — when we deliberately stay quiet.
We would rather tell you nothing than tell you something that isn't true. Insights only speaks up when there's enough data behind a change to be reasonably confident it's real and not just noise. If your dog's panel is quiet, that isn't a fault — it means we're still watching.
Every exercise builds its own history — the longest hold, the highest reps, the total work in each session. Those individual histories are then grouped by body area using the body-area tag on each exercise.
We deliberately do not use whole-session totals. If you combine two programs, add a set, or restructure a routine, session totals jump for reasons that have nothing to do with your dog. Exercise-level histories survive all of that.
When a session is saved we record both what the program prescribed and what your dog actually did. That lets us tell apart three quite different things:
The program has progressed — someone raised the bar because your dog was coping.
Your dog is completing what's asked — finishing sets they previously couldn't.
Your dog is exceeding what's asked — you pushed a little and they delivered.
The third is the strongest signal in the whole app, because nobody prescribed it.
Where you record thigh or limb circumference, we focus on the difference between left and right rather than the raw number changing over time. Both limbs are measured in the same sitting, by the same person, with the same tape — so much of the inevitable measurement error affects both readings equally and cancels out.
Measuring a dog's thigh with a tape measure is far less reliable than it feels. Published veterinary research has found poor agreement even between trained clinicians using spring-tensioned tape — and one study that immobilised the limb entirely still found large variation between and within observers.
That doesn't make measurements worthless. It means we treat them with appropriate caution: your figures are always charted for you and your physiotherapist, but Insights only comments when a change is large enough and sustained enough to stand well clear of ordinary measurement wobble.
Not enough sessions yet. The panel tells you roughly how many more it needs.
Not enough time yet. Some things — muscle in particular — take weeks to change no matter how diligently you log. Logging more often won't speed those up, and we won't pretend otherwise.
The change is too small to trust. Below the thresholds, we say nothing rather than dress up noise as progress.
Your dog has been uncomfortable. If you've recorded discomfort recently, we won't report rising numbers as straightforwardly good news. A dog can push through and still be sore, and that's the moment to talk to a professional — not the moment for an app to cheer.
A trend has to clear all of these before Insights will comment on it. We compare averages, not single sessions — one session is a noisy thing, depending on your dog's mood, the time of day, the surface, and how much time you had.
PawPhysio is a tracking tool, not a treatment tool. Insights describes what your dog did. It does not tell you what to do next. Specifically, it will never:
Say anything about healing, tissue, or what's happening inside your dog · tell you your dog is fine or that there's nothing to worry about · suggest new exercises, extra sets or a harder program · diagnose anything · compare your dog to other dogs · suggest you've fallen behind.
Rest days are never counted against you. There are no streaks to break. Recovery isn't linear, and an app that punishes a quiet week is an app that's wrong about recovery.
Every figure is measured from a starting point, shown at the top of the panel. By default that's your dog's surgery date, their veterinary clearance date if you've recorded one, or the day you started logging if you haven't.
You can change it in your dog's settings — useful if circumstances change, such as a second procedure or a break from training, when comparing against much older numbers would be misleading.
Exercise descriptions and body-area groupings are informed by published veterinary rehabilitation and canine conditioning literature, including Canine Sports Medicine and Rehabilitation and the Handbook of Canine Exercise Therapy, alongside peer-reviewed studies on canine limb measurement reliability and muscle activity during therapeutic exercise.
Referencing these works does not imply that their authors endorse PawPhysio.
Our thresholds are carefully reasoned, not perfect. There is very little published research on how much a dog's performance naturally varies from session to session at home. Our figures draw on comparable human rehabilitation research and on the practical question of what size of change is worth mentioning. As real usage data accumulates we expect to refine them — and we'll update this page when we do.
Rising numbers mean the work is building. They are strong evidence that things are moving in the right direction. They are not a clinical measurement of your dog's strength, and we're careful never to word them as though they were. For that, nothing replaces a veterinary physiotherapist with hands on your dog.
If something in your dog's Insights doesn't make sense, or you think a number looks wrong, please tell us at pawphysio@kali-rising.com. We would genuinely rather hear it.