The cat is now a connected device

From collar to cloud

The old technology stack for a cat was a bell on a collar and a microchip the size of a rice grain, readable only by a vet with the right scanner. That was it for half a century. The microchip was a database primary key without a database worth speaking of: one number, one registry, one phone call if you were lucky.

Then the smartphone supply chain did what it does to every adjacent market. GPS modules, accelerometers, batteries and LTE radios became so cheap that putting them on a two euro collar became a rounding error. The same component deflation that gave us the fitness tracker gave the cat a location beacon, an activity monitor and a sleep tracker. Nobody set out to instrument cats. It just became economically absurd not to.

The result is a category that went from novelty gift to serious purchase in about five years. GPS collars now report position within seconds across most of Europe. Smart feeders portion by gram and log every meal. Water fountains track intake. And at the top of the stack sits the device nobody predicted would matter most: the litter box.

The litter box is the product

Ask any vet what they wish they knew about a cat and the answer is rarely exotic. Cats hide illness until it is advanced; it is a survival reflex. The earliest reliable indicators are boring: weight trend, visit frequency, urine output, changes in routine. All of it happens in the litter box, daily, voluntarily, at home. The connected litter box is not a convenience gadget. It is a clinical instrument that the patient uses on schedule without being asked.

The modern versions weigh the cat on every visit, distinguish between multiple cats by weight curve, log duration and frequency, and flag deviations against the animal's own baseline rather than a population average. That last part matters. A model trained on your cat's normal beats a breed-average chart for the same reason personalised baselines beat population thresholds everywhere in medicine.

The conditions this catches early are the expensive ones: urinary tract issues, kidney disease, diabetes. In cats, kidney disease is common enough that a meaningful share of the senior population will face it. A two week head start on a weight trend is the difference between a diet change and an emergency clinic bill that costs more than the device by a factor of ten.

This is the same argument this publication makes about infrastructure, applied to a four kilo mammal. The scarce thing was never the sensor. The scarce thing was the baseline: continuous, boring, longitudinal data about what normal looks like. Whoever owns the baseline owns the diagnosis timeline.

Who owns the cat's data

Here is where it gets less cosy. Every one of these devices ships its data to a cloud you did not choose, under terms you did not read, into a business model that is rarely just hardware margin. The hardware is subsidised by the subscription, and the subscription is justified by the data. Your cat has a data broker.

The commercial logic is already visible. Aggregated, anonymised pet health data is valuable to insurers pricing pet policies, to food brands reformulating products, and to pharma companies running animal health trials. None of this is sinister by itself. But the pattern is identical to every other consumer IoT category: the device is cheap, the data is the asset, and the subject of the data has no seat at the table. In this case the subject is also the only family member legally classified as property, so do not expect the regulator to rush in.

The sensible position is not refusal. The health upside is real and the privacy exposure of a cat is, charitably, limited. The sensible position is the same one we argue for enterprise software: know what you are paying with. If the device is free and the subscription is the business, the data is part of the price. Price it in, and buy anyway if the arithmetic works.

• Check whether the device works locally when the subscription lapses, or bricks itself
• Check whether you can export your own cat's data, and in what format
• Check whether the company sells or licenses aggregated data, and to whom
• Check what happens to ten years of baseline data when the startup gets acquired

What the metrics should look like

Pet tech marketing measures engagement, because engagement is what apps measure. The metrics that matter for an animal are different, and they look a lot like the SRE dashboards we argue for elsewhere: baseline, deviation, time to detect, time to act.

Note the last one. The industry has built beautiful dashboards for owners and almost nothing for the vet, who is the only person who can act on the data. The export button, the vet report, the boring PDF: that is the feature gap. The first platform that treats the vet as the customer of the data, rather than the owner as the consumer of the app, wins the category.

What to change on Tuesday

And one for the industry. The cat is a connected device now. That sentence was a joke five years ago and a market map today. The companies that understand they are in the longitudinal health data business, not the gadget business, will build the vet integrations, the export formats and the trust. The rest will sell cameras with ears on the box.