Six in the morning, kitchen light on, the bag open on the floor by the door. The scoop goes in, comes out heaped, gets a small shake to level it off, and the lot goes in the bowl. That shake is the whole of the quality control. Nothing in the room weighs anything, nothing in the room knows what the dog did yesterday, and the dog is delighted either way.
That morning happens in a great many New Zealand kitchens, and it goes wrong in more of them than you would guess. In the largest local dataset, 24,247 dog records from 10 North Island clinics, 26.1% of dogs were overweight and a further 2.3% obese: 28.4% between them, more than one in four, carrying weight they should not be. (Gates MC 2019) The people holding the scoop are not careless. Most of them are measuring.
It looks like willpower
The obvious reading is that this is a discipline problem. Feed a bit less, skip the treats, use the cup properly, and the dog stays trim. Some of that is true, and it is the part everyone recognises. A biscuit here and a top-up there, all given out of love, add up over a year. A lot of overfeeding is simply affection with nowhere precise to go: you want to give the dog something, so you give it a bit more dinner.
But willpower is an explanation that assumes the owner knew the right amount and decided to go past it. Mostly they never had it. You cannot fail to stick to a number nobody ever gave you.
Then somebody measured it
For most of the time anyone has kept dogs, the cost of the extra bit has been a matter of opinion. A bit heavy is a judgement, not a quantity, and a judgement is easy to argue with at six in the morning.
Then it got measured, properly and slowly. A study followed 48 Labrador retrievers from seven litters, split into sibling pairs, with one dog in each pair fed 25% fewer calories than its littermate from eight weeks of age. It ran from 1987 to 2001, 14 years, the whole of those dogs’ lives. The lean-fed dogs reached a median lifespan of 13 years. Their siblings reached 11.2. The gap is 1.8 years, about 15% more life, and it came from nothing but the amount in the bowl. The lean-fed dogs also went far longer before their joints needed treating, to a mean age of 13.3 years, three full years later than their pairmates. (Kealy RD 2002)
It is worth saying what that study does not show, because it is the reason the number is worth trusting. It restricted calories. It did not test whether an exactly measured portion beats a roughly measured one, it ran a single breed in a single research colony, and it was paid for by the pet-food company whose food the dogs ate. So it is not evidence that scoops are bad. It is evidence of the size of the thing being decided every morning by somebody holding one, and the size is 1.8 years.
The harms in between are the ones that never present as a cause. Carrying extra weight tracks with insulin resistance and with osteoarthritis, and osteoarthritis in particular arrives so gradually that it gets read as age and often goes undiagnosed. (Association for Pet Obesity Prevention 2026) A 2018 analysis across 12 breeds found the same direction outside the colony: heavier dogs, shorter lives. One harm on that list needs correcting, because it is the one everybody reaches for first. Obesity is strongly tied to diabetes in cats, but canine diabetes is largely immune-mediated, so the honest chain from an over-full bowl runs through insulin resistance, joints, and years, and not through a diabetes diagnosis.
The unit is wrong
A cup is a measure of volume. What a dog needs is energy. The two do not line up, and nothing on the packet closes the gap. A heaped scoop holds a good deal more than a level one, and nobody levels a scoop of dog food except by shaking it. Two foods of different density fill the same cup with different amounts of dinner. Veterinary nutrition guidance has said for years to work out a dog’s energy requirement individually and to measure the food by weight rather than by volume, for exactly this reason. (WSAVA Global Nutrition Committee 2026) Almost no kitchen owns the scale. The scoop came free in the bag.
The bag cannot rescue it either. One printed number cannot be right for a whippet, a labrador, and a 15-year-old with a slow thyroid, so the bag prints a range instead, one to two and a half cups, which is the most honest thing a one-size label can do and no use at all for one specific animal. That is not a manufacturer hiding something. It is a label being asked a question it has no way to answer.
Nothing in the room corrects you
Every other instrument in the situation is just as blind. The dog cannot tell you it has had enough, and most dogs will eat cheerfully past the point. The feedback that would teach you is the animal itself getting heavier, which arrives over months, gradually enough that you stop being able to see it, and far too late to attach to any particular morning. In a lot of houses more than one person feeds the dog, with no shared record of what has already gone in the bowl, so the error compounds without anybody making it.
The few things that set a dog’s portion are all knowable: its weight, its age, whether it has been desexed, and how much it moved today. They are just not known in the kitchen at six in the morning. Work them out somewhere with a scale, hand back an exact amount for that dog rather than a range for its category, and the daily decision stops being a decision. That is what Portions is built to do, and it is the least interesting part of this.
The interesting part is what the arrangement does when it is left alone. An amount gets decided fresh each morning by a person holding a volume instrument, against a range printed for a category of animal, with no feedback for months and nothing traceable in it when it finally comes. A decision made under those conditions does not sit still. It drifts, and it drifts one way, because nothing in a kitchen pushes toward less: nobody has ever tipped food back out of a bowl to show a dog they love it. It took 14 years and 48 Labradors to put a number on what that costs. The harm was never hidden. It was only ever measured somewhere the person holding the scoop could not see it.