Some variation from one portion to the next is normal for most automatic feeders. A steady change in the average over weeks is not, and it usually has a cause you can find. Many feeders measure kibble by volume, not weight, so the grams in each portion move with the food itself. When you ask about automatic feeder portion accuracy over time, you are really asking two questions: how much do single portions scatter, and has the average drifted? Scatter is built in. Drift is a signal. A kitchen scale is the only reliable way to tell them apart.
No maker we could find publishes a figure for how much a portion is allowed to change over months of use. So this page does not invent one. It shows what the makers do say, what can change a portion's weight, and a simple test that gives you your own number.
What feeder makers actually say about portion size
Start with the manufacturers' own words, because their support pages are more candid than their product listings.
PETLIBRO defines a portion as the food released each time the feeder's rotor completes a rotation, and says a portion is set by the volume of the food rather than its weight (PETLIBRO support, read September 24, 2026). The same page lists the volume per model, for example about 20 ml, or roughly 1/12 cup, for its Granary models, and says the actual amount may vary with the size, shape, freshness and thickness of the food. Its feeder FAQ goes further for the Granary Smart Camera Feeder: one portion usually weighs 10 g with 4 g of "floating" either way, may fluctuate more depending on the food, and cannot be recalibrated because the feeder does not contain a load cell.
Whisker says its Feeder-Robot dispenses in 1/8-cup increments with the standard dispenser, or 1/4-cup increments with the larger one (Whisker support, read September 24, 2026). Those are volume measures. The page makes no statement about accuracy or variation.
Sure Petcare takes a different approach with its Microchip Pet Feeder Connect. It does not dispense food on a schedule. You fill the bowl, and built-in scales with LED guidance help you serve a set weight to within 1 g, according to the product page (read September 24, 2026). It then records how much the pet eats.
| Design | How a portion is measured | What the maker states about accuracy |
|---|---|---|
| Rotor dispenser (PETLIBRO) | Volume per rotation | Varies with the food; the camera feeder given as 10 g, give or take 4 g |
| Cup-increment dispenser (Whisker Feeder-Robot) | Volume, in 1/8 or 1/4 cup steps | No accuracy statement on the support page |
| Owner-filled bowl on scales (Sure Petcare) | Weight, filled by you | Within 1 g when filling, per the maker |
Notice what is missing from all three: any figure for change over time. Some review roundups publish gram-level drift numbers from a single unit tested for a month. One unit over one month describes that unit, not the category, which is why those numbers are not repeated here.
Scatter vs drift: the difference that answers "is this normal?"
Scatter is the spread between individual portions. Kibble does not settle into a dispensing pocket the same way every time, so each rotation can drop a slightly different weight. On PETLIBRO's own figure for its camera feeder, a single 10 g portion can land anywhere between 6 g and 14 g. That is a wide band, and it is the maker's stated normal, not a fault.
Drift is a change in the average. If your portions used to average one weight and now average a clearly different one, with the same food, something has changed. A dispenser that turns the same amount moves the same volume, so a lasting shift usually comes from what is filling that volume, or from something partly blocking it.
This is why a single weighed meal proves nothing. One light portion could be scatter. Several portions that all come out light, checked twice a few weeks apart, is drift.
What can change the weight of a portion over time
Work through these before assuming the feeder is wearing out. Most are quick to check.
- A new bag or a new formula. PETLIBRO names the size, shape, freshness and thickness of the food as causes of variation. A new recipe, a different kibble size, or a bag that has sat open for a while can change how much weight fits in the same volume.
- The hopper level. How much food sits above the dispenser, and how it has settled, may affect how each pocket fills. Note the fill level whenever you weigh, so you compare like with like.
- Moisture and clumping. Kibble that picks up humidity can stick together and fill pockets unevenly. Suspect this if the change arrived with a humid spell.
- Residue in the mechanism. Kibble dust and oils can build up on rotors, paddles and chutes, and a coated pocket may hold less. Cleaning the dispensing path to the maker's instructions rules this out.
- A partial jam. A lodged piece may block part of a pocket without stopping the feeder, so meals can keep arriving, just smaller.
- A settings change. Someone adjusted the portion count, or, on a feeder with interchangeable dispensers, the app setting no longer matches the dispenser that is installed.
- Another cat. In a multi-cat home, a bowl that looks emptier may belong to a feeder that dispensed normally and a second cat who got there first.
Mechanical wear is possible, but we found no maker documentation that puts a number on it. If everything above checks out and the average keeps moving, that is a support and warranty question for the manufacturer.
A kitchen-scale test for automatic feeder portion accuracy over time
This takes a few minutes and needs a digital kitchen scale that reads in grams.
- Place an empty container under the chute and zero the scale with the container on it.
- Dispense a single portion using the feeder's manual feed button or the app's feed-now option.
- Weigh it and write it down. Empty the container, zero again, and repeat until you have at least five single portions.
- Record the context: the date, the food brand and formula, when the bag was opened, and roughly how full the hopper was.
- Work out the average and the range, meaning the lightest and the heaviest portion.
- Repeat the test in about a month with the same food and a similar hopper level.
A simple log is enough:
| Date | Food and bag | Hopper level | Five portions (g) | Average | Range |
|---|---|---|---|---|---|
| First test | |||||
| A month later |
How to read it: if the new average falls inside the range you recorded the first time, you are looking at scatter. If it falls outside that range on two tests in a row, treat it as drift and work through the causes above. That gives you a threshold built from your own feeder and your own food, which is more useful than any general percentage.
The best time for the first test is the day you set the feeder up or open a new bag. If you missed that moment, today's test becomes your baseline.
When drift matters, and when it matters less
For a healthy adult cat on a flexible feeding plan, a modest shift matters less than the longer trend in your cat's weight and body condition, which your veterinarian can assess at a checkup.
It matters much more when a veterinarian has set a specific daily amount, for weight loss or a prescription diet. Measuring by volume has a documented accuracy problem. The University of Guelph's Ontario Veterinary College pet nutrition team (read September 24, 2026) summarizes two studies: a 2010 study led by Dr. Alex German found measuring-cup portions ranging from 18% under to over 80% over the intended amount, and a 2018 study by Dr. Jason Coe and colleagues found owner errors from 47.83% under to 152.17% over. Its advice is to measure pet food by weight with an electronic gram scale. A feeder that measures by volume works on the same principle as a cup, so if your cat's plan depends on exact amounts, weigh what the feeder dispenses and ask your veterinarian how to set it up.
Do not adjust portions on your own to chase a change in your cat's weight. Weight gain or loss, or a change in appetite, is a question for your veterinarian first. The feeder can carry out a plan. It cannot tell you what the plan should be.
What the feeder's log cannot tell you
A dispensing log records what the machine released. It does not record what your cat ate. A feeder can dispense perfectly into a bowl that a housemate cat empties, or that your cat leaves untouched, and the app will still say "fed."
The same principle runs through this site's thinking on litter box data: device numbers are a prompt to look closer, not a verdict. The guide to what litter box weight tracking is actually good for makes that case in detail, and it applies to feeder logs just as well. If more than one cat shares the feeding area, the account of running a self-cleaning litter box in a three-cat home is a useful reminder of how quickly shared devices blur whose data is whose.
Build the check into a routine
Drift is slow, so a fixed habit catches it better than a worried moment. Re-run the scale test whenever one of these happens:
- you open a new bag or switch formulas,
- you clean the dispensing mechanism,
- the feeder has jammed, even once,
- the season turns humid,
- or a month has simply gone by.
If you already keep a weekly look at your automated devices, add the feeder to it. The site's self-cleaning litter box maintenance schedule is written for litter boxes, but its core idea, a fixed check of the physical device instead of trusting the app, carries straight over to a feeder: hopper, chute, bowl, and a glance at what was dispensed.
Frequently asked questions
Do automatic feeders get less accurate over time?
They can, but most apparent drift has a findable cause, such as a new food, residue in the mechanism, a partial jam, humidity or a settings change. We found no manufacturer figure for accuracy loss over time. Weigh several portions now and again in a month to find out.
How much variation between portions is normal?
It depends on the feeder and the food. PETLIBRO states that for its Granary Smart Camera Feeder a portion usually weighs 10 g with about 4 g of variation either way, and more with some foods. A five-portion test tells you what is normal for your own feeder.
Why does my feeder dispense a different amount with a new bag of food?
Many feeders, including PETLIBRO's, measure by volume. PETLIBRO says the amount in a portion varies with the size, shape, freshness and thickness of the food, so a different formula or kibble size can change the grams in every portion. Re-weigh after any food change.
Can I recalibrate an automatic feeder?
Some cannot. PETLIBRO says its Granary Smart Camera Feeder has no load cell, so the dispensing amount cannot be recalibrated. Check your own manual. If recalibration is not possible, adjust the number of portions based on weighed results and your veterinarian's advice.
Is an automatic feeder accurate enough for a weight-loss diet?
Use it to deliver meals, not to measure them. Veterinary nutrition guidance from the University of Guelph favors weighing food with a gram scale. Weigh what your feeder dispenses and follow your veterinarian's plan.
The short version
Portions scatter, and that is normal. Averages should not wander, and when they do, the cause is usually the food, the mechanism or the settings. Weigh five portions, write the numbers down, and weigh again in a month. Let your own numbers define normal for the feeder, and let your veterinarian decide what your cat should be eating.