Cleaning
Robot vacuums, cordless sticks and carpet cleaners — where the headline numbers stop meaning anything.
Vacuum listings compete on three figures: suction in pascals, runtime in minutes, and bin size in litres. All three are real measurements and none of the three tells you whether the machine will clean your home. This page sets out what we compare instead, and where the specification sheet quietly stops being comparable between brands.
What this category covers
- Robot vacuums — with and without mopping, and with or without a self-emptying dock.
- Cordless stick vacuums — the main floor cleaner in most homes now.
- Handhelds — car interiors, stairs, upholstery.
- Carpet and upholstery cleaners — extraction machines and spot cleaners.
- Wet-dry machines — hard floors, spills, and combined vacuum-mop units.
The numbers that actually decide it
| Check | Why it matters | Where the spec sheet misleads |
|---|---|---|
| Suction, in pascals | It is the number every listing leads with, so it is the number buyers compare first. | Pascals are measured at the inlet with the brush sealed against a plate in a lab. On carpet, airflow and brush design matter more than the peak figure — and brands quote at different points in the airflow path, so two numbers side by side are often not the same measurement. |
| Runtime | Decides whether the vacuum finishes the home on one charge, or whether you get half a room and a recharging break. | Quoted at the lowest power setting, with a new battery, off carpet. At maximum power with a loaded brush roll, expect a fraction of it. Capacity also falls over a few hundred charge cycles. |
| Bin or bag volume | On a stick vacuum it decides how often you empty it. On a robot, it decides whether you have to be home at all. | Measured to the brim, excluding the pre-filter volume that fills first. On a self-emptying robot, the dock bag is the number that matters for how often you intervene. |
| Filtration and sealing | Determines how much fine dust goes back into the air, which is the whole point if anyone in the house has allergies. | The word “HEPA” is used loosely. Sealing matters as much as grade: an H13 filter in a leaky housing leaks. Look for both a stated grade and a description of a sealed system. |
| Brush roll and tangle | On long hair and pet fur, this is the thing that decides whether the vacuum actually gets used. | Effectively absent from the spec sheet. Check whether the brush comes out without tools, and whether the anti-tangle design is a shape or a marketing claim. |
| Battery and parts replacement | Decides whether the machine is repairable at year three or becomes waste. | Rarely stated. A vacuum with no user-replaceable battery has a design life equal to its battery, which is typically two to four years. |
Where the spec sheets mislead
Suction in airwatts versus pascals
These are different quantities, and some brands switch between them depending on which number flatters the product. Airwatts describe the power delivered to the airflow; pascals describe the pressure the motor can pull. A machine can score well on one and poorly on the other, and the one that matters depends on whether you are cleaning hard floor or deep carpet. A listing that gives only one, without saying why, is choosing a number rather than reporting a measurement.
“Up to X minutes”
“Up to” is doing real work in that sentence. It signals the lowest-draw configuration, with the display dimmed and the motor in its economy mode. Ask for the runtime at maximum power, on carpet, with a used battery — that is the number that describes your Saturday morning.
Self-emptying docks
A self-emptying dock does not remove the maintenance. It moves it: from a bin you empty every few days to a bag you replace every few weeks, plus a dock filter and a dock that needs its own cleaning. That is a convenience upgrade and a recurring cost, and it is worth knowing which of the two you are buying.
The cost that is not on the box
| Recurring cost | What triggers it | Why it matters |
|---|---|---|
| Dock bags | Self-emptying robot vacuums | Each bag holds several weeks of debris. Over a year, bag cost can exceed every other consumable on the machine combined. |
| Filters | Fine dust load, pets, washable versus replaceable design | Washable filters still have a replacement interval. A filter that is only nominally washable is a subscription in disguise. |
| Brush rolls and belts | Hair wrap, abrasive floors, worn bristles | The part that does the actual cleaning wears out first, and on some models it is not sold separately from the head. |
| Battery | Two to four years of normal use | Often the largest single lifetime cost, and the point at which many machines are quietly retired instead of repaired. |
| Mop pads and cleaning solution | Mopping robots and wet-dry machines | Pads are consumables with a wash limit. Proprietary cleaning solution is a recurring cost that only exists because the machine requires it. |
How we write this category
The starting point is the people who own the machine. We go through owner reviews, forum threads and support discussions looking for what recurs — a brush roll that tangles on long hair, a dock that jams, a battery that fades at eighteen months, a filter that clogs faster than the manual admits. A single review is noise; the same complaint from hundreds of owners over two years is a design fault, and it goes in the guide even when it is inconvenient.
Then we go back to the manual, which is where the qualified figures live — runtime at a named power level, filter grade, battery chemistry. That is how we tell a real fault from a misunderstanding, and it is why we sometimes decline to compare two figures at all when the brands measured different things.
What we do not do: we do not rank by commission rate, and we do not call a machine “best” without saying best for which floor, which home and which budget. See the disclosure for how the site is funded.
What happens after you buy it
Most robot vacuum pages stop at the purchase. The numbers that decide whether the purchase was a good one arrive later — a battery at year two, a sensor that behaves like it has failed, a part that quietly stops being manufactured.
We compared the published lifespan figures for this category and found they disagree by a factor of three, with none of the sources explaining where their number came from. How long do robot vacuums actually last? puts the four claims side by side, along with the repair-cost brackets that decide whether a machine is worth fixing at all.
Comparing two models?
If you have two listings open and the numbers do not line up, send them to us. Explaining why two “comparable” figures are not comparable is most of what this category is for.