Test method
How we test vacuum cleaners
The uses we score and the tests we run on every robot, cordless stick and handheld vacuum: dust, rice, sofas, hair, runtime, noise and mopping.
Every vacuum cleaner we review goes through the uses and tests below, in a real home, on the unit we bought or borrowed. Robots, cordless sticks and handhelds do different jobs, so some tests apply to one type only, and each test says so. Within a type, the same test means the same thing for every product on the site.
Each test result in a review links back to the matching test on this page. Floor area, floor types and the number of weeks of use are stated in each review, because a small 1BHK and a large 3BHK with rugs are different jobs.
What we score
- Daily dust on hard floors
- How well it lifts the fine dust that settles every day on vitrified tile, marble and Kota stone, plus spilt rice and grit along the skirting.
- Rugs and carpets
- How much fine dust and grit it pulls out of low-pile rugs and dhurries, and, for robots that mop, whether the mop stays off them.
- Pet hair and long hair
- How much long hair and fur it picks up, and how quickly hair wraps around the brush roll in daily use.
- Sofas, curtains and cars
- How well the handheld mode cleans fabric sofas, curtain pleats and car seats. Applies to sticks and handhelds, not robots.
- Mopping
- For robots that mop: whether it lifts dried stains, leaves dark tile free of streaks, and how well the dock keeps the pads clean and dry.
The tests
-
Fine dust on tile, one pass
We spread a standard patch of fine dust on vitrified tile (fine flour for robots) and report the share picked up in a single pass of the main floor head, or a single robot pass over the patch. Stick and handheld tests are run three times.
-
Dry rice on tile, one pass
We scatter dry rice on vitrified tile, including along the skirting, and report the share picked up in one pass. For sticks we also note how many passes it took to clear the rice completely.
-
Fine flour on a rug, one pass
We spread fine flour on a low-pile rug and report the share picked up in one pass, in the mode the vacuum uses on carpet. If a review uses a thicker rug or max mode, the result says so.
-
Embedded dust from a fabric sofa
Sticks and handhelds only. We work dust and crumbs into the seat of a fabric sofa, clean it once in handheld mode with the brush or crevice tool, and report the share removed.
-
Long hair on a rug
We spread long hair on a low-pile rug and report the share picked up, then check the brush roll for wrapping straight after the test and again after weeks of daily use.
-
Runtime, everyday mode
From a full charge until the unit stops or the robot heads back to its dock. Sticks and handhelds run in the lowest power mode with the main head fitted; robots run on standard suction. Corded vacuums are not timed.
-
Runtime, top power mode
The same as the everyday runtime test, in the strongest power mode. Makers' rated runtimes usually refer to the lowest mode, so this is the number to check if you clean rugs and sofas on full power.
-
Time to clean the test home
We time one full clean of the test home in the everyday mode and note its floor area and the battery left. Test homes differ between reviews, so read the result with the area in its note. For robots we also note any recharge-and-resume and any part of the floor left untouched.
-
Noise, everyday mode
We measure the sound level at one metre from the vacuum in its normal or standard mode while it cleans a tile floor.
-
Noise, top power mode
The same as the everyday noise test, in the strongest mode. For robots with a self-emptying or mop-washing dock, we note the dock's noise separately.
-
Days before the filter clogs
Sticks and handhelds. In daily use in a dusty home, we note how many days pass before suction drops noticeably, then check that tapping out or rinsing the filter, and letting it dry fully, restores it.
-
Stuck events per week
Robots only. Over the test weeks we log every time the robot gets stuck or needs rescuing from cables, rug fringes, doormats or slippers, and report the weekly average. Robots with obstacle cameras also run on a staged floor with cables, slippers and socks.
-
Highest door sill crossed
Robots only. We note the highest door sill or threshold in the test home that the robot climbed without help, and where it hesitated or stalled.
-
Dried chai stain
Robots that mop. We let a chai spot dry on vitrified tile or marble and count the mopping passes needed to remove it. A mop that does not lift it at all is recorded as a fail.
-
No streaks on dark tile
Robots that mop. We mop dark glossy vitrified tile at the medium water setting and check for streaks once it dries. Pass means none were visible; we note if a lower water setting fixed any streaking.