How we test

Deep Clean Baking Soda Test Vacuums R Us Test Lab

How we do it. This is how we test how well a vacuum pulls embedded dirt out of carpet: the fine stuff that settles down in the fibers where you can't see it.The test carpet is a 48 oz frieze. Frieze is a looser, twisted pile. It isn't bargain carpet and it isn't high end, it sits in the middle of what most homes have. We weigh out 100 grams of baking soda, pour it in a straight line down about six feet of carpet, and rub it in until you can't see it.Then we weigh the bag, or the bin on a bagless machine. The vacuum runs back and forth over the line for 30 seconds, usually three or four strokes, and we weigh the bag or bin again. The difference is how many grams it picked up. Whatever is missing is still in the carpet.Between machines, our best performer goes over the carpet for three to five minutes so the next machine starts on clean carpet.We tried other debris, other floors, other time frames and other ways of laying it out. Baking soda won because it's fine enough to work deep into the fibers and heavy enough to weigh reliably before and after. We settled on 30 seconds because at five or ten seconds every machine grabs the surface stuff and the results bunch together. Run anything for ten minutes and even a terrible vacuum eventually gets it all. Thirty seconds is short enough to separate the machines and close to how people vacuum at home.How grams become stars: the scale runs in a straight line from 0 grams (1 star) to 85 grams (5 stars). 85 grams sits just above the best normal result we've recorded, so the top machines land near 5. Anything over 85 grams caps at 5. The actual grams are written on every reading so you can check the math.So far the results line up with what we've watched come across the counter for decades, with a few surprises in both directions.

Equipment. Scale, Timer

Pet Hair Side by Side Test Vacuums R Us Test Lab

How we do it. Pet hair is one of the hardest things we test, because it weighs almost nothing. Our deep clean test works by weighing what a machine picks up. With pet hair the weight is so small the scale can't tell machines apart, so this test is visual: every machine runs over the same rug, we photograph what it leaves behind, and we compare the pictures.The rug is an ordinary short black throw rug, the kind you'd find at the grocery store. Short pile is sticky, so the test material grabs onto it the way hair grabs onto carpet at home, and white fibers on a black rug make whatever is left easy to see.For the hair we use kapok, a fiber that grows on a tree. It's the standard demo material in the vacuum business and about as close to pet hair as we can get without shaving our dogs. It isn't perfect. Stronger machines sometimes break the finest fibers into a powder through static, which leaves a thin white haze that takes a couple more passes to clear. When a note on a reading mentions haze, that's what it means. We tried weighing it out for more control, and it's too light for that to work, so we lay a thick, even layer over the whole test area by hand and break up any clumps. A big clump in front of one machine would throw its result.Every machine gets one push forward and one pull back over five seconds, on a timer. Some machines clean better going forward and some going back, so both directions count. The timer keeps us from playing favorites. If we loved Sebo, we could give it nine seconds and give the Shark three. The timer doesn't let us.Three baseline machines set the scale:* **Dyson DC14: 1 star.** It sold in huge numbers, and plenty of owners believe it's great on pet hair because they can see hair in the bin. What they can't see is everything it left in the rug. * **Dyson V8: in between, 2.9.** Stick vacuums in general aren't strong on pet hair. We ran this one on low. On high it seals down onto the short rug so hard you can barely push it. Even on low it took about seven takes to get one clean five second stripe. Low and high picked up the same amount of hair, and this machine is here as a measuring stick, not as a test subject. * **Riccar R17: 5 stars.** Non stretch belts and an aggressive brush. In our testing it's as good as it gets on pet hair, with only a few other machines up there with it.Every other machine runs on its highest setting. After each run we photograph the path, vacuum the area clean, lay down a fresh layer to match the baseline area, and bring in the next machine.Each machine is rated 1 to 5 by comparing its photo to the three baseline strips. The rating is the average pickup across the whole width the machine covers: a clean center with a stripe of hair down one side doesn't pass, because gaps and weak edges count against it. The one exception is the far edge of the rug by the sign. If a machine wasn't pushed all the way to it, we ignore that edge instead of blaming the machine.Some of the differences between the top machines are hard to see by eye, so we also run the high resolution photos through an AI comparison. A person reviews every result, and the star you see is the one a person signed off on. Each reading carries a short note on what the path looked like.

Equipment. Stop watch. AI agent for image comparison. Black rug, Kapok

Attributed Professional Opinion VRU Staff Opinion

Ratings produced under this methodology are statements of opinion, not statements of measurable fact. They reflect the good faith professional judgment of the named individual, based on extensive first hand observation and customer input, as of the date shown. Individual experiences differ, manufacturer policies and parts availability change over time, and a machine or brand may improve or decline after a rating is written.We publish the full reasoning beside every star precisely so you can judge the basis for yourself, and we say so when an opinion rests on limited information. Where a rating references specific facts, dates, policies, or documents, we cite or link them where available. If a manufacturer or reader believes something we have stated as fact is wrong, we want to hear it, and we correct promptly when shown better information. The opinions, however, remain our own.The absence of a rating on one of these topics is not a judgment against a product. It means we do not yet have enough direct experience to hold an honest opinion, and we would rather show you nothing than a guess.

How we do it. Some of the most important things about owning a machine cannot be put on a test bench. No instrument measures what a warranty phone call is like, whether a repair part will still exist in year eight, or how hard a machine fights the person trying to fix it. For these topics, our ratings are the professional opinions of individual named staff members, formed through extensive observation and customer input gathered in the daily operation of our sales and service centers.The judgment behind one of these ratings draws on, where applicable: hands on repair work across the thousands of machines that pass through our service departments; service check ins and post repair verification; ordering parts across every brand we service and tracking what arrives on time, what runs delayed, and what has quietly disappeared; direct experience with manufacturer warranty and support processes, including years spent operating under manufacturer service contracts; manufacturer documentation, published policies, and warranty terms; and what customers tell us at the counter about their own experiences, including with machines and brands we do not sell.When a staff member scores one of these topics, the questions they weigh are practical ones. What does this repair actually take, and who can realistically do it? Can the part be had, at what cost, and for how many years of the machine's life? What happens to a customer who files a warranty claim, and what did it take to get there? What exists to help an owner help themselves? How has the brand behaved over time, and in which direction is it moving? Scores are anchored: 3 represents average for the industry as we know it, 5 represents the best we know of in that category, and staff are asked to name their reference points so a score can be read against a field rather than in isolation.Every rating of this type is attributed to the specific person whose judgment it reflects, with their role and the basis of their experience stated alongside it. Staff members sometimes disagree. Where they do, each opinion is shown separately under its own name rather than averaged into a false consensus.

Instrumented Bench Measurement

This methodology produces measured values rather than a subjective assessment. Values are directly comparable between machines only when taken at the same measurement point with the machine in the same prepared state, which is why both are recorded with every reading.All readings are taken in Colorado at approximately 5,280 feet, where reduced air density affects any measurement dependent on air movement or pressure. Our figures may therefore differ from manufacturer specifications measured at sea level, and should be compared against our own readings on other machines rather than against published specs.Readings reflect the machine as tested with clean filtration and a known bag state. Performance under real world loading may be materially.Ratings reflect our honest assessment based on the evidence available at the time of publication. Testing data, product specifications, and models change, and errors are possible. Nothing here is a guarantee of performance for any individual machine or household. If you believe a rating is wrong or a detail is out of date, tell us and we will review it.

How we do it. Measurements are taken on the bench under consistent conditions using calibrated measuring equipment appropriate to the value being recorded. The machine is prepared in a known state, typically with a new or empty bag and clean filtration, run at rated voltage, and allowed to reach steady state before any reading is taken.The measurement point on the machine is recorded with every reading, since the same machine can produce substantially different values depending on where the reading is taken.The specific instrument used and the units reported are recorded with the individual measurement rather than fixed by this methodology, so that additional measurement types can be added over time under the same procedure.

Equipment. Calibrated measuring equipment appropriate to the value being recorded. The specific instrument used is recorded with each measurement.

Observational Performance Assessment

We aim to provide the most accurate star rating possible, and for transparency we link to third party test footage and publish our own in store testing video wherever available, so you can review and critique the results yourself.Sources included range from accredited laboratory programs to independent third party testers, our own comparative in store testing, and extended hands on use of the machine in our showrooms and service departments. A source is weighted only in the categories its testing actually measures.Absence of accredited laboratory data is not evidence of poor performance, since certification programs are voluntary and paid, and many capable machines were never submitted.The star value is our own distillation of the weighted evidence. The underlying sources are disclosed on each rating.

How we do it. Vacuum performance testing has no single industry standard, and results vary widely depending on method, test material, and conditions. Rather than comparing raw figures across incompatible tests, we rank machines against each other within each source's own results, using tests performed by the same tester under consistent controlled or semi controlled conditions. Sources are weighted by rigor, with accredited laboratory testing such as the CRI Seal of Approval program carrying the most weight. Our star ratings reflect the combined ordering across all weighted sources for that category.

Equipment. Varies by source.

Sealed Box Airflow Suction Test

How we do it. This test measures what a vacuum does at the cleaner head, the part that touches your floor. Motor ratings and brochure numbers tell you what happens inside the machine. The sealed box tells you what's left by the time the air reaches your carpet, after the filters and every leaky seal have taken their cut.The cleaner head sits on a box that seals around its edges, so the only way air gets in is through one opening of a known size. An anemometer in that opening reads air speed, and air speed times the size of the opening gives airflow in cubic feet per minute (CFM). Block the opening and a gauge reads suction, in inches of water lift.Some heads are built to leak on purpose, with vents for soft carpet or bleeder valves that open in certain modes. When a head won't seal, it gets taped or run in the mode that seals, and the reading says so. If a head still wouldn't seal, that goes on the reading too, because it affects the numbers.Airflow is how much air is moving to carry dirt into the machine. Suction is how hard it pulls when the air is choked off. Plenty of machines are strong in one and weak in the other, so we show both. Readings from the same box compare cleanly. Other testers' boxes differ a little, which is why every reading names who took it.

Equipment. The core apparatus is a sealed plywood airflow box built from 3/4 inch stock at 18 by 18 inches, glued and screwed tightly enough to hold an air and watertight seal under load. It uses a reversible dual-chamber design, with one large sealed compartment sized for full-size American vacuums and a separate smaller compartment for European and compact machines, selected simply by flipping the box over. Each face carries a measurement hole where the vacuum nozzle seals against the box and where an airflow meter reads velocity through the opening.