How we test robot vacuums

Every ranking on VacVerdict comes from the same process: real robots, the same conditions, and the criteria that actually matter when you live with one. Here is exactly how we reach our verdicts.

The same conditions for every robot

Each robot goes through the same routine in comparable conditions, so we are comparing real-world behaviour rather than the figures printed on the box. We run every robot across the same mix of hard floor and carpet, scatter the same measured debris — fine dust, oats, flour, larger crumbs and pet hair — and send each one over the same routes through cluttered and open rooms alike. We set them up the way you would at home, with the app installed and the dock placed sensibly, and we let them get on with it. That consistency is what makes a fair comparison possible: a robot that looks impressive on paper but strands itself under the sofa in a real room has nowhere to hide.

What we score

We judge the things that make a difference when you actually live with a robot, not just the spec sheet:

• Cleaning power on hard floor and carpet — how much of the scattered debris each robot actually picks up on both surfaces, and whether it manages the transition from one to the other instead of skating over edges and leaving dust behind.
• Navigation & mapping — whether it covers the whole floor in tidy, methodical rows rather than bouncing about at random, builds a usable map, and finds its own way back to the dock without abandoning half the room or getting itself stuck.
• Obstacle avoidance — how it copes with the real clutter a home contains: phone chargers and cables, stray socks, shoes, and the pet mess that turns a minor miss into a genuine disaster.
• How hands-off it is — how often you have to empty the bin or step in, how much maintenance the mopping function demands, and whether a self-emptying dock genuinely lets you forget about it for weeks.
• Build, app & ease of living with it — how solid the robot feels, how clear and reliable the app is, how easy scheduling and no-go zones are to set, and whether the whole thing is a pleasure or a chore to own.

Each robot's gauges and overall rating on this site come from these criteria, scored from 1 to 5. A high score is not about having the biggest suction number or the highest price — it is about being the best fit for the buyer the robot is aimed at.

Testing each robot for the home it's built for

A central part of our method is judging each robot against the home it is actually designed for. We do not penalise a budget bump-and-turn robot for lacking lidar mapping, because mapping was never part of its job; it is built to keep a small flat ticked over cheaply, and we score it on whether it does that well. Equally, we do not let a flagship off the hook for clumsy navigation just because it is expensive. We judge each robot on its intended buyer, and then tell you plainly which home it suits — and, just as importantly, which home it does not. That is why our reviews always say who a robot isn't for, not just who it is. A robot that is perfect for a tidy one-bedroom flat can be hopeless in a cluttered family house, and the honest thing to do is to tell you which one you have.

We treat spec sheets as claims to test

Manufacturer specifications are a starting point, not the verdict. A suction figure quoted in Pa tells us roughly how much grunt a robot is supposed to have, and a battery figure tells us roughly how long it should run — but both are best-case numbers, measured in ideal conditions at the most favourable setting. So we treat the spec sheet as a hypothesis to test rather than a result to report. A high Pa rating counts for nothing if the brush design flicks debris aside instead of lifting it from carpet, and a generous battery figure means little if the robot wanders inefficiently and never reaches the far rooms. Our scores reflect real behaviour: where a robot's performance matches its claims we say so, and where it falls short of its numbers, that gap is exactly what our hands-on testing exists to catch.

Our independence

We buy the units we review. We are not sent free robots in exchange for coverage, and manufacturers cannot pay for a place or a higher position in our rankings. The order on every page is decided entirely by how the robots perform against our criteria. VacVerdict is funded by affiliate commissions — if you buy through our links we may earn a small commission at no extra cost to you — but that funding never sways a verdict. A commission on a robot does not move it up the page, and a robot we cannot link to is not pushed down. The full detail is in our affiliate disclosure.

Keeping rankings current

The robot vacuum market changes every year as models are discontinued, replaced and rebadged under new names. We revisit our rankings regularly, update prices and availability, and swap in newer robots where they earn a place. If a robot we recommend is discontinued, we say so and point you to the best current alternative rather than leaving a stale link behind. We would rather show a slightly shorter list of robots we genuinely stand behind than pad the page with options we would not recommend, so a model only stays on our list as long as it remains the best choice for its buyer. To see our latest picks, head to the best robot vacuum ranking, and read more about us on our about page.