How accurate is a LiDAR scan of a home?
There is no single figure that holds for every room — accuracy depends on the room and on how you scan it. Here are the nine things that wreck it, and the one measurement that reveals whether your scan can be trusted.
Short answer: the accuracy of a LiDAR scan cannot be written as a single figure that holds for every home — it depends on the room, the light and how you move through it. So the only honest answer is: measure it afterwards. Take a tape measure, measure one long wall in the biggest room, and compare it with the figure in the scan. If that one measurement checks out, the rest usually does too. If it doesn't, there is almost always one of nine known causes — mirrors, glass, darkness, glossy floors, moving too fast, an interrupted scan, furniture along the wall, walls that aren't square, or high ceilings.
Why this article doesn't give you an accuracy figure
It would be easy to write "±1 cm" and leave it at that. But an accuracy figure for a LiDAR scan of a home is only true for the room it was measured in. A LiDAR scan of an empty, well-lit living room with matte paint on the walls and a scan of a bathroom with a mirrored wall and glossy tiles are two very different disciplines — same phone, same app, wildly different accuracy.
On top of that, a LiDAR scan has two kinds of accuracy that often get confused:
- Accuracy of a single measurement — how close the scan's figure for one wall length is to reality.
- Accuracy of the whole plan — whether the rooms sit correctly in relation to each other once you've walked through the entire home.
The two fail in different ways. One measurement can be completely correct while the whole plan slowly rotates, because the phone lost track along the way and picked it back up slightly off to one side. This is called drift, and it's the error that grows the further you walk. That's why a scan of a single room is almost always better than the last room in a scan of an entire house.
How a LiDAR scan works — and where it can go wrong
LiDAR sends out short pulses of light and measures how long they take to come back. The distance to each individual surface is therefore measured, not guessed from an image. That's its strength. The weakness lies elsewhere: the phone also has to know, at every moment, where it is in the room, and it works that out by recognising details in the camera image from frame to frame. Lose the details, and it loses itself — and then the walls move.
Everything that wrecks accuracy in a home scan hits one of these two things: either the light that doesn't come back correctly, or the phone that loses track of its own position.
The nine things that wreck accuracy
1. Mirrors
A mirror bounces the light pulse onward instead of reflecting it straight back. The phone therefore measures the distance to the reflection, not to the mirror. A mirror above a sink can give a bathroom a phantom room behind the wall, and a mirrored wall in a wardrobe can make the room look twice as deep. Mirrors are the single fault that produces the wildest results in a home scan.
What to do: cover the mirror with a towel, or scan the room without pointing straight at the mirror, and correct the room's measurements with a tape afterwards.
2. Windows and glass
Clear glass lets a good deal of the light straight through. A large patio door can end up being measured as the garden behind it — or not measured at all, leaving a hole in the wall. Windows in a scan are rarely accurate in the glass itself; it's the frame around it that carries the measurement.
What to do: draw the curtains, or scan the window wall at an angle from the side so you capture the frame. If it's evening and dark outside, the pane also starts behaving like a mirror — see point 1.
3. Darkness
The distance measurement itself works fine in the dark — that's one of LiDAR's whole points. But the phone's position tracking doesn't, because that relies on the camera image. Scan a dark basement or a room in the evening with a single lamp on, and the distances may be fine while the rooms drift out of place relative to each other.
What to do: turn on every light, including in rooms you're just passing through.
4. Glossy floors and shiny surfaces
A freshly varnished wood floor, glossy tiles, a polished worktop or stainless steel all act like a half-mirror. The light skates across the surface instead of bouncing straight back, and the measurement becomes uncertain or is missed entirely. Glossy floors typically produce a room where the floor height tilts — and therefore a storey height that doesn't add up.
5. You're walking too fast
This is the most common cause of a poor home scan, and it costs nothing to fix. Move the phone fast and the camera images blur, leaving the phone with fewer details to recognise. The plan then drifts. Walk at a strolling pace — slower than you'd think necessary — and turn corners smoothly rather than swinging the phone.
What to do: walk along the walls in a calm loop, hold the phone at roughly chest height, and let the camera see both the floor edge and the ceiling edge. Pause at every corner for a couple of seconds.
6. You close the scanner between rooms
Stop the scan in the living room and start a new one in the kitchen, and the app has no way of knowing how the two rooms relate to each other — it has lost the unbroken chain of positions. Each room ends up correct on its own and wrong together.
What to do: keep the scan running while you walk through the door. Move slowly through the doorway and let the camera see the door frame on both sides — it's the door frame that stitches the two rooms together.
7. Furniture along the wall
A bookcase, a cabinet or a sofa in front of the wall means the scan never sees the real wall there. Sometimes the front of the furniture becomes the wall; other times the app guesses at the wall behind it. A room with bookcases on three walls is the hardest room in the home to scan accurately.
What to do: scan at two heights — one pass along the floor edge and one pass higher up, where the wall is clear above the furniture. If the wall is completely covered, measure the room afterwards with a tape.
8. The room isn't square
Older homes rarely have four right angles. A scan measures the crooked reality, but afterwards that crookedness has to be translated into a drawing you can actually use — and that's where a choice has to be made: should the wall be straightened, or should the skew stand? Straighten it, and the drawing looks tidier while the area is slightly off. Leave it, and the area is right while the drawing looks messy. A room that's 3 cm wider at one end than the other is completely normal in a house from the 1930s.
9. Large rooms and high ceilings
LiDAR is a rangefinder with a range. The further away the surface is, the less certain the measurement — and a high-ceilinged room or a large open living room forces the phone to measure across long distances. The same goes for sloped walls and a collar-tie roof, where it's a long way up to the ridge.
What to do: get closer. Scan large rooms in two passes, so you're never standing more than a couple of metres from the wall you're measuring.
The check: one tape measure, one measurement, two minutes
Here is the test that decides whether your scan can be trusted. It takes two minutes, and it's worth far more than an accuracy figure from an article:
- Pick the home's longest unbroken wall in the biggest room. A long wall matters: a 2 cm error is a lot over 1.5 metres and very little over 6 metres, and it's the relative error you want to know.
- Measure it with a tape from wall to wall — that is, the internal measurement, at floor height. Write the figure down in centimetres.
- Look up the same measurement in the scan.
- Subtract, and convert the difference to a percentage: difference ÷ measured length × 100.
Example: the tape measure says 512 cm. The scan says 516 cm. The difference is 4 cm. 4 ÷ 512 × 100 = 0.8 %. You can now apply that percentage to the rest of the home: if the living room is 24 m² in the scan, expect the true area to be around 0.4 m² either side of that — the area error is roughly double the length error as a percentage, because the area is two lengths multiplied together.
If that one measurement falls within what you can live with, you can trust the rest of the scan. If it's badly wrong, look for the cause in the list above before scanning again — it is almost always a mirror, a glass wall or an interrupted scan, not the phone.
How accurate does it actually need to be?
The question "how accurate is a LiDAR scan" only makes sense once you know what the measurement is for. The three typical purposes call for very different levels of precision:
- House hunting and getting an overview: you want to know if the sofa would fit, and whether the home matches the listing. A scan is more than enough here — you're comparing rooms and getting a feel for sizes.
- Materials and contractor quotes: paint, flooring, plaster, tiles. Waste gets added on top here regardless, and a scan gives a solid basis for a quote. The contractor measures again on site before ordering.
- Something that has to be cut to size: a kitchen unit between two walls, a sliding door, a worktop. No scan is good enough here, and that goes for professional scans too. That one measurement gets taken with a tape or laser on site, every time.
This isn't a weakness specific to LiDAR. It's the same rule that applies to an old floor plan from an estate agent: good to plan from, never good enough to cut to.
How HouseSense handles it
HouseSense calculates in internal measurements — from wall surface to wall surface, exactly as when you stretch a tape measure across the room yourself. That means the figure you can check with a tape is the same figure the app shows you. That sounds obvious, but it isn't a given: official floor areas are measured externally in most countries — via the public building register — so the official area will always be larger than the sum of the rooms you can actually walk around in. That gap is not an error in the scan.
HouseSense also doesn't state a claimed accuracy in centimetres, and that's a deliberate choice: the app can't measure any better than what the phone delivers in that particular room, and a figure that can't be held to is worse than no figure at all. The principle behind the app is that the drawing may be tidied up, the figures may not. Draw the home by hand instead, using your own tape-measure figures, and the accuracy is exactly as good as your tape and your reading of it.
The short checklist
Want a LiDAR scan of a home as accurate as possible: turn on every light, cover mirrors, draw the curtains, scan at a walking pace, keep the scan running through doorways, do two passes in large rooms and rooms with bookcases — and afterwards check one long wall with a tape measure. That single follow-up check tells you more about your own scan than any general accuracy figure ever could.