All articles · 5 min read

How accurate is a floor plan I draw myself?

Exactly as accurate as your measurements — no more, no less. The drawing tool adds no error of its own; the tape measure, your memory, and the skipped intermediate steps do.

Where the errors actually come from

When a floor plan does not match reality, it is almost never because someone drew it crooked. It comes down to one of four things:

• Chain measuring. You measure 80 cm to the edge of the door, then 90 cm for the door, then 60 cm to the corner. Each measurement might be 1 cm off, and the errors stack on top of each other. Always measure from the same corner to every point — one zero point per wall — so the error cannot travel.
• Guessed wall thicknesses. "It's probably around 10 cm." If it is actually 15 cm, every wall is 5 cm wrong, and in a house with seven partition walls that shifts a third of a metre around.
• Rooms drawn square. Old houses are skewed. Draw 90 degrees in every corner and the plan looks tidier, but a kitchen ordered to fit it will not sit flush along the wall.
• Shared walls counted twice. Draw each room on its own and place them side by side, and the shared wall easily ends up counted in both the living room and the kitchen. The home grows by a couple of square metres on its own.

The drawing tool can help with the last one: in HouseSense, the walls of two neighbouring rooms snap together into one shared wall, precisely so the same centimetres are not counted in two places. The other three errors are yours, and only a tape measure catches them.

What different tasks require in precision

"Accurate enough" depends on what the number is for. Rough orders of magnitude:

• Comparing homes, or judging whether the furniture will fit: ±5 cm per room is plenty.
• Buying flooring, paint or tiles: ±2-3 cm per wall length. You add 5-10% for waste regardless.
• Redecorating, redesigning, planning a walkway: ±2 cm.
• Ordering a kitchen, sliding doors, built-in wardrobes or a staircase: a floor plan is not enough here. The supplier measures on site themselves, and they have to, because they are liable for the measurement.

The point: your own floor plan can handle almost everything. It just should not be used as an order measurement for anything that has to be squeezed in between two walls.

The maths that shows how much an error matters

A room measures 4.12 m × 3.05 m. The area is 4.12 × 3.05 = 12.57 m².

Measure 5 cm too much on both sides — so 4.17 × 3.10 — and the area becomes 12.93 m². The difference is 0.36 m², or just under 3%. If you're buying flooring for that room, 3% means nothing, because you're adding 5-10% for waste anyway.

But the error scales. Measure systematically 5 cm too much in every room of a 120 m² home, and the whole home grows by roughly 3% — about 3.6 m². That is an entire small bathroom that does not exist. This is why the systematic error is dangerous, while the random one is harmless. Systematic errors come from measuring at hip height instead of floor height, from measuring to the front edge of the skirting instead of to the wall, and from guessing the same wall thickness everywhere.

The check that catches almost everything: the sum has to add up

Measure the outside width of the house, and subtract the exterior walls. If the house is 9.60 m outside with 35 cm exterior walls, it is 9.60 − 0.70 = 8.90 m inside.

Now add up the internal widths of the rooms and the partition walls across the house: 4.12 + 0.11 + 4.67 = 8.90 m. If it adds up, your measurements are consistent. If it does not, something is missing — a service duct, a chimney, a built-in cupboard, or you have measured a wall wrong somewhere.

That check takes two minutes per floor and catches more errors than any other.

The diagonal: the only way to measure skewness

A square room measuring 4.12 × 3.05 m has a diagonal of the square root of (4.12² + 3.05²) = the square root of (16.97 + 9.30) = 5.13 m. Measure 5.21 m, and the room is skewed.

How much does 8 cm on the diagonal matter? Enough that a 4 m long kitchen worktop gets a gap of a couple of centimetres at one end, and enough that a drawing where everything is square cannot be used to plan around it. Measure the diagonal in every room. It is one extra pull of the tape and the cheapest quality check there is.

Is a scan more accurate than a drawing, then?

A LiDAR scan has one big advantage: it measures what is actually there. Skewed walls, alcoves, chimney breasts and bay windows come along without you having to spot them first. It is not that the scan's numbers are truer than a tape measure's — it is that the scan does not forget anything.

A hand-drawn plan contains exactly what you yourself saw and measured. Be thorough, and it is just as true to scale. Be sloppy, and no sensor can save it. This is also why a hand-drawn home in HouseSense is treated exactly like a scanned one: areas, takeoffs for contractor quotes, the 3D model and reports are all calculated from the same measurements.

Five habits that make your drawing reliable

1. Always measure at floor height, internally from corner to corner, in centimetres.
2. Always measure from the same corner — never in a chain.
3. Measure the diagonal in every room.
4. Measure the wall thickness once per wall type, in a door or window reveal.
5. Finally, check that the rooms plus walls add up to the house's internal measurements.

And remember the one limit no drawing can move: neither a scan nor a hand-drawn plan can determine whether a wall is load-bearing. That takes a building surveyor or an engineer on site.