Price per Square Meter: How to Calculate and Read It

A price per square meter looks like the simplest figure in the entire property industry: price divided by area. But precisely because it is so easy to calculate, it also gets used far beyond what it can actually bear. An average of DKK 32,000/m² for “the area” can cover anything from a freshly renovated detached house on a quiet street to a run-down flat facing a main arterial road — and an average tells you nothing about which of the two most closely resembles your property.

For anyone who has to defend a figure in front of a credit committee, a seller, or an investment committee, the question is therefore rarely “what is the price per square meter” — but “what lies behind it, and can I trust it”. This guide walks through how a price per square meter for an area is actually calculated, what pulls it up and down, and why the number of sales behind the figure is often more important than the figure itself.

What a price per square meter actually measures

A price per m² is the price of a realised sale divided by the property’s area. The first source of error appears right here, because which area? BBR (the Buildings & Dwellings Register) works with several area concepts — living area, total building area, basement, utilised loft — and they are not interchangeable. If one sale is calculated on living area and the next on total area including the basement, you are no longer comparing like with like. A price per square meter is only meaningful if the numerator and denominator are defined the same way across every sale in the dataset.

The second thing to keep firmly in mind: a price per m² is built on transaction history — registered, completed sales — not on asking prices. The asking price is the seller’s wish; only once the sale is registered do you know what the market actually paid. An area price based on active listings typically overstates the level, because it counts what sellers hope for, not what they got.

Rule of thumb: A price per square meter without a stated area concept and a stated time period is not a data point — it is a rumour.

How an area price is calculated — and where it goes wrong

An area price is an aggregate: you take a number of comparable sales within a geography and a time period and combine them into a single figure. How you aggregate determines how robust the figure becomes.

Average versus median

A simple average is vulnerable to outliers. A single extreme sale — a fully renovated property sold dear, or a family transfer sold well below market — can skew the average significantly, especially when only a handful of sales sit behind it. The median (the middle observation) is more robust, because it is insensitive to individual extremes. For most area analyses, the median is a more honest starting point than the average, and the difference between the two is itself a signal: if they lie far apart, the dataset is skewed and should be treated with caution.

Geography and time window

Two dials control everything: how large an area, and how far back, you go.

• Too narrow / too short → too few sales → the figure becomes random and unstable.
• Too broad / too long → you blend in incomparable neighbourhoods or outdated price levels → the figure becomes diluted.

The right balance depends on how transaction-active the area is. In a dense urban setting with many similar flats, a radius of a few hundred metres and 12 months may give plenty of basis. In the countryside or in a low-turnover detached-house neighbourhood, you have to widen both the radius and the time window simply to get enough sales — at the cost of falling comparability. Which sales belong in the dataset is a discipline in its own right; we have gone through the criteria in how to select comparable sales.

What pulls the figure up and down

An area price is a weighted result of the properties that happened to trade during the period. As a result, two things move the figure without the market itself having moved at all:

• The mix of sales. If a given quarter happens to see more large, newer homes sold than usual, the area price rises — not because prices have gone up, but because the mix has changed. This is the most common trap in quarter-on-quarter comparisons.
• Condition level. Condition is the factor most people underestimate. Two identical homes on the same street can differ considerably in price on internal condition and quality alone. When you read an area price, you are therefore reading a blended condition level — and your property rarely sits exactly on the average. How much condition actually shifts the figure is something we have covered in condition level and price per m²: what a building’s condition does to value.
• Location at the micro level. Quiet street versus busy road, view, proximity to water, shade from neighbouring buildings. An area price averages across these differences; your specific property does not.
• Area and area type. Smaller homes typically have a higher price per m² than large ones, because part of the value lies in having a home at all, not only in the floor area. An area with many small flats will therefore have a higher average price per m² than a detached-house neighbourhood — without either one being “more expensive” than the other in any meaningful sense.

The point is not that an area price is useless, but that it is a starting point, not a verdict. It tells you the level; it does not adjust for how your property deviates from the average. That step — from area price to property-specific value — is the whole exercise in how a property is valued on the existing market.

Why the number of sales decides whether you can trust the figure

The most important figure about a price per square meter is often not the price, but how many sales it is built on. An average of three sales and an average of three hundred are two entirely different kinds of figure, even if they show the same number.

With few sales behind it, the price is unstable: a single additional sale can move it noticeably, and you have no real way of knowing whether the level is representative or accidental. With many sales, the individual fluctuations smooth out, and you approach a figure that actually describes the market. Every area price should therefore be accompanied by a count — and if the count is low, the conclusion must be phrased as an estimate with corresponding caveats, not as a fact.

This is also where geography and time return as a trade-off: you widen the radius or the time window to get enough sales, but you pay with falling comparability. A good analysis makes this trade-off explicit instead of hiding it behind one tidy figure. A practical rule to remember: always state the price, the number of sales, the geography, and the period together. If one of the four is missing, the figure cannot be assessed — only repeated.

Area price, public assessment, and market value

An area price is derived from the actual market and thus differs from the public property assessment, which is an authority-set basis for taxation with its own model and often marked deviations from real transaction prices. Never confuse the two: the public assessment is not a price-per-m² reference, and why the two diverge is something we have gone through in market value versus the public property assessment. Nor is an area price in itself a market value for a specific property — it is the raw material you adjust from. And it says nothing about where the market is heading; it only describes the realised market over the period it covers.

From a manual calculation to a basis you can defend

Calculating a credible area price by hand means gathering registered sales, standardising the area concepts, choosing the radius and time window, weeding out the incomparable sales, and deciding whether the median or the average is the most honest figure — and then making the count behind it clear. That is hours of work per area, and it is easy to make one of the errors this guide has described.

That is precisely the work the Ejendomshandler (Comparable sales) module automates. You place a point on the map, choose a radius (up to 4 km) and a period (up to 24 months), and get a basis built on 6.3 million comparable sales — with the number of sales behind each figure visible from the start, so you can judge for yourself whether the basis holds. The result can be exported to PDF and dropped straight into a case file. It does not replace your professional judgement of whether the sales are comparable — but it removes the hours of data collection, so you can spend your time on interpretation instead.

If you want to see what an area analysis looks like on a specific address, drop by Arcili or book a walkthrough.