What Truly Makes Two Properties Comparable?

The same price per square metre does not mean the same property. A newly renovated flat on a quiet side street and a run-down one of the same size on a busy main road can show up with an identical price per m² in the statistics, and be hopeless to compare when a concrete valuation has to hold up. The number is the same; the properties are not.

This is where most valuations are really won or lost. The calculation itself, sale price divided by area, is trivial. The professional discipline lies one step earlier: in deciding whether two properties even belong to the same comparison basis. A price per m² inherited from a sale that is not comparable bleeds into the entire valuation, however neatly it is calculated. Comparability between properties is therefore not a technicality, it is the factor that decides whether a comparable is valid or noise.

The four axes that determine comparability

When you need to judge whether a sale can be used as a comparable, four axes carry most of the explanation in practice: use, location, condition, and age/area. They are not of equal weight, use and location are hard filters, while condition and age are adjustment variables, but none of them can be left out.

Use, the hard filter

Use comes first, because a difference here can rarely be adjusted away. An owner-occupied flat, a terraced house and a detached single-family house are priced in fundamentally different markets, even on the same street. The same applies to the distinction between a year-round home and a summer cottage, and between residential and commercial. Use is recorded in BBR (the Buildings & Dwellings Register), and it should be the first criterion that screens out a sale, not an adjustment you add afterwards.

Watch out for mixed-use properties. A property registered with both residential and commercial floor area does not necessarily follow either of the two pure markets, and a price per m² drawn from such a sale should be treated with considerable caution.

Location, at the level that actually matters

Location is the second hard filter, but it needs to be understood more finely than “same town” or “same postcode”. The relevant unit is often smaller: the neighbourhood, the specific street, even the side of the street. Traffic noise, view, distance to a station, a school or a green space can separate two properties a hundred metres apart. Two sales in the same postcode can belong to entirely different sub-markets.

This is also where timeframe and search radius come into play: the wider you search geographically, the more sales you get, but the more you mix sub-markets together. That trade-off deserves its own treatment, and we have written about it in Radius and timeframe: how wide should you search for comparable sales?.

Condition, the most overlooked adjustment variable

Here lies the classic trap from the introduction. Two properties with the same area, use and location can differ markedly on condition, and the difference in condition is rarely visible in the raw sales data. A thoroughly renovated home and one that needs everything appear identical in a price-per-m² table, yet they represent completely different products.

That is why serious valuations work with explicit condition levels rather than a loose estimate. By splitting into several steps, from neglected through standard to fully renovated and near-new, you force yourself to place both the valuation object and each comparable on the same scale, so the difference in condition becomes a deliberate adjustment and not a blind source of error.

Rule of thumb: if you do not know the condition of a comparable sale, you do not know its real comparability either. A price per m² without condition context is a number without an address.

Age, area and physical features

Year of construction, gross floor area, plot size, number of rooms and particular physical features (basement, balcony, parking, energy label) are the fourth axis. Most of these can be adjusted for, a smaller home typically has a higher price per m² than a large one of the same type, but only if you know them. And here it is worth remembering that register data does not always match reality: the area concepts in BBR can be misunderstood, and misregistrations are more common than many people think. We have collected the typical pitfalls in BBR data in property development: what it shows, and where it gets things wrong.

Comparability is a gradient, not a binary test

It is tempting to treat comparability as a yes/no question: is the sale a valid comparable or not? In practice it is a gradient. A perfect comparable, same street, same use, same condition, sold last month, almost never exists. The realistic goal is to gather sales that match on the hard axes and that you can adjust for on the soft ones.

It also means you have to be able to explain every single comparable. Why is this sale included? What is it adjusted for, and why? A comparable you cannot justify is a comparable you cannot defend, whether to a client, a bank or a counterparty. The selection discipline itself, which sales make the cut and which are weeded out, we cover in How to select comparable sales for a credible valuation.

Quantity and quality go hand in hand

The tighter you filter on comparability, the fewer sales remain. It is a real dilemma: a handful of near-perfect comparables gives a sharp but fragile valuation, one outlier shifts the average markedly. A broader selection gives statistical robustness, but at the cost of precision per sale.

There is no single right number, but there is a professional minimum below which a price per m² no longer holds. That balance, between a few precise and many robust, is worth understanding before you trust a figure, and we go into depth on it in How many comparable sales does a price per m² need?.

In practice this means comparability and quantity must be assessed together. A valuation built on three sales requires those three to be close to the object on every axis. A valuation built on thirty can tolerate greater spread, but only if the spread is known and controlled, not hidden in an average.

How Arcili automates the manual exercise

This whole discipline, screening out on use and location, placing each sale on a condition level, adjusting for age and area, and judging whether the quantity holds, is professional work that takes time every single time. It is exactly the exercise that Arcili’s Boligvurdering (Valuation) puts into a system: it pulls comparable sales from the surrounding market, weighs them on 72 factors spread across the axes described here, places the object on one of seven condition levels and gives a confidence level that reflects how solid the comparison basis actually is. The valuation is made on the existing market, not on a forecast, and the underlying sales can always be reviewed one by one, so you can defend every comparable.

The professional judgement still rests with you. But the time-consuming part of finding, filtering and adjusting the right sales need not be done manually from scratch every time.

Want to see what it looks like on a specific property? Book a walkthrough, and we will show you the basis behind the number.