Company

Built because the data didn't exist yet.

We didn't build another valuation model. We solved one problem: the Turku archipelago has no public register of sale prices. The honesty that demanded became the core of the tool.

Coverage

Parainen to Taivassalo, one comp at a time.

Every added sale sharpens the next estimate. The library grows from agents' own sales.

KustaviKemiönsaariNaantaliTaivassaloParainenNauvoKorppooHoutskariIniö
Municipal boundaries: Statistics Finland (CC BY 4.0)

The problem

Finland doesn't publish sold prices. So we stopped pretending otherwise.

Realised prices aren't available from any single source. Consumer estimates patch the gap with regional data that knows nothing about a single plot.

What was available in April 2026

MML transactions APIbeta, synthetic
KVKL hintatiedotpaywalled
Etuovi / Oikotieasking prices only
Tilastokeskus StatFintoo coarse for one plot
No per-property sold-price source
So the model runs on a manually curated corpus instead

The tool rests on what an archipelago agent already has: own sales, a network, the local papers. Out of those grows a structured, weighted body of evidence. Every added sale strengthens the next estimate.

What we won't do

Three rules we don't bend.

01

Always a range

An estimate is always a range. Single-figure precision doesn't exist on this data.

02

Everything sourced

Every figure has a source, most have a link. You can check them yourself.

03

Refusal over guessing

Under five sales the tool gives no figure. It asks for more sales.