Methodology & trust

We give an independent, free read on whether a property's asking price is realistic. For each valuation we publish a 0–100 Reality Score, a fair-value range (low–high) with a most-likely value, and a confidence level. Estimates are informational — not a certified appraisal or financial advice.

Estimates are built from observable market signals rather than official transaction records (which are scarce here): listing asking prices, neighborhood rents, local income levels, and regional benchmark prices per m². For every neighborhood we show the sample size behind its data and when it was last updated, so you can judge its freshness.

We blend three internationally recognized approaches: the income approach (value implied by rent and a capitalization rate), the sales-comparison approach (price per m² adjusted for size, finishing and floor), and an affordability anchor (local income). The low–high range reflects uncertainty: the thinner the data, the wider the range.

The score (0–100) is a weighted blend of four factors: rental yield, affordability versus income, listing behavior (relisting and price cuts), and neighborhood context. Bands: 80+ realistically priced, 60–79 reasonable, 40–59 questionable, 20–39 likely overpriced, under 20 highly speculative.

Confidence depends on how much data backs an estimate and how fresh it is — not on whether the price is “good.” When a neighborhood has little or stale data, confidence drops and the range widens. The model can't see every property's unique details (view, condition, seller urgency), so treat the value as a guide, not a verdict.

No one can pay to change a score or a value. The platform is free and ad-free, and we don't sell biased valuations. We only ever read Facebook data through the official API with permission — we never scrape posts.

Every report carries a reference you can verify online, and we track our performance in the open on the Track record page by comparing estimates to realized sale prices. If you spot an error in any valuation, the “Report an error” button on the report sends it to us for review.