Two things, and they are one argument: how we establish that two listings are the same house, and which surface we divide prices by so they can be compared. The first is the evidence that the second is needed.
These pages compare listings from different agencies for the same property. Every figure is the agency's own, linked to its listing, so any reader can check it in one click.
No estimate appears on these pages. No OMI bands, no negotiation discounts, no valuation of ours. We compare the agencies to what they themselves published.
The index below is a different thing, and we say so plainly: to compare listings with each other we have to bring them to one surface, and that conversion is a declared choice of ours. Which is why it is published as a range, never as a single number, with the rule written out below.
| Reference number | The agencies use the same reference. Their identifier, not our inference. |
| Photographs | At least two identical photographs, compared by eye one pair at a time. A single shared image is not enough: agencies reuse pictures, and the same view can be photographed from different houses. |
| A non-round price | A figure like € 110,625 does not occur twice by accident. A round price on its own proves nothing. |
| Descriptions | Where two listings describe the same house unmistakably — same structure, same history, same grounds — even with no photographs in common. |
If you are an agency and believe two listings shown here are not the same property, write to us and we will re-check. If you are right the page comes down; if the data has changed it is updated. Either way we reply within 7 days.
The same applies in reverse: if a figure here is wrong because the listing has since been corrected, tell us.
Write to: correzioni@casazebra.it — see also Report an error.
The data is re-read from public sources periodically. A page reports what was published at the time of reading, and listings change: the link to the source is always the reference.
Every Italian property portal publishes a price per square metre. None of those numbers can be compared with each other, because the denominator — the surface — is defined differently by every agency, and is almost never written down anywhere.
The case that proves it is in our own archive. Two agencies ask the identical price for the same villa in Anghiari and publish €508/m² and €3,265/m² — six times apart. Not because they disagree about the value, but because one divides by a "commercial surface" of 3,150 m² that includes a 2,600 m² park. With the land removed, the real disagreement is between 550 m² and 490 m², about 12%. That is a question a buyer can put to an agent. The 6.4× never was.
One rule does most of the work: land never enters a surface figure. Gardens, parks and agricultural land are reported separately, as what they are. Nobody walks on a park.
Our reference surface is the internal habitable area — internal floor area measured to the inside face of the external walls. It excludes accessories and all land.
Beside it we always publish the agency's own price per m², calculated with the surface the agency itself stated. The two figures sit side by side on every page. The gap between them is the point.
This table is a declared choice, not a measurement. We did not derive it from the market: we wrote it down, and we apply it identically to everyone. That is precisely what the agencies do not do. The garage row (50%) and garden row (10%) match the rule Immobiliare applies to its own listings — inconsistently, from one agent to the next; the remaining rows are ours.
| Component | Weight | Cap |
|---|---|---|
| Internal habitable area, h ≥ 2,40 m | 100% | |
| Sub-height space, 1,50–2,40 m | 50% | |
| Below 1,50 m | 0% | |
| Habitable annex / dependance | 100% | |
| Veranda, enclosed and unheated | 60% | |
| Covered loggia / portico | 35% | |
| Open balcony / terrace | 25% | ≤ 25% of sia |
| Cantina / magazzino, non-habitable | 25% | ≤ 15% of sia |
| Garage / box | 50% | ≤ 40 m² |
| Covered parking space | 30% | ≤ 25 m² |
| Ruin / unrestorable annex | 0% | reported separately |
| Private garden ≤ 1.000 m² | 10% | ≤ 15% of sia |
| Land / park > 1.000 m² | 0% | reported as land |
| Pool, tennis court, well, vineyard | 0% | reported as amenities |
This table is not yet in use. Applying it requires an itemised breakdown of the surface, which public listings do not carry. We publish it anyway, because an agency is entitled to see the standard it will be measured against before we are able to measure it.
Publishing a confident number from an unconfident input is exactly what we object to. So every listing states how solid its index is.
| A — measured | Surface breakdown available. A point value. No listing currently reaches this level. |
| B — inferred | One surface figure and a typology. We publish a range, never a number. |
| C — insufficient | No price, no usable surface, or a typology too variable to infer from. No index at all. The page shows the agency's figures and explains why there is no number of ours. |
For Tier B listings we convert the stated surface to internal habitable area like this:
| Type | Factor |
|---|---|
| Apartment | × 0.78 – 0.88 |
| Whole-building house | × 0.72 – 0.85 |
| Restored casale / farmhouse | × 0.65 – 0.85 |
| Terratetto / townhouse | × 0.72 – 0.85 |
Villas with grounds have no factor, and that is not an oversight. For that category the conversion would run from 0.30 to 0.80 — a range wide enough to say nothing. It is also the category where the problem is worst. So a villa is either broken down by hand or it carries no index. We would rather publish a declared gap than a number that does not hold.
Every listing enters its comune's band as an interval, not a point, and it stays an interval all the way to the page: what we publish is X to Y, never their midpoint.
A band is published only with at least 8 listings from at least 2 different agencies, and only if the central interval is no wider than 28.0% — that is, no wider than the uncertainty of the single worst conversion feeding it. Below that threshold the comune has no band, and we say so.
With every band we publish two widths: how uncertain our conversion is and how much the market itself varies. Across the Valtiberina the first runs 13–27%; the second runs 52–98%. A reader should be able to see how much of the range is us and how much is the market — and here the market is by far the larger.
It is not a survey or an appraisal, and it does not claim to be. In Italy a perizia is a regulated act carried out by a qualified technician, on site. This is an index: figures the agencies have already published, brought onto one surface by a written rule, so that they can be compared with each other.
We do not say what a house is worth and we do not say what to offer. We say what was published, and what happens to those numbers when the same measure is applied to all of them.