How to assess a property location before you buy
Buying a home is the largest single decision most people make on the thinnest intelligence. The listing tells you about the house. Almost nothing tells you about the location — and the location is the part you cannot renovate. When we started looking for a house ourselves, we did what we do for clients: treated the address as an intelligence target. A crime baseline, a natural-hazard read, and — the layer that surprised even us — the political one. What follows is the method, run on our own house hunt — including the commune it ruled out. The communes stay anonymous; the checks do not.
Where does a property assessment actually start?
Not with the property. It starts with the decision: what are you actually choosing between, and what would change your mind? “Which of these two houses” is a different question from “which commune,” and both are different from “which canton has the schools we want at a crime level we can live with.” Name the real question first, because each layer below is read against it. Then establish the baseline — what is normal for this place. A quiet street is only meaningful if you know whether it is quiet the way a stable neighbourhood is quiet, or quiet the way a street is quiet the year before a development project breaks ground. The same discipline as any risk read: no judgement without a baseline to grade it against.
One input belongs in the frame from the start, because it changes what you can afford: tax. In Switzerland the bill is a location decision, not a national one — income and wealth tax is set by canton and multiplied by a communal coefficient, and vehicle tax, family deductions and church tax vary the same way. Two communes fifteen minutes apart can differ by thousands a year on the same salary. That difference is worth knowing before you fall in love with a house, not after.
In our case the frame was two questions stacked: first which commune — we compared a village commune in the lower Rhône valley against Lausanne, where we had a baseline of lived experience — and only then which property within it. Running the commune comparison first mattered, because it is the layer where the biggest differences hide.
What does crime data actually tell you — and what doesn’t it?
Less than the rankings suggest, and more than most buyers extract. Aggregate “safest towns” lists are close to useless: they average away exactly what you need. What matters is granularity and rhythm — which offence types, on which streets, at which hours, moving in which direction over five years. A commune with average statistics can contain a station district that produces most of them and a hillside that produces none. And the direction of change outweighs the level: a low-crime area trending up is a different purchase from a moderate one trending down. The honest question is not “is this town safe?” but “what is normal here, and is normal moving?”
Our comparison used the federal police crime statistics at commune level — not the cantonal aggregate, which averages a city and its villages into the same number. The gap was larger than intuition suggested: on the two most recent statistical years, the urban commune ran at three-and-a-half to four times the offence rate of the village commune per 1,000 inhabitants. Both are, by any international standard, calm places — which is precisely the point. “Calm” is not one thing. Two addresses a Swiss buyer would file under the same word sat a factor of three apart on the measured baseline, and only the commune-level number showed it.
How do you read a natural-hazard map?
In Switzerland the cartography is excellent and public — cantonal hazard maps grade flood, landslide, rockfall, and debris-flow exposure parcel by parcel. The data is not the problem; reading it is. Zones carry codes a layperson skims past, and the difference between them is the difference between an insurable nuisance and a parcel you should walk away from. Two checks matter beyond the map itself: whether the map has been revised recently (hazard maps are re-drawn after events, and an old sheet can understate a known risk), and what the building’s insurance history says — the claims record is ground truth the cartography can lag behind.
And the hazard layer is not only nature’s. One area we assessed sits across the river from a major chemical-industrial platform. That is a risk with its own federal regime — the ordinance on protection against major accidents (OPAM) defines consultation perimeters around such sites, and they are queryable in the public cadastre of public-law restrictions (RDPPF), parcel by parcel. A listing will mention the river view. It will not mention which side of an OPAM perimeter the parcel sits on, what the site’s population-protection plan asks of residents, or how the prevailing wind runs. All of it is public; none of it is volunteered.
What is the political layer nobody checks?
The one that found us. Every commune publishes its planning intentions — zoning plans, revision schedules, development concepts, public-inquiry notices. Almost no buyer reads them. Yet a single line in a zoning revision — a parcel reclassified from single-family to multi-storey residential — rewrites a property’s future: the view, the light, the traffic, the resale story. This is primary-source work at its purest: the information is not hidden, not paywalled, not speculative. It is sitting in the public record, in the minutes of a planning meeting, years before the crane arrives. The same file tells you about the planned road, the school extension, the commercial zone creeping closer. No listing will ever mention it, because the seller’s agent has no duty to read the commune’s drawer — and often hasn’t.
This is the layer that decided one for us. In a commune we assessed — and ultimately ruled out — there is a quiet lane of single-family houses, the kind of street a buyer reads as settled, permanent, done. The commune’s public record says otherwise. A global revision of the zoning plan has been under way for years: blocking zones in force since 2022, the draft plan published for public consultation that autumn, cantonal approval still pending. And in the draft, the sector around that lane moves from single-family zoning toward small collective buildings. Nothing about the street announces it. But buy there, and over the following decade the neighbourhood you paid for — the houses, the gardens, the parking, the traffic count — is replaced around you by a denser one, at exactly the moment you would want to resell into it. Set that beside the other layers — the crime trend, the school picture, the industrial platform across the river — and the layers did what layers do when they agree: they closed the question. Not a bad commune in the abstract; the wrong purchase against our decision frame, knowable entirely from public sources before a single visit. One honest caveat, in our own method’s terms: the sector-level reading is ours, made from the published maps — the decisive documents here are cartographic, and a desk read of a map is an inference until the commune confirms it in writing. Any buyer can put that one-letter question to the urbanism service, which must answer what applies to a parcel today and what the pending revision foresees. Almost none do.
How do schools fit a location assessment?
As a comparative question, answered early — it sorts communes before you ever visit a house. But grade the sources: official performance data where it exists, capacity and catchment rules (a good school you cannot enrol in is scenery), and the trend line again — a school system is a slow variable, which makes its direction unusually reliable once you establish it. For an expat family the catchment question doubles: local system, bilingual stream, or international school, each with its own geography and its own waiting list. The school layer is where “which city” gets decided; the crime and hazard layers then vet what the school layer proposed.
What can only feet tell you?
Two things the record never holds. First, the neighbours. A short doorstep conversation — honest, unhurried, before you commit — surfaces what no register contains: the dispute over the hedge, the music every evening, the smell in the stairwell, the neighbour everyone quietly routes around. Ask who has lived on the street longest, and what has changed. People tell you. Second, the routes your life would actually run — to the office, the school, the motorway ramp — walked or driven at the hours you would run them, because the 8 a.m. school corridor and the Sunday-afternoon viewing are two different streets. This is the same discipline as assessing a route through a high-risk city, pointed at an ordinary commute: for clients we grade each route with ground imagery, timed runs, and an independent alternate. The free version is your own feet and an honest notebook.
What does the assessment leave you with?
The same thing any of our assessments leaves you with: not certainty, but a graded, honest picture — a baseline for the place, each risk layer checked against its primary source, and an explicit ledger of what is confirmed versus what remains open. Some findings kill a purchase. Most do something quieter: they change the price you are willing to pay, the questions you ask the seller, or the clause your notary adds. The location is the part of the house you cannot renovate — but it is also the part you can fully assess before you sign, because nearly everything that will happen to a neighbourhood in the next decade is already written down somewhere public. The work is knowing where, and the discipline is showing what you checked — including the checks that came back unconfirmed.
Want to run this yourself? The Location Dossier is free: thirteen checks with where each one is written down, plus the FR/EN letter template that obliges the commune to answer. And this is the method we run for clients relocating or buying across borders — composed, like every Aegilo report, for the specific decision in front of you.