Data Sources & Methodology

Why we publish our sources

Every number on this site — every price per square meter on the regional map, every occupancy figure in the yield tables, every cost line in the acquisition breakdown — traces back to a source you can check yourself. That's a deliberate choice, and it costs us something: when a source is public, a reader can catch us being wrong. Good. That's the point. Property research aimed at US investors is thick with unnamed "market intelligence," "proprietary indices," and figures that appear fully formed with no parents. We keep a leather ledger, and a ledger shows its entries. This page is the index to ours.

What follows is each source we use, what it actually gives us, how often we pull it, what it costs to access, and — just as important — where it runs out of road. No single source in Austria tells you what a chalet in Sölden is worth. The picture comes from triangulation: official statistics for the ground truth, the land registry for the transactions, listing platforms for the current ask, tourism bodies for the rental demand, and the national bank for the credit weather. When those five agree, we publish with confidence. When they disagree, we publish the disagreement and say so. And at the end of this page, we'll tell you exactly what our data cannot see — because an honest methodology page that skips its limitations is marketing wearing a lab coat.

Statistik Austria: the ground truth, once a year

Statistik Austria is the national statistical office — the Austrian counterpart to the US Census Bureau and the Bureau of Economic Analysis rolled into one — and it is the spine of our price work. The series we use most is its residential property price data: median transaction prices by district, built from actual registered sales rather than listings or appraisals. For Tyrol, that means district-level medians for Innsbruck-Stadt, Innsbruck-Land, Kitzbühel, Imst, Landeck, Lienz, Reutte, and Schwaz — the administrative geography that contains every resort and valley we cover.

The cadence matters, so here it is plainly: the data updates annually, and it arrives late. The 2023 transaction medians were published in October 2024. So for part of every year, the most recent "official" truth is more than a year old, and the freshest numbers are always ten months stale on publication day. We don't pretend otherwise. What Statistik Austria gives us is the anchor — the slow, reliable, methodologically clean series against which everything faster gets validated. When our listing-platform data suggests Innsbruck apartments are asking 8% above last year's medians, it's the Statistik Austria median that defines "last year." When we say the Kitzbühel district trades at the top of the Tyrolean range, that's the official series talking, not our hunch.

Access is mostly free. The headline series are published on the office's open data portal; some detailed tables sit behind modest paid downloads. Anyone reading this from Ohio can pull the same district medians we use, and we encourage exactly that. If our price map and the official medians ever part company without explanation, the map is wrong — and our corrections page exists for that event.

The Grundbuch: every transaction, one query at a time

Austria's land registry — the Grundbuch — is the most underused data source available to a US buyer, and the most definitive. Every property transaction in Tyrol is recorded here, because under Austrian law a sale isn't legally complete until the registry entry happens. Notary-reported, court-maintained, parcel by parcel: ownership, encumbrances, easements, and the fact of transfer. If a chalet changed hands in Neustift in March, it's in the Grundbuch. There is no off-the-record in Austrian conveyancing — only off-our-radar, which is a different problem we'll address in the limitations section.

The catch is access. The Grundbuch is a registry, not a statistics service: you query it property by property, at roughly €20 to €50 per query depending on what documents you pull and whether you go through the official portal or a professional intermediary. There's no bulk download of "all Tyrolean chalet sales 2024." So we use it the way a careful appraiser does — for targeted verification. When a benchmark property trades, when a comparable sale anchors a valuation, when we need to confirm that a "sold" listing actually closed, we pull the registry extract. It is the slowest source we use and the only one that constitutes legal truth.

For our site, the Grundbuch feeds the comparable-sales checks behind the price map and the validation layer beneath the yield assumptions. A listing platform can tell you a seller wants €11,200 per square meter in the Kitzbühel Alps. The registry tells you what buyers actually paid last quarter on the street you're looking at. The distance between those two numbers is where buyers either save money or donate it.

ImmoScout24 and Willhaben: the asking-price layer

If Statistik Austria is the anchor and the Grundbuch is the ledger, the listing platforms are the weather vane. Two dominate Austrian residential property: ImmoScout24 and Willhaben, the latter a general classifieds giant whose property section is genuinely the larger of the two in several Tyrolean districts. Together they carry the overwhelming majority of publicly marketed residential listings in Tyrol, from Innsbruck studios to Ötztal lodges.

We track asking prices on both platforms for our six measurement points — more on those below — and we treat the data with the skepticism it deserves. Asking prices are aspirations. So every platform figure we use is validated against transaction data, and here's the relationship we've measured consistently: asking prices in Tyrol typically run 8% to 12% above closing prices. The spread widens at the top of the market — a €4 million lodge has more negotiating air in it than a €315,000 studio — and narrows in hot micro-markets where multiple bidders compress the discount. When you see a price on our site derived from listings, it has been adjusted downward by that measured spread unless we explicitly label it as an asking figure.

What the platforms give us that nothing else can: speed and texture. New supply shows up here months before it appears in any statistic. Price reductions — a seller cutting €90,000 off a Stubaital chalet after ten weeks — are visible in near real time and tell us where the market's resistance actually sits. And the listing text itself carries information no database captures: whether a property is marketed as a Zweitwohnsitz (second home, with its regulatory baggage), whether short-term rental permission is claimed, whether the roof is from 2003 or 1974. We read listings the way brokers read them. You should too.

Tourismusverband Tyrol: the demand side of the yield story

A property's price is a capital question. Its yield is a tourism question. For the rental side of our analysis, the primary source is the Tourismusverband Tyrol — the regional tourism association — which publishes occupancy and overnight-stay data by region, season, and accommodation category. The headline figure for 2023: 52.4 million overnight stays across roughly 80,000 commercial beds in Tyrol. Let that ratio sink in, because it's the entire investment thesis for the province in one line. Tyrol's commercial bed base is small relative to its demand, which is why well-located rental property achieves occupancy figures that make US vacation-rental markets look sleepy.

We use the tourism data at the regional level, not the provincial headline. The 52.4 million is the whole of Tyrol; what matters for a Sölden lodge buyer is the Ötztal's winter and summer split, the average length of stay, and the bed-occupancy rate in the commercial category — because a private apartment competes with those 80,000 commercial beds for the same guests. The tourism association's regional breakouts feed our occupancy assumptions on the seasonal yield page: when we model a chalet at a given annual occupancy, the figure is anchored to the published regional rate for its accommodation class, adjusted for the discount that a single private property suffers against professionally marketed commercial beds.

One caution we apply internally and repeat here: tourism statistics count commercial accommodation. Your chalet is not in the sample. The association's numbers tell you the tide; they don't guarantee your boat floats. A well-priced, well-presented property in a strong season beats the regional average; a stale listing with 2019 photos sinks below it. We model conservatively, and we'd rather hand you a yield figure that disappoints upward than one that requires heroics to achieve.

The Oesterreichische Nationalbank: credit weather and the foreign-buyer picture

The Austrian National Bank — the OeNB — is our source for everything monetary: interest rate data, bank lending standards, and the official estimates of foreign buyer share in Austrian residential transactions. When we state on the acquisition costs page that non-resident borrowers see 60–70% loan-to-value limits and 10-year fixed rates in the 2.8–3.5% band, the rate environment and the lending-standards backdrop come from the OeNB's publications and its regular lending-standards surveys of Austrian banks.

The foreign-buyer estimates deserve a paragraph of their own, because they're the least precise figures we use and the most frequently misquoted. Austria restricts property purchases by non-EU buyers — Tyrol more strictly than most provinces — and the OeNB tracks the resulting flows as part of its financial-stability work. The estimates are exactly that: estimates, built from approval data, registry analysis, and survey work rather than a clean census. We cite them as ranges, never as point values, and we flag the confidence level. If a source's uncertainty is high, the honest thing is to say the number is soft and explain why we'd rather show you a soft number labeled soft than a hard number that's fiction.

Practically, the OeNB feeds two things on this site: the financing assumptions in our cost models, and the macro context in our market commentary — when Austrian bank lending to residential buyers tightens or loosens, transaction volumes in Tyrol follow with a lag, and we'd rather tell you that in advance than explain it afterward.

The six measurement points

All of the above is collected at six locations we chose deliberately and revisit on every refresh cycle. They're not the six prettiest places in Tyrol. They're the six that, together, represent the investable market across price tiers and demand profiles.

Innsbruck is the urban anchor: year-round economy, university, the only market in our set where locals, not tourists, set the floor. €8,400 per square meter on our map. Kitzbühel is the prestige tier: international buyers, the steepest prices we track, €11,200 per square meter, and a winter season that drives the entire annual yield. Sölden represents the high-altitude resort market in the Ötztal — glacier skiing, a long season, strong commercial occupancy. Obergurgl, its higher, quieter neighbor, is our snow-reliability benchmark: when lower resorts have a thin January, Obergurgl's numbers tell us what altitude is worth. Neustift in the Stubaital is the valley-floor family market: moderate prices — the Stubaital runs €5,400 per square meter on our map — a glacier in the backyard, and a four-season rental pattern. And Mayrhofen in the Ziller Valley is the volume play: the most liquid resale market of the six, €4,900 per square meter, with a summer season strong enough to matter.

Why six and not sixty? Because depth beats breadth when the data collection is partly manual. We'd rather know six markets properly — registry checks, listing histories, occupancy cross-checks, local agent conversations — than sixty superficially. When we add a seventh point, it will be because we can resource it to the same standard, not because a map looks better with more dots.

A practical note on how the six points behave as a set, because it shapes how you should read every figure we publish from them. Innsbruck and Mayrhofen are the liquid pair — enough transactions each quarter that a median means something statistical. Kitzbühel and Sölden are the mid-liquidity pair — plenty of listings, fewer closings, wider error bars. Obergurgl and Neustift are the thin pair — a single estate sale can move the quarterly average by double digits, and we flag it in the text when that happens rather than letting a quiet outlier pose as a trend. When you compare figures across the six, compare the price tier first and the movement second. A 6% annual rise in Mayrhofen, built on volume, is a fact. A 9% rise in Obergurgl, built on seven closings, is a hypothesis. We treat them differently in our own models, and the weighting is written into the method notes.

How the sources feed what you see

Let's trace the pipeline, because a methodology page that doesn't connect sources to outputs is a bibliography, not a methodology. The price map — the regional €/m² figures — starts with Statistik Austria's district medians as the base layer. We then adjust for the resort premium using our platform data: listings in each measurement point, discounted by the measured 8–12% asking-to-closing spread, and cross-checked against registry pulls for recent comparable sales. If the three disagree by more than a few percent, the map shows the registry-anchored figure and we note the tension in the accompanying text.

The yield tables run on a different blend. Occupancy assumptions come from Tourismusverband regional data, discounted for single-property performance. Nightly-rate assumptions come from platform tracking of comparable rental listings — what studios, chalets, and lodges actually ask per night by season, not what revenue management software dreams of. Cost inputs come from the acquisition stack on this site and from the OeNB's rate environment for the financed scenarios. The output — the yield percentage you read — is only as good as the weakest input, which is why we publish the assumptions rather than just the result.

Refresh cadence, stated flatly: platform asking-price data, monthly, with price-reduction tracking continuous. Tourism occupancy, quarterly as published, with the big annual revision when the full-year figures land. Statistik Austria price medians, annually in October when the prior year publishes — the single biggest update in our calendar. OeNB rate and lending data, quarterly. Registry pulls, on demand, whenever a benchmark property or comparable trades. And the whole site gets a full consistency pass each October after the Statistik Austria release, because that's when the anchor moves and everything tied to it must move too.

The reference table

Everything above, compressed for the ledger. Bookmark this table; it's the page in one view.

Source	What it gives us	Refresh cadence	Cost of access
Statistik Austria	Median transaction prices by district; the annual anchor	Annual — prior year's data published in October	Free for headline series; modest fees for detailed tables
Grundbuch (land registry)	Legal record of every transaction; title, encumbrances, comparables	On demand, per property	€20–50 per property query
ImmoScout24 / Willhaben	Asking prices, supply, price reductions; typically 8–12% above closing	Monthly tracking; reductions continuous	Free to browse
Tourismusverband Tyrol	Occupancy and overnight stays by region (2023: 52.4M nights, ~80,000 commercial beds)	Quarterly; full-year revision annually	Free publications
OeNB (Austrian National Bank)	Interest rates, lending standards, foreign-buyer share estimates	Quarterly	Free publications

The limitations: what our data cannot see

Now the section most methodology pages skip, and the one we'd insist on reading if the roles were reversed. Our numbers have edges. Here they are.

Listing data is asking price, not closing price. We've said it twice already and it bears a third: the platforms show what sellers want. Our 8–12% discount is a measured average, not a law. In a hot micro-market the real discount can be 3%; on a stale luxury listing it can be 18%. When you use our figures to value a specific property, you're applying a market average to an individual case. Adjust for the property, or accept the error bars.

Transaction data has a six-month lag. Registry entries take weeks to complete after closing, and any aggregation built on them — ours included — is looking at the market as it was two quarters ago. In a stable market that's fine. In a turning market it's dangerous, because the official series will confirm a top or a bottom long after the platforms have already shown it. That's why we run both. When the fast data and the slow data disagree, believe the fast data about direction and the slow data about magnitude.

Off-market transactions aren't captured. A meaningful slice of Tyrolean prestige property — especially around Kitzbühel — trades without ever touching a public platform. Family transfers, quiet sales between acquaintances, deals assembled by lawyers before a listing exists. The Grundbuch eventually records the transfer, but often without a clean market-price signal we can interpret. Our data describes the visible market. The invisible market is smaller than rumor suggests and larger than zero, and we won't pretend to have measured it.

The foreign-buyer approval process adds delays we haven't quantified. Non-EU purchases in Tyrol require a land-transfer approval, and the process adds weeks — occasionally months — to a closing. The delay varies by district, by property type, and by the applicant's file. We know it happens; we see it in transaction timelines; we cannot give you a defensible average because the data isn't systematically published. What we can tell you is the practical consequence: model your closing at the long end of the 6–10 week range, keep your funds liquid through the wait, and don't schedule anything that depends on a specific registry date.

How a US reader should sanity-check our numbers

You shouldn't trust this page because it's transparent. You should trust it because it's checkable. Here's the checking routine we'd follow in your chair — a reader in Chicago or San Jose with a browser and an hour.

First, pull the Statistik Austria district medians yourself. The open data portal has English navigation for the headline series. Compare the Innsbruck and Kitzbühel district figures to the €/m² values on our price map. They won't match exactly — our map is resort-adjusted, theirs is district-wide — but they should be within shouting distance, and the rank order should be identical. If it isn't, write to us; that's what the corrections page is for.

Second, open ImmoScout24.at and Willhaben.at and search our six measurement points. Count the listings in your target property class. Look at the asking prices. Apply our 8–12% haircut. You're now holding our listing layer in your own hands, and if your haircut-adjusted range lands far from our figures, one of us is wrong — find out which before you wire money, not after.

Third, check the tourism story independently. The Tourismusverband publishes its occupancy and overnight figures publicly, and Statistik Austria carries accommodation statistics as well. Our 52.4-million-overnights figure for 2023 is a public number. If you're modeling yield on a Stubaital chalet, pull the Stubaital's regional occupancy and hold our assumptions against it.

Fourth, for anything transactional — a specific property, a specific comparable — spend the €20–50 on a Grundbuch query. It's the cheapest due diligence in European real estate, and it's the only step in this routine that produces a legally authoritative document. Everything else is research. The registry is the record.

Do those four things and you'll have replicated the skeleton of our process in an afternoon. That's not a threat to us — it's the design. A reader who can verify our work is a reader who catches our errors early, and both of us end up with better numbers. Which brings us to the two pages that complete this one: the detailed notes on how we calculate what we calculate, and the public log of every time we've had to fix it.

Next: Method Notes — the calculations behind the tables

Corrections — every fix we've published, on the record