Transparency

Methodology & data sources

Which data underlies every rating, how it is aggregated, and where the analysis has limits — as completely as possible.

Data sources (all Tier-1, CC-BY-compatible)

GISA — Gewerbeinformationssystem Austria

Source: data.gv.at · Dataset "OgdAufrechteGewerbeberechtigung" · Licence: CC-BY 4.0

The official register of all active trade licences in Austria, maintained by the Federal Ministry for Labour and Economy. Updated monthly, broken down by trade + state + postcode. The primary data source for density, trend and churn.

→ Dataset on data.gv.at

Statistik Austria — population figures

Source: statistik.at · STATcube "STAT03" · Licence: CC-BY 4.0

State population figures as of the reference date 1 January of the current year. The basis for the density calculation (trade licences per 10,000 residents). The figure is updated once a year — our densities are correspondingly at annual, not monthly, granularity.

City of Vienna — district population

Source: data.wien.gv.at · Dataset "Population since 1869 - Vienna districts" · Licence: CC-BY 4.0

For the 23 Vienna districts we use the City of Vienna's own time series (Statistik Austria has no district-level publication for the same reference date). Current reference date: 1 January of the current year.

OpenStreetMap — visible competitors

Source: Geofabrik Austria extract · Licence: ODbL 1.0

The "OSM-visible competitors" map filters POIs by matching tags per trade (e.g. shop=hairdresser for hairdressers). Important: OSM is crowd-sourced and incomplete — we show it only for 14 of the 20 trades where coverage is high enough for meaningful cross-comparisons.

Not shown (too patchy at district level): IT services, business consulting, advertising agencies, taxi, master builders, life and social coaching.

No commercial data, no WKO company-directory scraping, no Google Places API — the entire analysis is reconstructed from open, licence-compliant sources.

How we calculate

Density per 10,000 residents

density = (number_of_licences / residents) × 10,000

A simple normalisation — it makes regions of different sizes comparable. For Vienna districts we use the City of Vienna's district population, for states the Statistik Austria figure.

Saturation rating (5 classes)

We cluster the densities of all regions per trade using quantile cuts:

low: below the 25th percentile
medium: 25th–50th percentile
high: 50th–75th percentile
very high: above the 75th percentile

These cuts are trade-relative — a "high" value for hairdressers is not the same as a "high" value for photographers. We always also show the absolute figures so you can compare for yourself.

13-month trend

The trend is the net change (additions minus closures) over the last 13 monthly reports. We do not show the absolute delta percentage across the entire period, because GISA changed its publication methodology in May 2025 (from annual to monthly reporting) — that break would distort any trend spanning before/after May. Instead: the time series itself, with an explicit methodology-break indicator.

Churn indicator

churn = (additions_12m + departures_12m) / average_stock

Measures how much "moves" relative to the total stock. High churn at a stable total means: there are often new entrants AND existing ones often quit. That can be an easier market than a crusty market with low churn.

Limitations (honestly)

Density ≠ market demand. High density means many providers, not automatically strong market potential. Customer purchasing power, tourism effects and border-location demand are not captured.
GISA ≠ real activity. An active trade licence is not the same as an actively trading business — some licences are reported dormant or run as a sideline.
District accuracy varies. Some trades have only a handful of businesses per district. Statistical fluctuations can then trigger larger percentage shifts than a real market shift.
OSM coverage is trade-dependent. See above — for 6 of 20 trades we show no OSM layer, because the crowd tags are too patchy.
Niche positioning is not assessed. A high overall saturation does not rule out specialised micro-niches (e.g. "hairdresser for Afro hair only"). The analysis gives the macro baseline; you have to sharpen your own positioning yourself.

Frequent questions about the methodology

Why do you use GISA and not the WKO company directory?

GISA (Gewerbeinformationssystem Austria) is the official register of all active trade licences in Austria, maintained by the Federal Ministry for Labour and Economy. It is freely licensed under CC-BY 4.0 (data.gv.at). The WKO company directory (Firmen-A-Z) has a more restrictive licence and additionally contains voluntary WKO memberships — not every trade is a compulsory WKO member. GISA gives the legally more precise, more complete figure.

What exactly does "density per 10,000 residents" mean?

We divide the number of active trade licences in a region by that region’s population and multiply by 10,000. The result is comparable across states of different sizes. Population figures come from Statistik Austria (states) and the City of Vienna (districts, "Population since 1869 - Vienna districts", CC-BY 4.0).

How is the 13-month trend derived?

GISA publishes a new snapshot every month. We store each snapshot locally and aggregate the net trade licences per niche + region over the last 13 months. A jump in May 2025 is visible because GISA switched there from an annual-report logic to monthly reports — we flag this as a methodological break explicitly on all trajectory charts.

Why does the churn indicator sometimes read high even though the total is stable?

The saturation figure alone only tells you how "full" a market is. The churn indicator measures the movement — how many trade licences are added in a month and how many are terminated at the same time. A crusty, stable market has low churn (hard to break into). A dynamic market with high churn can be easier to crack, even at high saturation — because seats regularly become available.

Is the OSM competitor map complete?

No, OpenStreetMap (ODbL licence) is a crowd-sourced database. Some trades are systematically underrepresented (IT services, business consulting, advertising agencies, taxi dispatch, master builders, life and social coaching — all trades that often have no physical walk-in address). For these 6 trades we don't show OSM data and mark the corresponding district pages with noindex, so as not to communicate misleading gaps.

How often is the data updated?

GISA publishes monthly reports in the first week of the following month. Our pipeline runs automatically the day after release: fetch → transform → SQL-load → Astro rebuild → IndexNow ping. New monthly snapshots typically appear on the website 3-5 working days after month-end.

Why aren't all 2,000 GISA trade codes on the platform?

We started with 20 active trades that together cover the majority of new registrations with real competition — where a saturation analysis is genuinely decision-relevant. For heavily regulated trades (pharmacists, notaries, chimney sweeps with territorial protection) a density analysis doesn't always make sense; those are left out for now. You can request trades any time via the contact form.

Can an analysis be wrong?

Yes. The analysis is based on publicly available data and statistical aggregation. What it does NOT capture: local customer loyalty, niche positioning, price-level differences, seasonal fluctuations below monthly granularity, undeclared work, cross-border demand. The analysis is meant as a first, data-driven sanity check — not as the sole basis for a decision to start a business.

Who checks the content?

Textual content is created under the motto "automated analysis, human editorial oversight": data interpretation and FAQ answers are based on template logic but pass a manual review before going live. Methodological changes (a new trade, a changed churn definition) are documented traceably in the /en/glossar/ and here on /en/methodik/ — with date and reason.

Methodology change history

2026-04-19: Initial publication. 20 active trades, 9 states, 23 Vienna districts, monthly GISA snapshots, 13-month trajectory.
planned: post-launch: Seasonal churn decomposition (removing month-fixed effects so that summer-vs-winter patterns are not misinterpreted as "market movement").