Data Study Built on German Public Registers

A study earns links when it contains a number a journalist cannot get anywhere else. That is the entire test, and about one concept in three fails it at the data pull, which happens in week two and which we tell you then, before a designer has been booked or a landing page built. Killing a study in week two costs you the scoping fee. Publishing one that fails the test costs you the whole budget and a page nobody cites. Where the numbers come from: The GENESIS-Online API at Destatis, Eurostat, the open series from the Deutscher Wetterdienst climate data centre, Overpass queries against OpenStreetMap, municipal Mietspiegel documents that exist only as PDFs and have to be parsed, the Europace house price index, Bundesbank residential property series, and licensed listing or short-let data where the question needs it. The work that produces the finding is usually a join nobody has bothered to make: a federal series against a municipal one, a weather series against an occupancy series, a planning-permission series against a completion series two years later. Original survey work happens only when no register answers the question, through a panel provider with the sample size and field dates printed on the page. The kill point: At the end of the data pull we state the finding in one sentence and test it on a specialist editor we know. If a person who covers the sector for a living is not surprised, the study is dead and we stop. We would rather return a scoping report saying the data does not support a story than produce a well-designed page that says rents went up. What ships: A landing page at a stable URL you commit not to move, with the finding in the first sentence and the method underneath it. The dataset as CSV and XLSX under CC BY 4.0 with a citation line written for the journalist to paste, which is the single change that turns a mention into a link more often than any outreach tactic. A methodology note in German and English covering sources, field dates, sample, and what the data cannot show. Charts as SVG with PNG fallbacks, sized for a newsroom CMS. An embed snippet. A press kit with the numbers broken out per federal state, because a regional desk will run the Bavaria figure and not the national one. The pitch: A named list of twenty-five to forty editors built for this specific finding, pitched individually in German, with the state-level cut attached to the regional titles and the national cut to the trade press. No mass send, no embargo games, no exclusivity offered to two outlets at once. What we will not publish: A study whose conclusion happens to be that your product is the answer. A three-hundred-respondent panel survey presented as research. An asset that is a designed infographic with no underlying data file. A finding that requires the reader to accept a proprietary index we invented in order to be interesting. Timing: Ten to fourteen weeks from scoping to first pitch, of which two to three are the data work and one is waiting on a Landesamt to answer a data request. If you need links inside the quarter, this is the wrong instrument and reclamation or bylined contribution will serve you better. Who should not buy this: Anyone who needs to approve the conclusion before the data is pulled. Anyone whose legal team cannot sign off on publishing a number that is unflattering to the sector they sell into. And any team without one person who can own a page that will be updated annually, because the studies that keep earning links are the ones that come back with the next year's figure.