Data Studies German Trade Press Cites and Links To

A study earns coverage when it answers a question a journalist would otherwise have to research themselves. That is the test applied before we agree to build one, and it disqualifies most survey ideas that arrive: a panel poll of 1,000 consumers about their intentions gives an editor nothing they can check and dates within a month. Where the data comes from: Administrative and registry sources, because a reporter can verify them without asking you. For automotive that means the Kraftfahrt-Bundesamt monthly new-registration series and the Bundesnetzagentur charging-point register. For travel, Destatis Genesis-Online accommodation and arrivals tables, Eurostat and published airport traffic statistics. For B2B software, public procurement notices and job-posting text where hiring is the leading indicator of a budget shift. Where we do field a survey, it is because no register holds the answer, and the sample, field dates and exact question wording go into the appendix where an editor can see them. Method: The question and the analysis plan are written before the data is pulled, so the finding cannot quietly become whatever flatters the client. If the result is dull or contradicts your positioning, you hear it and decide whether to publish. We have shelved two studies this way and would do it again rather than defend a number to an editor who can download the same file. What gets published: A page on your own domain with the finding in the first 40 words, the method beneath it, the full dataset as CSV, and charts as SVG plus the underlying table so a graphics desk can rebuild them in house style. A copy of the dataset is deposited with a DOI so the study is still citable in three years. No gate and no email form: a study behind a form is a lead magnet, and lead magnets are linked by nobody. German and English versions get separate headlines, because the finding that leads in Handelsblatt is rarely the one that leads in a UK trade title. Taking it to desks: An embargo pack goes to a short list of named journalists five working days ahead, carrying the finding, the one chart that matters, the method note and a slot with whoever ran the analysis. Data desks want the file. Trade desks want the sector cut. Both get a version written for them rather than the same release twice. How links are counted: Editorially placed root domains only, verified in Ahrefs and by hand. Syndicated copies of the same article, wire repost sites and aggregators are listed separately and excluded from the count, because they inflate the number and nobody serious is fooled by it. The report gives you domain, anchor, followed status and the coverage that carried no link at all, which is often the coverage that mattered. Not included: Link building involving payment, exchange or guest-post placement. Shops that promise a link count are buying part of it, and a client of ours collecting a manual action would cost us more than the account is worth. We also do not build studies whose finding is that a named competitor is worse. It reads as marketing and desks treat it accordingly. Not for you if: You need the study to conclude something specific. You need it live in three weeks: registry acquisition, cleaning, review and a five-day embargo do not compress below six. Or you cannot host it on a domain you control, in which case the links accrue to somebody else's asset and you have funded their SEO.