Project description

BR Data is the data journalism team of the public german broadcaster Bayerischer Rundfunk. Since November 2018 we are part of the investigative editorial unit (BR Recherche / BR Data) and thus are not only working on conventional data journalism projects, but also on investigative stories that require strong programming or technical knowledge. The data-team consists of approximately eight members, which have a journalism, informatics, statistics or programming background. For the publication we usually team up with investigative journalists from our editorial unit that complement our work with classical investigative skills and bring the stories to TV and radio.

As projects for the DJA application we chose nine very different stories that show the broad approach we have in our work:

1. Increased risk: A crowd sourced investigation in one of the broadest used algorithms in Germany – the Schufa algorithm for credit scoring.
2. Gift im System (Poisoned System): A text analysis backed investigation about the admission of pesticides in the European Union.
3. Kein Spiegelbild der Gesellschaft (No Reflection of Society): A data analysis on the representativeness of the Bavarian parliament
4. Das verbaute Land (land loss country): A data analysis on land loss in Bavaria
5. Deutsche Honorarkonsuln im Offshore Business (German Honorary Consuls in the Offshore Business): A follow up to a data analysis story that showed, that many of the companies registered on the island of madeira are letterbox companies.
6. Bayer AG von Hackern ausgespaeht (German industry giant Bayer AG was hacked): By conducting a port scan on the biggest German companies we showed that the servers of the Bayer AG were infected with spyware.
7. Die kranke Rechnung (The Ill Calculation): A tap through story optimized for mobile presentation that explains why the German calculation, how many doctors are needed in each county, has nothing to do with reality.
8. Wahlplakat-Analyse (Election Posters Analysis): We crowd sourced the collection of photos of election posters to find patterns ahead of the Bavarian elections for parliament.
9. Datenleck bei VOI (Data Leak at VOI): We discovered a data leak at the Swedish E-Scooter start-up VOI with 460.00 users affected. The data leak was closed after we informed the company.

What makes this project innovative?

We are using different techniques depending on the investigation or the story we want to tell. In the last 12 months, we focused on algorithmic accountability reporting (Schufa story), on automated text analysis (pesticides story), crowdsourced data collection (Schufa and election posters stories), investigative tech reporting (Bayer hack and data leak at VOI) as well as experimental forms of data visualization (representativeness of the Bavarian parliament) and new formats of mobile storytelling with data (doctor calculation). And of course, basic skills like statistics and data visualization are always playing a big part in our projects.

What was the impact of your project? How did you measure it?

Some examples for the impact of our stories: Our investigation into the Schufa algorithm evoked an instant reaction from the German Minister of Justice, demanding more transparency from the Schufa. Our story on the Bayer hack was reported worldwide, including Reuters and the NY Times. Our story on the data leak at VOI was reported heavily in Scandinavian Countries and may have prevented that the data was stolen by criminal hackers. We are measuring our success on the one hand by counting clicks and visits to our websites, but also by monitoring the TV and radio coverage that is coming out of our reporting. We are also tracking our impact on other media as well as reactions from high ranking politicians. Success for us also means that we always want to develop our individual skills and as a team, e.g. by using machine learning algorithms and building deep knowledge of the technical infrastructure in our (data driven) world.

Source and methodology

The data we use can be either public, crowdsourced or exclusive in forms of data we requested from third parties or information provided by whistle blowers.

Technologies Used

We are constantly using the toolbox of different machine learning algorithms, that are helping us with tasks such as OCR, text analysis and image recognition. For the data analysis we are working with R, Python, Javascript or whatever gets the job done. For the crowsourced election posters collection we built bots for What’s App and Telegram that helped people uploading their pictures. For the bigger projects we are building customized websites using mainly HTML, CSS and Javascript.

Project members

Robert Schoeffel, Uli Koeppen, Oliver Schnuck, Maximilian Zierer, Elisa Harlan, Steffen Kuehne, Maximilian Richt, Niels Ringler, Sabine Wieluch, Hakan Tanriverdi


Additional links


Click Follow to keep up with the evolution of this project:
you will receive a notification anytime the project leader updates the project page.