Expert Hearing on Data Strategy at the Federal Chancellery of Germany (2020)

Mr. Federal Minister Braun, Ms. State Minister Bär, I am very pleased to have been invited and to be here today. I am co-founder and CEO of Datarade. We are a global matchmaker for commercial data and also a member of the German AI Association. I would like to answer the two questions separately.

Data Usage in Germany

Where do we stand today, from my perspective or from our perspective, regarding data usage in Germany? Frankly, studies on data usage vary widely and are very contradictory. A 2017 study by TERRA Data states that Germany and German companies are international leaders in the field of Big Data and Analytics. The German Economic Institute (IW) calls Big Data uncharted territory for companies. According to their 2019 survey of companies, only eight percent of German companies conduct Big Data analyses. I believe I have also heard other study results on this topic.

From our perspective as a globally operating data company, there is a data usage gap in Germany. That's how I would describe it. On the one hand, we see extremely ambitious startup projects and software companies. They are building innovative and data-driven business models and technologies. Examples include Celonis, Datameer, Exasol, Zeotap, Scoutbee, SAP, and the healthcare cloud, to name just a few. On the other hand, there are traditional medium-sized companies that are still in the process of digitalization. They first need to create structures for internal data acquisition and use.

Challenges

From our perspective, this leads to the following challenges: Fewer German companies are willing to share their internal data and use external data, for example. This impacts the availability of data. Our internal research shows that for every one data provider in Germany, there are ten data providers in the US alone.

Second challenge: Access to data is more difficult. Commercial and public data offerings are sometimes very hard to find and hardly standardized. This makes them complex to integrate with internal data repositories. The US Open Data Portal contains 260,000 different datasets, I checked yesterday. The EU's Open Data Portal only lists 14,000, roughly 20 times fewer. Does this mean we have less data in the EU? I don't think so. And thirdly: Yes, I believe a challenge is the shortage of skilled workers. Even when internal and external data are available, and infrastructure is in place, many German companies lack the expertise and skills in the areas of Big Data and AI. They need specialists like data scientists or machine learning engineers who can extract information, knowledge, and intelligence from data.

Measures

Let's move on to the measures. And here I would like to be very specific. We want to become leaders in data use and AI in Germany. To achieve this, the free flow of non-personal data must become a priority and the fifth fundamental freedom of the EU single market. The legal, technical, and financial implementation requires close cooperation between government, industry, research institutions, associations, and civil society. Specific recommendations for action: We must create more data transparency with initiatives that promote and demand the discoverability of and access to publicly and commercially available data.

Let's start with the commercial register. Company data is extremely valuable for KYC processes, fraud prevention, and B2B sales planning. Data can be accessed individually via handelsregister.de, though some of it is subject to a fee. However, it is neither machine-integrated, for example via an API, nor freely reusable under licensing agreements. From our experience, I can tell you that such company data is among the most valuable data that companies in Germany want and need to use.

Secondly, we must jointly create data standards that enable the free exchange of data. They must also facilitate integration with data sources. Let's take company data as an example again. There is no open and stable identifier for companies. This complicates data maintenance and the exchange of company data. Here, the private sector relies on and is dependent on proprietary standards, such as the Dun & Bradstreet D-U-N-S Number.

Thirdly, we must jointly promote more positive attention, investment, and inspiration in the areas of data, AI research, data-driven innovation projects, novel data technologies, and use cases. Increased collaboration between technology startups and established companies can also help close the data usage gap.