“Think Faster, Implement Faster”

Mr. Mayer-Schönberger, is data the crude oil of the 21st century?

No. That is nonsense. Many large companies have seen this cover of The Economist and believe the most important thing is to equip all their machines with sensors to record as much data as possible. The problem is that the data is then not used. A study has shown that 85 percent of the data collected by companies in Europe is never used once.

So it’s lying around on some servers?

That is correct. In contrast to the physical resource oil, the added value of information-based goods such as data is only created through their repeated use. For example, if you type a search query into Google, you get a search result. This is the first use. But Google also uses search queries to train the Google spellchecker. As a result, Google has developed the best spelling checker in the world. This is how data can be reused, and Google does it all the time. Here’s another example: While Google's autonomous vehicles are driving around, they collect around one billion data points per car, per second — including data about WiFi networks that they drive past. This creates a large database of WiFi networks worldwide. If you use an Android phone, it not only uses the GPS signal, but also detects open WiFi networks in the area and triangulates your exact position from these and the database. Here again, the data is used multiple times.

That’s all very well for Google, but how can comparatively small European companies keep up?

They can. The most important thing is that they change their mindset. They need to understand that they need data to make better decisions. They have to collect data and use it. That means they have to encourage people, especially within the company, to use the data in different ways for different topics. Here’s an example: Lufthansa aircraft fly on autopilot most of the time. To do this, weather data is continuously collected and passed on to the autopilot. A few years ago, Lufthansa began reporting all of this meteorological data to the German Weather Service. This improved the German weather forecast by seven percent in one go. Lufthansa also benefited from sharing the data, because with more accurate weather forecasts, they can plan flights better. This is how you gain value from data.

This means that using data is successful if you share it.

That is correct. In companies, it is important to know what data is available, to make it accessible and to share it - across all silos. In addition, even very innovative companies do not generate all their good ideas within their organization. Sometimes they are discovered outside, at a start-up. Therefore one should consider how to enable data usage beyond one’s own organization. Companies like PALFINGER have an unbelievable amount of data that they could make available. Since this is usually sensor data, it does not relate to particular persons, so there are no data protection issues.

How do you handle the worry that this data could flow to the competition?

Quite simply with innovation. By making sure that everyone else with their reverse engineering is lagging behind the new products and services. Chinese companies are excellent at this, but Intel - for example - largely doesn't care because Intel makes 90 percent of its profit from chips that came onto the market in the last six months. That means by the time the others have implemented that, the real innovator has already moved on again. This also means that companies like PALFINGER must think faster and turn ideas into products more quickly.

So this data is the oil, because it drives innovation. Does it also drive value creation, cooperation and collaboration?

That’s right. Basically, data only has the function of helping us make better decisions. It's about asking the right questions and not just collecting information to clarify existing questions. Big data, machine learning and artificial intelligence help us do this because these tools do nothing but draw our attention to specific patterns in the data from large amounts of data. This enables us to generate better questions, as demonstrated by Canadian researcher Carolyn McGregor. She focuses on premature babies, a disproportionate number of whom died from infectious diseases that were identified too late. The pattern she found in the data was surprising: If the vital signs become stable from one second to the next, there is a high risk 24 hours later. If it were not for her work, we would never have asked the question: “what does it mean if the vital signs suddenly become stable?” These are precisely the questions that we don't generate often enough on our own. Big data and artificial intelligence help us to become aware of these surprising patterns and to generate better questions from them.

You mentioned mindset before, the internal sharing of data, you also mentioned data protection — how much of a hindrance is it?

It is not as legally restrictive as we assume it is. According to the General Data Protection Regulation (GDPR), if you have collected personal data, you may only use it for a specific purpose. But there are exceptions to this. For example, data that has been legally collected for a specific purpose may also be used for statistical evaluation for other purposes. The GDPR makes it easy for large companies that collect a lot of data during normal operations to use it for other purposes. This not only helps the Googles of this world, but also companies like PALFINGER. In other words, everyone who has a lot of customer relationships because they have lots of devices with lots of sensors that collect a lot of data. The GDPR makes it difficult for very small start-ups that do not have large amounts of data.

This means that it could be a business model for large companies to provide start-ups with anonymized data.

That is correct. You can join forces with start-ups in a data pool and offer data for a monthly fee. Or you make the raw data you've collected available free of charge, and work out together how to gain valuable insights from it. This value-adding process is worth paying for. Information-based goods work differently and require different mechanisms to physical goods. The latter requires ownership, purchase, and all the classic transactions that we have developed for physical goods. However, it no longer works that way with information goods. Just think of ownership: If you read a book, which I then read  too, I'm not taking anything away from you. There is no point in having an absolute exclusive right to read the book.

This means that through the purpose-specific and shared use of data, we are opening up a whole new world and myriad opportunities.

And that can be surprisingly easy, as long as the focus is on what comes next. There is a Swedish start-up that builds trucks that drive autonomously on the highway. This is technologically feasible, and they drive in a convoy without any problems. As soon as they leave the highway, a driver who is located at the company’s headquarters and is connected to the truck via 5G takes over and steers the last few kilometers to the warehouse or to the customer. As a result, the truck drives completely autonomously 85 percent of the time. This also means that the logistics company can get by with significantly fewer truck drivers. Because the driver is only needed for the first and last five to fifteen minutes of the journey. This means that one truck driver can operate several trucks. This is technology that works today, so there is no need for complete autonomy, it is just about making the most of the opportunities that the data opens up.

Thank you very much for the interview.