The Self-Driving Company: A Summary of the Book

In 2021, Springer Gabler published my book "Das selbstfahrende Unternehmen. Ein Denkmodell für Organisationen der Zukunft" (The Self-Driving Company: A Thinking Model for Organizations of the Future). It describes how companies will hand over the majority of their routines and decisions to software by 2035, and what role remains for people. With this post we begin a series in which we explore the book's core theses one by one; we start with a summary of the thinking model.

Whatever is technically possible and useful will eventually be implemented

The history of humankind can also be read as the history of its tools. From the wheel to the steam engine, the automobile, and the personal computer, one pattern holds: whatever is technically possible and proves useful in practice will sooner or later be put to use. The peoples, and later the companies, that adopted a new technology first gained considerable advantages from it.

Today we stand at another such threshold. The global networking of data that has emerged over the past 25 years meets artificial intelligence and computing power that continues to grow rapidly. This fundamentally shifts the task of computer systems. Until now, that task has mainly been to capture, store, and prepare data. In the future, software algorithms will make decisions based on this data and execute them independently, faster and more precisely than people could.

An estimate by the International Data Corporation that I cite in the book (IDC, via Statista 2020) shows the size of the mountain of data we are dealing with: the global volume of data is expected to grow from 33 zettabytes in 2018 to 175 zettabytes in 2025. The problem for companies is not storing these volumes of data but actively using them. Without self-learning algorithms, this complexity can no longer be mastered; with them, we remain the decision-makers.

The role of people in organizations changes as a result. More and more, their task will be to set strategy, develop creative solutions, and take care of culture and communication. This communication no longer takes place only between people, but across a variety of interfaces with software systems and mechatronic robots. Routine tasks that give no one a sense of achievement disappear.

Why "self-driving": the analogy to the vehicle

The term is deliberately striking. The analogy to the self-driving vehicle creates a clear mental image: just as the individual components of a car work together intelligently, highly digitized subfunctions in the self-driving company communicate with each other in real time and steer the overall system with a precision that cannot be achieved manually. Semantically, "self-adapting" would be more accurate; in the book I keep the more memorable term.

The analogy has a second layer. We are close to leaving the driving to a car. Only when it comes to the company do we still believe it is better to do everything by hand. But the logical consequence of the pattern described at the outset is this: everything that can be self-driving will sooner or later be self-driving, controlled by learning, reliable systems that are deployed exactly where they are superior to people.

According to the book's definition, a company is self-driving when roughly 80 percent of its business functions and decisions are automated and controlled by sufficiently intelligent algorithms. The remaining 20 percent are special cases, personal advice, complex work on site at the customer's premises, and all tasks that require empathy and creativity.

The four dimensions of the company of 2035

The vision reaches further than the current change project and the five-year strategy. It comprises four dimensions, each of which contributes decisively to success and which I will return to again and again in the coming posts of this series.

• Autonomy. The starting point from which everything else follows: the company of 2035 acts largely autonomously. Many operational and tactical decisions are no longer made by people but by software. The biggest obstacle on the way there is not the technology but the uncertainty that arises from a lack of knowledge. Anyone who already uses translation services or voice input has long been working with artificial intelligence without calling it that.
• Sustainability. The book's thesis is that an autonomous company in 2035 must operate sustainably without exception, and that it can do so better than a company steered exclusively by people. People decide in the short term and cannot see all the long-term consequences. An organism that is fully networked internally and externally continuously computes all scenarios on the basis of all data and ensures sustainability in the sense of ecological and economic goals.
• Resilience. The world has become less predictable since the end of the Cold War; many companies can hardly plan for the medium term anymore. Real-time data, projections, and simulations allow forward-looking management: risks are recognized early, and the company keeps functioning even in an economic crisis or a pandemic.
• Humanity. The company of 2035 puts people at the center in every respect. The worry that computers might come to rule over people is unfounded: software has no motives and no will of its own; it does what it was programmed to do, although it keeps learning at a high pace. People set the direction. Because operational decisions are made by the system, they have their heads free for humanitarian concerns. And because the system is fully transparent, unfavorable practices can no longer be concealed.

Up to 2035, people will continue to decide where the journey goes; the company will handle these decisions in a self-driving way.

The road there: the levels of autonomy

As with vehicles, a company's technical maturity can be described in levels of autonomy. Individual teams, departments, and divisions develop at different speeds; we only speak of a self-driving company once the majority of all areas have fully transformed.

An example from the book makes the levels tangible: the incoming invoice. If it arrives by mail, there is no digital basis to build on. If it arrives as a PDF by email, the company is one step further; an algorithm can learn to recognize and assign the invoice, verify the service delivered, and release the payment. If the supplier changes its logo, a person checks once, and from then on the system knows the new version. In the fully self-driving company, the invoice disappears as a document: platforms exchange the data, together with the associated contracts, directly between the partners' systems, just as today no bills change hands in the flow of money; data is simply rewritten.

This example shows why many companies today are further from the goal than they assume. The scanned invoice is digital, but it is only an image. It frees the archive from paper, not the process from manual work. Most organizations are at an early position between the analog and the digital world, even though they have been working on digitalization since the mid-1990s. Whoever has 80 percent of their data available in machine-readable and processable form has cleared the first real hurdle and is already leaving the competition behind.

At the highest level, a further effect comes into play. While the human ability to absorb and process information is limited, the algorithms of the self-driving company process almost unlimited volumes of data, from the company itself and from its relevant environment. Decisions thereby become not only faster but better, and thanks to full transparency they remain traceable at all times.

The exponential function: why getting started now matters

The transition to the self-driving company has begun and will accelerate considerably within this decade. If we perceive digitalization as rapid today, that is because of a planning horizon that ends with the conversion of analog data into digital data. That is only the first step on a much longer road.

In the book I describe this development as an exponential function. The physicist Al Bartlett called the inability to understand the exponential function the greatest shortcoming of the human race (Bartlett 2006). Applied to companies, this means that for a long time hardly any progress is visible because a foundation is being laid. As soon as more and more parts of the system become independent, learn, communicate with each other, and process growing volumes of data in real time, the curve rises steeply. My assessment in the book: it will take until about 2030 for these changes to become visible. After that, self-driving companies will not only overtake the half-digitized ones but keep accelerating as they do.

Three practical consequences follow:

1. Think big, start concretely. It is not enough to set up a small digital unit as a lighthouse project and hope it outshines the idle rest. What is required is the restructuring of the entire company, not of the business model: what works today will also work self-driving. The starting point is wherever a digital basis already exists, for example a well-maintained CRM system in sales.
2. Start early, especially as a large company. Anyone who wants to replace their old ERP system only once the curve is already rising must allow four to five years. The larger the company, the earlier it has to prepare. Smaller companies will have an easier time because they can source cloud-based products with intelligent functions from the market.
3. Decide deliberately. Not every business model needs the self-driving company. The workshop that makes tailor-made lederhosen by hand deliberately decides against global scaling. Large companies do not have that choice; those who do not go along disappear from the market, as Nokia and Kodak have demonstrated.

Seven theses on the company of 2035

The book condenses its findings into seven theses. They form the common thread of this series:

1. People will continue to work in and for companies; they perform the empathetic and creative tasks.
2. Intelligent software algorithms are better suited to repetitive tasks than people.
3. There will be no more processes: processes belong to the 19th and 20th centuries, algorithms are the 21st century.
4. Middle management will be replaced by software; leadership tasks remain with people.
5. Software and IT are the basis of the self-driving company. Central IT departments in their current form will dissolve; development and operations will move into the responsibility of the business units.
6. Jobs that disappear will be compensated for by demographic change and by market growth resulting from productivity gains.
7. Companies will still serve people in 2035.

The methodology of our management consulting at ReqPOOL is derived from this vision. Based on a precise analysis of the circumstances in a company, we show which routines and processes can be controlled digitally and in what order. This relieves employees and at the same time continuously increases value creation.

The next step

The book is aimed at decision-makers, entrepreneurs, and politicians who want to implement the aspects of the thinking model relevant to them over the coming ten to fifteen years. You will find an overview of the book and its topics on the page The Self-Driving Company. If you would like to know which level of autonomy your organization is at today and where the first sensible starting point lies, set aside an hour: Book an expert meeting.