2.2.2 Digitalization and System‐Cybernetic Complexity Management Have a Common Basis
Digitalization and the management of complex systems have the same hour and place of birth. The power of holistic, system‐cybernetic management far surpasses the mechanistic linear management—many today call it “completely new”. Its origin lies exactly in the time and place that also saw the development of today’s computer technology. Partly the same pioneers were responsible who also build the fundament for modern computer technology and simultaneously for the proper management of societal organizations: They realized early on that the same principles apply to both areas.
One of them is mathematician Norbert Wiener, founder of modern cybernetics, with his book “Cybernetics: Control and Communication in the Animal and the Machine” (1948). Also the British neurophysiologist Ross W. Ashby with his revolutionary book “Design for a Brain. The origin of adaptive behavior” (1952). Add John von Neuman, the inventor of the modern computer, Claude Shannon’s information theory and Heinz von Foerster’s Causal Circularity.
With their early theories on information and communication, algorithms and heuristics and on the design and navigation of complex systems, they have created the prerequisites for today’s real “cyberspace” as well as for the cybernetics of the management of complex systems. It was still too early for this new system‐cybernetic management though. For decades, the industrialized society’s mechanistic notion of management would continue to dominate, and it would be taught at thousands of universities and business schools—until today.
2.2.3 Organizations as Living Organism
The current challenge of the Great Transformation21 forces the renunciation of mechanistic management. The prevalent notion of the last decades that the company is a machine which can be steered with the linear principles of cause and effect blocked the necessary progress.
The more helpful notion is the organization as a living organism in its evolutionary environment. Management then is responsible for enabling organizations to self‐organize and self‐regulate wherever numerous and ever more people work together to reach common goals. Contrary to wide‐spread fears this is exactly what allows people the freedom for the first time to unfold their intelligence and creativity in the digital world in a new and better way.
The new goal is the adaptive viability of a flexible organization that far surpasses the notion of sustainability. What needs to be done has been identified—and system‐cybernetic management provides the How and Whereby. Hence, digital interconnectivity and system‐cybernetic management of complexity will become the very societal functions that enable people to effectively exploit the new possibilities and opportunities.
© Springer-Verlag GmbH Germany 2018
Claudia Linnhoff-Popien, Ralf Schneider and Michael Zaddach (eds.)Digital Marketplaces Unleashedhttps://doi.org/10.1007/978-3-662-49275-8_3
3. Preface: New Computing in Digital Marketplaces Unleashed
Florian Leibert1
(1)Mesosphere Inc., San Francisco, USA
Florian Leibert
Email: [email protected]
We are at an inflection point where consumer technologies and innovation have forced their way into Global 2000 enterprises. People are asking themselves why enterprise software should be so much worse than someone’s experience on an ecommerce travel site or searching for something on Google. And disruptive technologies such as new digital consumer devices and services have disrupted the datacenters at the heart of every large company in the world.
These advances in enterprise IT require new methods of computing, and a new breed of entrepreneur able to capitalize them. The unleashed digital market places are an exciting time that arguably began with the rise of the Worldwide Web just more than a decade ago, but really accelerated with the rise of social networks and mobile devices in the more‐recent past.
Think about the shifts that have occurred since the 1990s, and have really picked up their pace in the past decade. One of the biggest is the proliferation of computing and the ability (or requirement) to reach consumers where they spend so much of their time—on their smartphones, tablets and laptops, or via connected devices that interface with those computers. A large enterprise application in 1995 might have involved thousands of client desktops connected to a single big‐iron application server or database server. Now tens of thousands to millions of laptops, tablets, smartphones and other devices have to access a single application at any given time.
What happened was that giant web‐scale companies such as Google, Facebook, Yahoo! and Twitter—all successful consumer web companies—grew so big, so fast that they had to invent distributed computing systems to prevent themselves from falling over or drowning in complexity. Because they were solving new classes of problems that legacy systems could not manage, they typically invented new open source frameworks themselves or they jumped on nascent projects coming out of top computer science universities.
Without distributed computing on the backend systems, there is no way any application could handle that much traffic—much less store, process and serve, in any reasonable amount of time, all the user and log data that millions of users can generate. This is why from Apache Spark to Apache Mesos to Apache Cassandra, almost every new technology of any real utility (and popularity) over the past decade is a distributed system.
Distributed computing is so commonplace in some areas that a single Google search or Facebook update on your smartphone touches dozens of such systems at very large scale—each spanning the equivalent of thousands to hundreds of thousands of nodes—in order to process data, rank results, serve ads, tap the social graph or knowledge graph, and much more. In fact, companies such as Google, Facebook, Microsoft and Amazon are so adept at managing and automating entire datacenters full of computers that they don’t even think about units as small as server racks anymore. They buy machines by the shipping container and compete for bragging rights over who has the most, biggest and best datacenters.
3.1 From the Fringes to the Mainstream
Today, the types large‐scale distributed systems, real‐time data processing and service‐oriented architectures these web companies built are business imperatives in corporate digital transformations. Executives and boards around the world are demanding their companies deliver products that can compete with the likes of Google, or at least can mimic Google‐like efficiencies and innovation within their own IT departments. This leaves CIOs looking for new IT allies that truly understand the technologies their enterprises will need in order to deliver on these mandates.
While few organizations will ever reach that level of scale, the techniques these companies have developed to manage their datacenter environments are very useful for small startups and large traditional enterprises, as well. That mundane startup news website or innovative mobile game you love, for example, is probably running on some combination of, say, MongoDB, Elasticsearch, Spark and Amazon S3. Traditional enterprises are getting increasingly hip to distributed systems, too, targeting data analytics with systems such as Hadoop and Spark, or faster application performance with a distributed database.
More and more enterprises are exploring Docker containers and want to deploy microservice applications—that is, packaging each component of an application (sometimes dozens of them) as connected, but loosely coupled and separately‐managed services. Microservices are frequently more about improving developer agility and simplifying IT operations than they are about scalability, but containers themselves have proven remarkably useful for technology pioneers such as Google, Facebook and Twitter. But even established enterprises like Verizon, Disney, and GE are embracing big data and re‐architecting their computing around microservices and a containerized infrastructure.
Example
They have seen the business‐changing, if not world‐c
hanging, applications that are possible when we monitor data from mobile devices, park visitors and jet engines—along with just about everything else in our physical world. They have seen the multi‐billion‐dollar potential of building first‐of‐a‐kind and user‐friendly consumer experiences. And they know the technologies that can help them deliver on these goals.
3.2 Distributed Computing is Hard—Good News for Entrepreneurs
For these reasons, I believe the movement toward this new distributed computing stack is inevitable. But it is not without its challenges. This creates business opportunities for entrepreneurs.
Most new distributed systems technologies require deep technical expertise to use in their raw forms. Many are open source software, oftentimes developed at large web companies and now managed by the Apache Software Foundation. They’re appealing because they’re free and proven to work at large scale. But you need teams of very smart engineers—who are hard to find and expensive to hire—to make it all work.
I experienced this firsthand at Twitter and Airbnb, where technologies such as Apache Mesos and Apache Hadoop helped us dramatically improve our ability to manage an ever‐increasing number of servers (and the applications running on them) and process big data. However, there is no way we could have achieved those end results without some of the smartest computer scientists in Silicon Valley adapting those open source technologies to fit our specific production environments.
Example
Truly capitalizing on the distributed computing movement will require new ideas about how applications are built and new technologies to make them possible. Mainstream businesses will require software that lets developers unleash their ideas by tapping into a platform composed of thousands of servers and complex data‐processing systems—without having to understand everything that’s happening below the surface. They’ll need software that automates the day‐to‐day experience of operations staff, so they can kiss goodbye the late‐night outage emergencies and confusing catalogs of which applications are running on which servers.
3.3 A New Class of Startups for a New World of Computing
This means there is a golden opportunity for a new class of enterprise IT startups that is born out this datacenter‐scale world. These are companies that understand the technologies at play—their founders have often worked with or created them at large consumer web companies—and understand how to turn them into products that large enterprises will actually buy. From Day One, this new class of startups has operated with the knowledge that their success hinges on how well they can balance cutting‐edge technology, real‐world requirements, and an ecosystem of partners working on other pieces of the new distributed stack.
Open source software is a common, although not entirely necessary, component of the business model for many of these companies. The depth of a company’s open source commitment is often tied to the technologies upon which its product is built, and to the level of software stack at which it plays. Popular infrastructure‐level software, for example, is often open source today (for reasons that have more to do with innovation and lock‐in than desire for free software), whereas application‐level software is less frequently open source.
For startups, like Mesosphere, that sell products based on open source software, a lot of value lies in the community engagement and crowdsourced innovation that open source technologies facilitate. Smart enterprise IT startups understand that they don’t know everything and it’s impossible for them to stay current on all the latest technologies. There’s a lot to be learned (and gained) from internalizing feedback from a diverse set of users, and from letting subject‐matter experts bolster platform capabilities in areas in which your company is not an expert. This is a departure from many legacy IT practices, but also from the practices of web giants, who often build tools or distributions designed specifically for their own environments before releasing them to the world under an open source license.
Turning remarkably scalable, remarkably complex and remarkably valuable systems into easily consumable products is a huge challenge, requiring a remarkably different type of IT company. As you watch the current generation of enterprise IT startups continue to grow, to partner with one another and to innovate on the business side, you’re actually watching the maturation of the next wave of IT giants in the unleased digital marketplaces—tomorrow’s Microsoft, Oracle and SAP. Their technologies and their methods might seem foreign at times, but that’s all part of coming of age in today’s technology landscape. When it’s all said, these are the companies that will dominate the era of datacenter computing.
Part II
Introduction
© Springer-Verlag GmbH Germany 2018
Claudia Linnhoff-Popien, Ralf Schneider and Michael Zaddach (eds.)Digital Marketplaces Unleashedhttps://doi.org/10.1007/978-3-662-49275-8_4
4. Welcome to the Age of Spontaneous Business Models: Start Shaping or Be Shaped
Claudia Linnhoff-Popien1 , Ralf Schneider2 and Michael Zaddach3
(1)Institut für Informatik, LMU München, Munich, Germany
(2)Allianz SE, Munich, Germany
(3)Munich Airport, Munich, Germany
Claudia Linnhoff-Popien (Corresponding author)
Email: [email protected]
Ralf Schneider
Email: [email protected]
Michael Zaddach
Email: [email protected]
4.1 Proliferation – Market Places Branching out
Markets are undergoing continuous diversification. They are becoming more and more individualized. Customization of offerings is a crucial factor for success. The old economy knew standard products. If you were lucky enough, at least there used to be a small group of products to choose from. Both products and services were tailored for mass production and mass consumption. They maybe were aligned to target a specific group of customers. Today, however, clients are being addressed individually and receive highly personalized offers.
Proliferation takes place. Products and business models keep branching out more and more – just like a cauliflower. One has to look for those buds of critical mass capable of attracting millions of customers. These are the buds, i. e., products, worthwhile developing. It does not matter where these customers are physically located. The only thing that counts is their common interest in and need for the same product a vendor can develop and bring to market.
An airline cannot build an own airport for everyone eager to fly. But it can search for groups of travelers willing to take a flight from Munich. Munich Airport itself has highly individualized offerings. There are vegan restaurants and steak houses. Some visitors and passengers only want to go shopping, others seek the tranquility of a lounge.
4.2 Marketplaces Unleashed Becoming Ever Faster
Market places are developing rapidly. They are digital. Even when looking for a physical product, I will choose the provider based on how fast it is able to deliver. I can get everything I want within 24 h. Customers demand instant gratification. Upon being pointed to a specific product, a customer wants it be delivered right away and in highest quality. This concept let corporations like Amazon grow big. They are highly reliable. They know that also for customers, “time is money” holds true. Indeed, they are able to deliver within 24 h or less.
And this time span shall become even shorter. Amazon recently filed a patent for flying warehouses to further reduce physical distance to its customers. The Internet merchant plans to set up so called “Airborne Fulfillment Centers” (AFC), floating up in the sky like an airship. Dependent on customer needs, different goods can be provisioned. They will be brought to the customer by drones, which do not have to queue up in the congested streets of a metropolis, but offer way shorter delivery times by air. Amazon makes a promise – and they will deliver on it.
4.
3 The Emergence of Spontaneous Business Models
Our time has become fast moving. Vendors often do not exactly know yet what they want to offer or what they will be able to do, but they are already looking for first feedback on early thoughts and ideas. If an idea is well received, a corresponding product will be developed. Sometimes an existing product only has to be upgraded with new features or an existing service optimized. In any case, however, resources will only be spent if an idea goes down well. Market analysis in the classical sense takes a step back and is conducted only concomitantly. New business models are given a try. They evolve spontaneously and will be extended if they are successful and scale. Only under these circumstances the growth rates required for a product to be worth developing are possible.
The market’s complexity is unforeseeable. It cannot be reliably predicted which development comes next. Instead, one has to try an innovation in the wild. Public awareness and interest for a new product that customers are not used to yet have to be created. Even more, they have lived well without it until now. One has to succeed in putting it on the public agenda in order to gain publicity and acceptance.
Digital Marketplaces Unleashed Page 5
