Designing delivery: A new model of control
As companies have increasingly automated internal processes and customer interactions, IT has become part of the control mechanism itself.
As companies have increasingly automated internal processes and customer interactions, IT has become part of the control mechanism itself.
Post-industrialism challenges twenty-first-century businesses to become more open, responsive, experimental, and resilient. Companies still, though, exist to generate profit. They aren’t indifferent to the results of their efforts. Profit isn’t something companies are happy just to let ebb and flow naturally. They direct all their efforts toward controlling it and ensuring that it continually moves in the right direction.
Profit allows a company to perpetuate its existence, and to replenish and grow itself. Growth lets public companies satisfy the imperative to create shareholder value. Even small, private companies pursue profit and growth in order to generate financial fuel for personal goals like sending your children to college or affording retirement.
One might say that control is the most basic activity that defines a corporation. Much of its activity will be directed toward controlling the parameters that contribute to profit and growth. In the quest to maximize profit, companies must control both internal and external parameters. Internal parameters include cost of materials, lead time, product quality, information flow, operational procedures, and employee behavior. External parameters include market share, income, customer behavior, and stock price.
IT came into being to aid the quest for control. Traditionally it provided information needed by internal and external control mechanisms. As companies have increasingly automated internal processes and customer interactions, IT has become part of the control mechanism itself. Understanding how post-industrialism changes the nature of control is critical to understanding IT’s new mandate.
As much as anything, the industrial era will be remembered for the emergence of powerful business control mechanisms. The industrial model of control is largely unidirectional. Information flows down and out along hierarchical management chains. Production flows forward through assembly lines and out to customers. Marketing (the attempt to control customer behavior) flows out from design agencies to media outlets to consumers.
Industrial control systems focus on maximizing efficiency in order to optimize unidirectional flow. Scientific management techniques such as Taylorism and Fordism drew inspiration from time-and-motion studies. They looked for ways to wring inefficiencies out of the production process. This approach implied that it was possible to analyze a process, identify and remove waste, and then operate the improved version in an ongoing, steady-state fashion. Having identified a more efficient operating model, one could implement that model and then run it unchanged for lengthy periods of time.
Separating design from implementation and management from execution was central to Taylorism and Fordism. Industrial control relies on decomposing global mechanisms into subcomponents and removing variation within them. On a human level, industrial control mechanisms view workers as integral parts of an overarching machine. The production process has primacy over the people participating in it. Designers develop an overall control model, and then break it down into component parts. Managers instruct and oversee workers to ensure work happens according to plan.
Corporate IT arose as a tool for managing industrial-style control systems. It reflected the push model of product companies. IT helped companies track business-control parameters such as inventory, cost, sales, and defect rates. Workflow management systems helped manage procedures that flowed through complicated-system organizational and process structures. Manufacturing automation systems increased efficiency and quality on industrial shop floors. IT even developed industrial-era control models such as ITIL1 to manage itself.
Post-industrialism is forcing companies to rethink their reliance on industrial control mechanisms. It challenges the effectiveness of designing static, long-running production processes. It questions whether optimization is feasible or even desirable. It confounds organizational structures based on hierarchy and unidirectional information flow. It eludes the capabilities of workforces that are at once conformist and disconnected.
Service, digital infusion, complexity, and disruption all contribute to challenging the applicability of industrial control mechanisms to the post-industrial economy. Service necessitates conversation and collaboration between company and customer. Controlling the customer relationship becomes a matter of satisfying rather than convincing. In order to satisfy customers, a service provider must be able to listen. The entire organization must be internally aligned so it can hear the customer accurately and then cooperate to generate an appropriate response.
Compartmentalized, hierarchical corporate organizational structures, characterized by one-way information flow, impede the ability to listen, hear, and respond. This impedance mismatch shows itself through all-too-familiar customer satisfaction failures. We’ve all experienced the frustration that comes from asking a company representative for help only to be told “that’s not my department.” That response might be appropriate from the point of view of optimizing internal efficiency. From a global control perspective, however, it has backfired. By frustrating customers, the interaction incentivizes them to find an alternative service provider and take their money elsewhere.
Digital infusion further stresses compartmentalization. No longer are systems of record disjointed from systems of engagement. Not only must the systems interconnect; their design, development, and operational processes, and the people who run them, also must interoperate. IT organizations generally manage systems of record using rigid processes that intentionally slow down the rate of change in response to constraints such as regulatory rules. Systems of engagement, on the other hand, must respond to marketing pressure for quick and continual change. The digitally infused business must find a way to blend or at least synchronize these differing approaches.
The industrial company is the epitome of a complicated system. It thrives on predictability and stability. Industrial systems management views failure as something to prevent. It tries to create fail-safe systems by structuring components into rigid layers. Each layer depends on the robustness of the layer below it. Robustness within a layer becomes a matter of trying to maximize the mean time between failures (MTBF) of each component.
Top-down industrial control poses the greatest threat to complex systems such as post-industrial organizations and IT systems. These systems excel at responding to failure, but wither in the presence of too much stability. Fail-safe design solutions create brittleness instead of robustness by thwarting the open, dynamic connections that characterize complexity. Due to their sloppy, failure-prone yet resilient nature, complex systems need the ability to absorb component-level failures instead of trying to prevent them. They also need the ability to continuously recalibrate their survival strategy in the face of ever-changing environments.
The assembly line is the central image of industrialism. Even nonmanufacturing companies use the assembly line as a design metaphor for business processes. In the age of disruption, the assembly line is problematic. It relies on significant up-front investment in design and optimization. It assumes the ability to develop a stable production mechanism, which can then be operated without major changes for long periods of time. It makes rapid change difficult. As a result, it amplifies the Innovator’s Dilemma by incentivizing companies to keep doing what they know how to do in the ways that they know how to do it.
Twenty-first-century businesses still need to generate continuous profit and growth. Now, though, they need a new way to drive the economic engine. They need the ability to synchronize themselves with their customers and the market much more intimately. When brands become conversations, the problem of controlling your brand requires a new strategy.
To accomplish the level of synchronization that brand conversations need, post-industrial businesses must minimize the delay between incoming signals and outgoing replies. They need the ability to continually change themselves in response to those signals. They need a mechanism that allows them to navigate uncertain, unknowable, continually changing environments. In other words, they need to turn the familiar, industrial model of control completely on its head. Because IT is integral to controlling the digitally infused business, it must fully participate in this transformation.
Ironically, in the search for a post-industrial model of control, we can look all the way back to the dawn of computing in the mid-twentieth century. The work of Norbert Wiener and others in the 1940s and 1950s offers a powerful model for adaptive control via conversation. Wiener was a major contributor to the original conceptions of computing and information theory. In the process, he led the creation of a new, holistic, cross-functional discipline called cybernetics.
During World War II, Wiener helped the US military with its attempts to develop automatic targeting for anti-aircraft guns. It was his idea to use knowledge about physics, airplane design, and pilots’ cognitive processes to predict an attacking plane’s flight path. Wiener recognized that a plane couldn’t instantaneously change its position from one random position in the sky to another. If it did, it would tear itself to pieces. Its evasive capacity was also influenced by limitations on the pilot’s ability to calculate new maneuvers. The more quickly he had to respond, the more likely he would be to resort to habitual maneuvers. Wiener believed it should be possible to mathematically compute probable flight paths, and use those computations to automatically point the gun at a location that was likely to intersect with the plane’s path.
Wiener’s research confronted him with two problems. First, it wasn’t possible to perfectly predict the path of a human-guided physical object through physical space. Second, even if that were feasible, it still wasn’t possible to perfectly aim a large, heavy gun barrel. Depending on environmental parameters such as air temperature, humidity, and even the age of the grease in the gun turret’s ball bearings, the barrel might swing a little too far or not quite far enough.
Wiener compensated for these unavoidable inaccuracies by building a feedback mechanism into the targeting system. It fed information about the gap between the intended and actual aim, and the predicted and actual flight path, back into the targeting system. Rather than just guessing the plane’s future location as best it could and then pointing and firing as best it could, the system successively approximated the desired aim, correcting itself along the way.
After the war, Wiener collaborated with Arturo Rosenblueth, a researcher in physiology at Harvard. Together they explored various biological functions, including one known as proprioception. The human body uses proprioception to control physiological processes such as walking. It loosely refers to muscle sense, or feedback from muscles, which allows the brain to judge progressive movement toward a target. A field sobriety test, for example, tests for impairment in your ability to successively move your finger toward your nose.
When you walk, your brain evaluates information telling it how far your foot still is from the ground. It then moves your foot closer, and reevaluates the distance to the ground until your body makes the next step. This process comprises a continuous, feedback-based control loop. The brain directs the leg to act, then listens to information “fed back” to it about the result of its action, then directs the leg to act some more, and so on. This process is illustrated in Figure 1-1.
Wiener’s work on weapon targeting, adaptive behavior, information theory, and proprioception all contributed to the development of cybernetics. Cybernetics formalized the science of feedback. The name comes from the Greek word for steersman. It shares its etymology with the word governor.
Cybernetics treats control as a dynamic process of maintaining homeostasis by continually processing and responding to feedback. It uses information about outcomes in the past to control actions in the future. If the gun barrel swung too far, by how much was it off? How inaccurate was our prediction of the plane’s next evasive maneuver? How far do I still need to move my foot to reach the ground?
Circular causality is a key characteristic of cybernetic control. Feedback loops push information about behavior from the past forward into future decisions. Cybernetic processes such as proprioception define circular relationships between the brain and the body. We generally think about the brain–body relationship in industrial terms, with the brain as manager and the leg as worker. According to proprioception, though, the brain paradoxically depends on the leg to inform its decisions.
Cybernetics assumes a constantly changing, unpredictable world through which we must steer our way. We can’t perfectly know either the present or future state of our environment. We must therefore incorporate listening into our actions so that they become inseparable parts of a continuous process. A sailor or oarsman, for example, must continually compensate for changes in wind or current. Without going in a perfectly straight line, the boat nonetheless wends its way under the steersman’s guidance to the desired destination on the other side of the lake.
Cybernetics provides a language that expresses the essence of post-industrial business. In addition, that language describes the fundamental nature of post-industrial IT. Contemporary methodologies such as Agile and DevOps, for example, use feedback loops to let development and operations teams steer in response to continually changing a requirements.
Wiener and other early cyberneticians participated in a series of symposiums in the 1940s and 1950s called the Macy Conferences. These conferences brought together researchers from a wide variety of disciplines, including mathematics, biology, psychiatry, and sociology. Together, they hatched the beginnings of systems thinking, complex systems studies, and cognitive science. They explored the potential applications of cybernetics to everything from modeling the human mind, to rethinking psychological counseling, to managing businesses, to steering entire national economies.
The first-generation cyberneticists tended to take a mechanistic approach to controlling systems of all kinds. Wiener introduced the new discipline to the world in 1948 with the publication of his book, Cybernetics: Or Control and Communication in the Animal and the Machine. Much of the original cybernetics work thought more in terms of control and machines than communication and animals.
This approach could become problematic when applied to complex systems such as people and societies. On the one hand, it does makes some sense to think about mental health as operating within a desired range of behavior. The psychologist could be seen to provide a feedback-based regulatory function, helping the patient bring themselves out of neurosis and back into homeostasis. On the other hand, an acceptable doctor–patient relationship must respect the intelligence and autonomy of both parties. The therapeutic relationship itself becomes ill if it views the patient as a device to be controlled.
Very quickly though, some of the original cyberneticians began thinking at a higher level. Margaret Mead and Heinz von Foerster introduced the concept of second-order cybernetics, also called the cybernetics of cybernetics, or New Cybernetics. It distinguished, in von Foerster’s words, between “the cybernetics of observed systems” and “the cybernetics of observing systems.”
Second-order cybernetics is inherently holistic. It recognizes that a first-order cybernetic system such as a thermostat is itself cybernetically controlled, in this case by the human who programs it. A thermostat has no purpose without the room that contains it, along with the air it’s controlling. The room in turn exists within a house. The person who lives in that house sets the thermostat so that it’s warm in the morning when they get up to go to work. They go to work because they’re employed by a company. They drive to work in a car that was made by a car manufacturer, on a road that’s maintained by the town in which they live.
On another level, second-order cybernetics restates relationships such as therapy in terms of interactions between autonomous systems that are themselves complex. It emphasizes the animal and communication over the machine and control. It presents a model that is interactive and mutual rather than manipulative and hierarchical.
Von Foerster’s concept of “the cybernetics of observing systems” captures the fact that complex systems always interact with one another from their own points of view. None of them holds objective knowledge. Instead, they negotiate shared understanding through conversation. That shared understanding is also doomed to imperfection. It continually breaks down and must be repaired as part of the dynamic process of life.
Cybernetics views existence as a co-creative conversation with your world, rather than a disconnected process of ingesting and interpreting information from it. Though it might sound like something out of postmodern philosophy, this worldview is directly relevant to post-industrial business. Digitally infused service breaks down traditional boundaries between vendor and customer, between virtual and physical, and between internal and external systems and processes. In order to provide quality service in such an environment, we need to augment our traditional, industrial-era reductionist thinking with holistic, integrative systems thinking.
In order to change our approach to value from delivery to co-creation, and our approach to brand from broadcast to conversation, we need to embrace the cybernetic view of reality as conversationally co-constructed. Business no longer consists of making things and then trying to convince people that these things are what you claim they are or useful in the ways you claim. Particularly in the era of social media, things are what people perceive them to be. The nature of a service is inseparable from its use; its usefulness arises from the collaboration between vendor and customer.
Complexity defies objective knowledge. Not only can we not fully model a complex system at any point in time and not only does the system we’re trying to model continually change, but our attempts to model it contribute to changing it. Rather than managing complex systems from above, we engage with them as participants. Our understanding of the systems we create, manage, and operate reflects our particular perspective as observers. It is necessarily relative and provisional.
Given that complexity increasingly characterizes our businesses and the IT systems we use to manage them, we need a new model of control. Cybernetics provides a model that accurately reflects the interactive nature of our relationships with these systems. That model frees us from the temptation to try to harness complexity through industrial-style control. It allows us to understand the limits of our ability to predict complex behavior, and a concrete method for replacing our attempts at prediction-based, top-down control with feedback-based steering.
Cybernetics’ insights into the co-creation and circular causality found their fullest expression in the work of Humberto Maturana. Maturana is a Chilean biologist who studied with Warren McCulloch, von Foerster, and other founding cybernetics luminaries. In particular, it was the biological experiments that he helped McCulloch conduct that set Maturana on the path to his unique contribution to biological cybernetics.
Maturana and McCulloch were investigating the visual systems of frogs. They made a surprising discovery: a frog’s vision is specifically constructed to see small, fast-moving objects (such as flies), and to ignore large, slow-moving objects. In other words, the frog sees the world in a particular way not because of how it interprets reality, but because of its internal structure.
These experimental results led Maturana to reconsider his views on reality, knowledge, and cognition. Being a biologist, he found himself struggling with the essential definition of what it means to be a living system. He arrived at a definition of life that, in typical cybernetic fashion, required the invention of a new word derived from Greek. In his case, it was autopoiesis.
Autopoiesis means self-production. Maturana distinguished it from allopoiesis, or other-creation. A thermostat exists in order to change the temperature of the air in a room. A word processor exists in order to produce text documents. A living system, on the other hand, exists in order to perpetuate its own existence.
Autopoiesis is the process by which components of a system work together to create the conditions for their own production (the process is illustrated in Figure 1-2). A human being, for example, exists through the interaction of lungs, heart, brain, and muscles. By continuously finding food and shelter, the human makes it possible for the lungs, heart, and brain to continue to function. If I develop lung cancer, I need to get treatment for it. If I don’t and my lungs fail, I will die. I die not just because I have lungs but because they provide a function upon which the other parts of me (as well as me as a holistic living system) depend. Without my lungs, the rest of my organs can no longer work together to keep me alive.
Autopoiesis may seem like a solipsistic view of life, but it only represents part of the story. Maturana explored autopoietic systems and their relationships to their environments in a series of books he coauthored with his student and colleague, Francisco Varela. According to them, a system’s autopoiesis must be appropriate to its environment in order to be sustainable. They called this requirement structural coupling. If a frog is constructed to see flies, for example, but there are no flies in the frog’s environment, then it will die.
Structural coupling happens by way of perturbation. Living systems induce behaviors within one another as part of sharing an environment. Frogs jump and stick out their tongues because they see flies. Antelope look around and flick their tails because they share water holes with lions. People sleep in tents with netting because they live in environments with mosquito-borne malaria.
Structural coupling is not, however, a static process. Autopoiesis always occurs within an environment that is continuously changing and evolving. In order to maintain its structural coupling with its environment and thus its viability, a living system must also change and evolve. Maturana and Varela used the term self-steering to refer to the process of autopoietic adaptation.
The concept of self-steering posits that living systems use a cybernetic process to maintain their autonomy within their environment. An individual’s response to a lung cancer diagnosis is such a mechanism. By taking a leave from work to receive chemotherapy treatments, that person is changing his situation in order to maintain autonomy—that is, to stay alive. Self-steering paradoxically lets a living system change, or adapt, in order to stay the same.
Self-steering implies that living systems don’t operate in slaved response to stimuli from the outside world. Their structural coupling with their environment causes them to respond to outside perturbations through adaptation. In the process they perturb their environment, causing the other autopoietic systems in that environment to adapt in turn. Autopoietic systems swim around, as it were, co-evolving along with other living systems, jointly creating a shared, dynamic reality.
Maturana and Varela introduced the theory of autopoiesis in the book Autopoiesis and Cognition. Their use of the word cognition refers to the claim that cognition is nothing more nor less than the mechanism by which a living system (re)-creates itself through self-steering. Cognition is therefore common to all autonomous systems. According to Maturana and Varela, the presence or lack of an advanced nervous system is irrelevant to the ability to cognize. Companies and the organizational subsystems (teams, departments, pods, etc.) that compose them can self-steer in the same manner as cells, fish, or people.
Cybernetics concerns itself with adaptive, purposeful systems. The purpose of an automatic gun turret is to shoot down planes. The purpose of a thermostat is to keep the temperature in a room comfortable. Maturana and Varela’s special contribution to systems thinking was their insight that a living system is not just a static arrangement of interrelated things. Its structure is a continuous, dynamic process of cognition for the purpose of staying alive. Because its surrounding environment is continually changing, the system must continually change itself in order to fulfill that purpose.
You may be wondering about the relevancy of Maturana’s and Varela’s work. It sounds abstract and esoteric, and revolves around a word that’s nearly impossible to spell! To appreciate its applicability, you need only consider the nature of business. What is the purpose of a company? On one level, a company might have a noble, or perhaps just a cynical, goal within its community. On a more primal level, though, if the company doesn’t stay in business, it can’t achieve any of those larger goals.
A company exists as a network of employees, IT systems, physical plant, teams, departments, and divisions, all of which conspire together to create products and services that the company can sell. Gaining revenue through sales allows the company to pay for its employees and IT systems and physical plant in order to create and sell more products and services.
One could thus view a company as an autopoietic living system. Furthermore, one might claim that the post-industrial economy—characterized as it is by service conversations, complexity, and disruption—requires a self-consciously autopoietic business strategy. When continual change becomes the primary characteristic of your environment, continual internal and external adaption must become the primary capability of your operating model. The key success criterion of IT in turn becomes its ability to power a business’s autopoiesis.
The history of Apple offers a perfect example of the corporation as an autopoietic system. During his introduction of the Apple Watch, CEO Tim Cook repeatedly used the word personal. Many readers may not be old enough to fully appreciate this reference. In the 1980s, Apple shipped the Apple IIe in a box with the slogan “The Personal Computer” printed on the side. The idea was that the IIe and the Macintosh after it were very personal devices. The IBM PC, by contrast, was (at least in Apple’s mind) a very impersonal device.
Since then, Apple has undergone dramatic changes. They’ve gone through three profound CEO transitions: from Steve Jobs to no Jobs, from no Jobs back to Jobs, and from Jobs to Cook. The company has hired and fired people, restructured its org chart, and built new buildings. Recently, it’s gone so far as to recruit fashion industry executives and Australian furniture designers. It’s shifted from being a computer company to, according to some analysts, primarily being a telephone handset company. It’s on the way to becoming—who knows? A music industry company? A banking company? A jewelry company?
Through all of these changes, though, Apple has stayed in the business of creating personal experiences. It is the quintessential self-steering organization. On the one hand, it’s changed in nearly every way. On the other hand, it’s undergone these changes precisely for the purpose of maintaining its essential identity in the face of a profoundly transformed environment.
Maturana’s definition of living systems as autopoietic continued the tradition of cybernetics’s fascination with circular causation. A living system embodies a circular relationship between the system and its components. Maturana’s biological background led him to define that relationship in terms of processes. Components participate in the operation of a holistic process that in turn creates the components. Life requires continual action; stasis results in death. The dynamic nature of viable systems applies equally to cells, people, businesses, and societies. In each case, what they are is what they do and how they change in response to their environment.
The theory of autopoietic self-steering carried on the simple yet profound conception of reality as a continual unfolding that marks all of cybernetics. Although first-generation cybernetics tended to think in terms of control, its understanding of the nature of control was revolutionary. The most basic definition of cybernetics contains Ranulph Glanville’s observation that “the controller is controlled by the controlled.” Every cybernetic process implicitly involves conversation and holistically points upward toward systems and away from reductionism.
The thermostat, for example, is the “Hello, World” of cybernetics. The truth is that thermostats don’t really control the temperature of the air in a room. It’s not as if a thermostat can force the air molecules to move at a specified average speed for some amount of time. Instead, it listens to the current temperature (by being physically deformed by the air around it). Based on what it hears, the thermostat influences the air to change its temperature by introducing warmer air from a furnace. The air in the room happily lets itself be influenced, but thanks to the second law of thermodynamics, refuses to behave and quickly starts cooling off again. The thermostat is thus forced to continually adjust. It is the controlled as much as the controller.
By considering Maturana’s work on autopoiesis, autonomy, and self-steering, along with revisiting the nature of thermostats, we can restate cybernetics from a less mechanistic, control-oriented perspective. Cybernetic systems interact with one another by way of conversations. In order to maintain existence, autonomy, and integrity, one must be able to listen. Autopoietic systems must listen to one another and to themselves (or more accurately to their component parts). Living systems co-create reality through circular influence. Circular influence is, and happens via, a conversation.
Cybernetics was extremely popular and influential in the 1950s and 1960s. For various reasons (some having to due with Norbert Wiener’s eccentric personality), it began to lose influence and has largely been forgotten. Its fate is ironic, given the popularization of innumerable words that start with the prefix cyber.
Regardless of its history, cybernetics is more relevant than ever. By interleaving planning and implementation into a unified process, cybernetic systems transform execution into continuous learning. This kind of control as conversation is exactly what post-industrial businesses need to confront the challenges posed by service, infusion, complexity, and disruption. In fact, we can see the trace of cybernetics in many of the methodologies companies are adopting to confront twenty-first-century business imperatives.
Lean Startup is a popular new product development methodology that exemplifies the cybernetic model. Lean Startup warns against committing up front to large-scale, long-term plans for implementing a product vision. If we do commit too early, we risk wasting resources building something that doesn’t match the market and therefore doesn’t deliver value either to the company or the customer.
Instead, Lean Startup counsels us to build the smallest possible version of our vision that lets us test it against the market as quickly as possible. Feedback from our tests provides insight with which to refine our vision and our implementation. By executing this build-test-learn cycle repeatedly, we steer our way toward a truly viable product.
Lean Startup incorporates the idea of a pivot. Sometimes what we learn is that we’re trying to solve the wrong problem. We might have misunderstood the problem in the first place or it may have changed. In either case, we need to pivot and start learning/steering in a new direction.
If this all sounds familiar, it should. Imagine the customer as a plane you’re trying to capture. Your product vision is your prediction of the customer’s flight path: if I build something with functionality X, it will match up with the customer’s needs at point Y. The build-test-learn cycle reflects the recognition that you can’t perfectly predict the precise location of point Y, nor can you perfectly aim your product there. You need to incorporate feedback in order to successively correct your aim. Pivoting happens when you learn that you mis-predicted customers’ goals and desires (or that they changed them in the midst of your product development).
First-order cybernetics helps us understand how to manage individual activities such as product development projects. Second-order cybernetics offers a model for thinking about businesses as wholes. We can view companies and their customers as autopoietic, self-steering systems. Companies exist for the purpose of maintaining their own existence (sustained negative cash flow leads to a company’s death, whereas profit allows it to replenish and grow its components and thus itself).
Customers, for their part, need to interact with companies as part of their own autopoiesis (without someone to deliver heating oil, I can’t use my thermostat to keep my house warm). Companies self-steer in the face of threats to their autonomy: a company introduces a new product to ward off a competitor. Customers do the same: if heating oil is too expensive for me to afford, I try to find a less expensive supplier or switch my heating system to use natural gas.
Post-industrial businesses need to replace the assembly line with the thermostat, the turret gun, and the living organism as their organizing metaphors. No longer can they design their systems and processes for linear efficiency. No longer can they afford to devote significant time to up-front development under the assumption they can run a single “assembly line” at length once they’ve optimized it. Instead, they must design and evaluate systems and processes for their ability to process and respond to feedback: to continuously recognize and measure inaccuracies in their customer understanding, and to recalibrate themselves based on those measurements.
In the age of disruption, efficiency becomes a measure of how quickly and accurately businesses and their component parts can adapt. Companies’ ability to survive as autopoietic systems depends on their ability to self-steer through conversation and the effectiveness with which their IT organizations can power that conversation. Even the simple thermostat can quickly and easily be reprogrammed at the whim of the home’s occupant. In the post-industrial economy, the quality of the thing gives way to the quality of the conversation.
In order to help customers accomplish their goals, digital service businesses need a unity of purpose across the entire organization. At the same time, complexity and disruption necessitate the ability to flex and change internally. Autopoiesis, expressed as self-steering through conversation, provides a unifying perspective that lets companies resolve these apparently contradictory needs. The metaphor of self-steering captures the mechanism by which organically organized, complex adaptive businesses interact with their environments.
Autopoietic systems function through circulatory causality. Their components exist because of and in service of the system. Lungs have no purpose without a body and can’t survive outside of it. At the same time, the system exists because of and in service of its components. The body contains lymphatic and circulatory systems that keep the lungs working. Without all of those subsystems, there is no body. The body also comprises eyes and ears that let it sense the environment. When the ears hear a lion and the legs make the body run away, the eyes and lungs (along with all the other body parts) survive to face another day.
We can think of a post-industrial company in similar terms. Teams and departments have cybernetic, conversational relationships with each other. They are structurally coupled with each other and adapt to each other’s perturbations. The corporation as a whole, and its vision and mission, define the environment within which those conversations take place.
The company simultaneously carries on a conversation with its customers by way of its components. Marketing, finance, product development, and IT all must converse with one another and with the customer. Based on feedback from customers, those teams all contribute to helping the corporation adapt and survive through its own structural coupling with the competitive market.
Just as a biological system self-steers by way of mutual service between system and components, so too does a corporation. The company co-creates value with customers through service. In order to make that exchange of value possible, the parts of the organization (marketing, design, development, operations, finance, etc.) all need to co-create service value with one another. Service at all levels is mutual and conversational, and thus cybernetic.
Successful self-steering depends on empathy. Feedback represents an outside perspective. In order to properly respond to feedback and thus steer productively, an organization must accurately understand that perspective on its own terms.
It’s important to understand that empathy does not mean understanding someone else’s needs, nor does it mean feeling sorry for them. Empathy is rather “the intellectual identification with or vicarious experiencing of the feelings, thoughts, or attitudes of another.” The ability to empathize is the ability to understand a situation from another’s point of view.
This definition doesn’t imply wallowing in another’s pain. Just because we can empathize with a depressed friend doesn’t mean we become similarly paralyzed by gloom. Although explanations of empathy often use painful experiences to differentiate it from sympathy, this definition doesn’t necessarily imply a relation to pain at all. For example, it’s possible to empathize with other people’s aspirations as well as their frustrations.
Empathy can also function on a much more mundane level. Why does someone navigate a user interface the way they do when there are “more efficient” ways to do it? Why do they take a certain route to work when there’s a shorter way?
Empathy is the basis for systems thinking. Systems emerge from relationships between components. Relationships happen through conversation. Conversation requires the ability to comprehend what another party “is trying to say to you”; in other words, the ability to see things from the other’s perspective. Otherwise, the participants in a relationship are just talking at each other without actually relating to each other. Without relationship, there is no system.
A self-steering system maintains structural coupling by allowing itself to be perturbed by its environment. In a sense, it “lets its counterpart in.” By seeing itself from another’s point of view, a living system can steer toward continued life and away from impending death. This impetus applies just as much to businesses relating to their customers and competitors as it does to frogs and human beings. The purpose of a business might be selfish; without the ability to see itself from the perspective of the market, however, a company will lose customers, and thus revenue, and thus viability.
According to designer and researcher Seung Chan Lim’s book Realizing Empathy, true innovation requires empathic conversation. Without it, innovation remains merely a generator of novelties that are interesting without being useful. A successful business doesn’t just need the ability to change; it needs the ability to change in the right direction, at the right time, for the right reason. Empathy makes change meaningful rather than random.
Lim describes the development of empathy as an unfolding process of speaking, listening, and processing. The participants in an empathic conversation progressively uncover each other’s perspectives. In other words, they use feedback loops to steer their way toward mutual recognition: “No, that’s not what I meant; what I really meant was this…”. One could say that cybernetics generates empathy, and that empathy is cybernetic.
Cybernetics’s approach to control is so radical because it incorporates listening and responding. The word conversation comes from the Latin for “to turn with.” In order to converse with someone, you have to continually switch back and forth between speaking and listening. A feedback loop cycles between acting and asking. Successful control, in the form of accurately understanding and responding to the answer that comes back from the environment, requires empathy. As Heinz von Foerster succinctly said, “I like cybernetics: its intrinsic circularity helps me see myself through the eyes of the other.”
While people often think of cyborgs when they hear the word “cybernetics,” nothing about it necessarily implies artificial intelligence or even computerization. Cybernetics is a conceptual model for the way various kinds of complex systems relate to the world. Particularly in Britain, cybernetics evolved as a way of thinking about human interaction and organizational management. Stafford Beer, a leading British figure in the field of operations research, applied cybernetic principles to the management of companies and even entire economies.2
Modern businesses do, however, rely on computerized systems as a critical aid to managing themselves. IT both reflects and makes possible certain kinds of information flow, activities, and relationships within organizations. Business operations and IT mirror each other. A business that seeks to manage itself and its customer relationships cybernetically needs an IT organization that takes the same approach to its systems, practices, and relationship to the larger business.
Digital infusion makes the relationship between IT and the surrounding organization even more intimate. IT already has become the mechanism through which companies operate themselves and communicate internally; now it needs to extend itself to become an enabler for empathic customer engagement. In the digital service economy, companies’ ability to co-create value with customers, as well as their ability to sustain themselves through adaptation, both depend on IT.
Post-industrial companies need a medium through which they can conduct the digital conversations that define their brands. This medium must be cybernetic in nature. It must enable self-steering, allowing companies to respond fluidly to perturbations from their customers and from the market. It must also allow the components of those companies to respond to perturbations from each other. In order for IT to take on that role and become a digital conversational medium, it will have to deeply absorb the cybernetic perspective on systems and relationships.