The podcast is back after a delay related to today’s article. Last week’s episode expanded upon the recent Follow the F—ing Instructions article, which ironically received much lower than average reads.
Oh well. I understand that nobody likes being told they have to do something they don’t want to, but such is the real nature of freedom: everything has a price. You don’t have to like it!
For my next trick, here’s yet another thing you won’t like, but is true.
One of the most useful principles in developmental work is also one of the least intuitive: when a person or organization reliably produces the same unwanted result, the first question should not be why the system keeps failing.
The better question is “What mechanics makes the result so reliable?”
Consistency is evidence of structure.
People tend to understand written systems when they produce good results, but tend not to understand the invisible ones that produce bad ones.
Structurally, there is no difference:
Consistent results, good or bad, are a result of repeated patterns. Intentional or not. Written or not. It doesn’t matter.
This is where people fail, because they don’t want to look at the system behind the consistently bad results; understandably so, because it’s usually messy. But it must be done.
I’ve been customer of Hawaiian Telcom for over ten years and I say without exaggeration that it offers consistently the worst customer service I have ever experienced. Stunningly so. And recently I had the misfortunate to become more intimate with its machinations.
I live in rural Maui, where occasional outages are not especially surprising, but recently had one persist for days. A technician, who chewed tobacco and spit frequently, eventually determined that neither the equipment at my house nor the line immediately serving it was responsible.
He drove a mile away to try to fix it and came back reeking of weed. The problem was elsewhere and required escalation to the splicing team, he explained, while I considered whether to call him out on being stoned or not. I didn’t, thinking it was best to keep him on my side.
I called the next day and the work was scheduled for that afternoon Great! The CSR, however, told me that schedule is “tentative,” and it could take longer. Could I call those people to find out? She gave me a number that I would dial many times, but it only ever refrained an unbranded, music-less recorded message about how important my call was and that someone would be with me shortly. No one ever answered, and the recorded message was so hollow and creepy I couldn’t last longer than ten minutes waiting. It reminded me of the recording of the numbers in the TV show Lost.
The next morning, I drove into town on a mission and dropped $350 on Starlink to use as a backup, which I should have done a long time ago, having barely managed with a cell phone booster and tethering headaches. But this was the longest outage ever by far and I’d already spent $40 on increased cell data.
I could still get no commitment from HT to a fix and, more remarkably, could not obtain a meaningful explanation of what was wrong or when it would be repaired.
The experience was frustrating, but the customer-service interactions were more interesting than the outage itself. The amazing thing about Hawaiian Telcom is not that their service is bad, but that it’s bad in such a reliable way.
Different employees and different levels of management produce variations of the same result, as if it were right out of a training manual. Direct questions produce indirect answers. Escalations produce additional conversations without additional authority. Promises to investigate replace commitments to act. Departments could send messages to other departments but apparently could not hold them accountable for responding.
Eventually, the pattern became sufficiently consistent that it deserved to be treated as a system, one that I find myself inside of, and I begin remembering Kafka’s nightmarish novels about bureaucracy based on his own work experience. By the way, if you haven’t seen The Onion’s Kafka International Airport, it’s a must-see filled with easter eggs.
The Kafka Problem
Franz Kafka understood something important about bureaucracy. Its absurdity does not require malicious people, and it does not even require particularly incompetent ones. It requires a structure in which information, authority, action, and responsibility have been sufficiently separated that everyone can perform a role while nobody remains responsible for the whole.
The resulting conversations can become almost comical. I was told, for example, that my repair could not be completed because a bucket truck was required. When I asked what was actually wrong with my internet, I was told again that they needed a bucket truck. I explained that I understood the logistical requirement, but wanted to know what physical problem required the truck. The answer, in various forms, remained that the repair required a bucket truck.
Kafka would have recognized the structure immediately. The information exists somewhere, as does the authority to act upon it, but the person with whom one is permitted to speak possesses neither. The people who possess one may not possess the other, and the customer is prevented from reaching any location at which information and authority converge.
This produces an organization that appears continuously active. Tickets are opened, notes are entered, messages are sent, issues are escalated, supervisors are consulted, departments are notified, and apologies are offered. I can’t tell you how many times I was thanked for my patience, despite not exhibiting any whatsoever, which is of and to itself a bit maddening. Considerable activity occurs in a bureaucracy without anyone becoming responsible for producing the actual result.
This suggests an important distinction between procedural accountability and functional accountability. Procedural accountability asks whether the prescribed process has been followed. Functional accountability asks whether someone owns the outcome. An organization can have elaborate systems of procedural accountability while possessing remarkably little functional accountability. The former is a kind of “responsibility theater” and functions as a kind of front-end for an out-of-sight backend that does God-knows-what.
The distinction is easy to recognize in language. Procedural accountability says that a ticket has been opened, the matter has been escalated, another department has been contacted, or someone will look into the problem. Functional accountability sounds different: I own this until it is resolved. This is what happens next. This is who is responsible for the next step. This is when we expect it to occur, and this is what I will do if it does not.
The first describes activity. The second establishes ownership.
How Responsibility Disappears
So, based on my analysis, here’s how you create a bureaucracy with Swiss-watch level reliability. I’m talking about a customer experience that is as consistent as The Ritz Carlton, but instead of reliably great, it’s reliably terrible.
First, separate customer contact from meaningful authority. The people who speak with customers should be capable of documenting and routing problems but unable to resolve anything outside a narrow range of ordinary circumstances. Then divide the organization into specialized departments, each responsible for a portion of the process but none responsible for the customer’s final outcome. The customer must always own that.
Next, restrict communication across those departmental boundaries. Customer service may be able to send a message to dispatch, dispatch may be able to assign something to a technician, and the technician may be able to escalate something to network operations. As long as nobody owns the chain from beginning to end, however, responsibility fragments and the signal degrades like a game of telephone as the problem moves through the organization.
The people with actual decision-making authority should also be impossible for customers to reach directly. Their names and contact information must remain unavailable, and customer-facing employees should be prohibited from making commitments on their behalf. The employee can therefore acknowledge the problem while remaining structurally unable to assume responsibility for its resolution.
Budget allowing, there should be several layers of management that the customer can escalate through, but none should be significantly more competent than the last so that customer loses hope and energy as they move through the ranks. Ideally, anyone the customer has access to should speak with a strong accent, unusual grammar, and be otherwise difficult to understand which serves to further fatigue the customer into giving up.
Common sense must be either unknown or ignored by employees so that the customer must repeatedly explain how and why they expect even the most basic and fair treatment. Employees must not enter such domains of conversation and instead quote company policy which they are bound by along with the customer.
Questions asked by customers that require a Yes or No answer must be not be answered as such. Instead, tire the customer out by repeating previous or already known information in a different way. CSRs should speak laboriously, and always unnecessarily repeat back what the customer has said to provide the appearance of listening while delaying progress and aggravating them, but not in any specific way that could be called out.
When escalating/transferring calls, do not hand off the call to protect the customer from having to repeat their plight. Communicate a sincere desire to help while the customer has to repeat their story and act a little bit surprised and slightly offended if and when this upsets them.
Finally, everyone should be trained to apologize.
This last feature matters more than it initially appears. An apology can provide the interpersonal appearance of responsibility without establishing the structural reality of responsibility. The employee may sincerely regret what is happening while possessing neither the authority nor the obligation to change it. Sympathy therefore substitutes for ownership, and the organization can appear responsive while remaining functionally inert.
The result is a peculiar achievement: everyone can truthfully claim to have performed the responsibilities of the role while the organization as a whole remains responsible for nothing. It’s really quite impressive to behold how much activity can produce no result.
The Customer as Project Manager
Responsibility does not actually disappear from such a system. It migrates.
In my case, there remained one person with end-to-end responsibility for restoring my internet service: me. I had to remember whom I had spoken with, keep track of the ticket, call again when nothing happened, explain the situation repeatedly, discover that one department did not know what another had done, request escalations, follow up on those escalations, locate higher levels of management, and continue asking what would happen next.
The organization possessed the personnel, infrastructure, vehicles, scheduling systems, technical expertise, internal communication systems, and authority required to repair its network. Yet I had effectively become the unpaid project manager responsible for coordinating its response to its own failure.
This is more than poor customer service. It is a structural transfer of responsibility. The organization retains the resources and authority while exporting the burden of integration to the person least capable of performing it: the customer.
Once this is recognized, repeated organizational dysfunction begins to look different. When something goes wrong once, an individual mistake may explain it. When the same basic result appears repeatedly across different employees, departments, managers, and circumstances, individual incompetence becomes progressively less persuasive as an explanation. The consistency of the outcome itself becomes evidence that something systematic is producing it.
Reliable results come from reliable organization.
This remains true even when everyone involved dislikes the results.
If customers are repeatedly bounced among departments, the organization contains structures that reliably produce bouncing. If employees repeatedly cannot answer direct questions, information is distributed in a way that reliably produces ignorance at the point of contact. If managers repeatedly cannot make commitments, authority has been organized in a way that reliably produces powerlessness. If customers repeatedly have to pursue resolution themselves, responsibility has been organized in a way that reliably externalizes ownership.
The organization may sincerely wish that these outcomes were different. Its employees may dislike them, and its executives may publish values statements explicitly opposing them. None of this changes the functional reality. Systems are revealed less by what their participants intend than by what they reliably produce; this is the structural analog of “Actions speak louder than words.”
Structure Is a More Reliable Expression of Values
Organizations routinely claim to value customers, accountability, communication, excellence, responsiveness, and integrity. Such statements tell us relatively little because values expressed abstractly cost almost nothing. Structure is more revealing because structure determines what actually becomes possible when values encounter competing pressures.
If an organization claims to value accountability, the useful questions are structural. Who owns a problem when it crosses departmental boundaries? Who has authority to make a commitment? Can the employee who learns of a failure cause something meaningful to happen? Can one department hold another accountable for responding? Does anyone remain responsible until the customer receives an actual result? When the same failure occurs repeatedly, what mechanism causes the organization itself to learn?
These questions expose the difference between declared values and operative organization. An organization may genuinely believe that it values accountability while being designed in ways that systematically prevent accountability from occurring.
If we want to understand an organization, therefore, we should not begin by asking what it says it values. We should follow the movement of responsibility. Where can responsibility travel, where does it stop, where does it fragment, and where does it disappear? Those boundaries reveal far more about the organization than its stated intentions do.
Bureaucracy Is Made of People
It is tempting to imagine bureaucracy as an impersonal monster imposed upon otherwise responsible individuals, but bureaucracies are composed of people, and systems and individuals continuously shape one another.
A system that chronically diffuses responsibility requires people who are willing, or have gradually become willing, to participate in that diffusion. This does not mean that every employee in a dysfunctional bureaucracy is personally irresponsible. People need jobs, inherit systems they did not design, and frequently possess little power to change them. A highly responsible person can certainly find themselves inside an irresponsible structure, but one with integrity would be actively looking for another job.
Nevertheless, such a structure exerts developmental pressure upon the people within it. Organizations recruit, reward, punish, exhaust, normalize, and select for particular forms of functioning. Someone with a strong internal standard of ownership may repeatedly collide with a system that prevents ownership, attempt to change it, leave it, or eventually adapt to it. Over time, phrases such as that’s not my department cease to seem like descriptions of organizational limitations and begin to seem like descriptions of reality itself.
This is one reason dysfunctional systems can become remarkably stable. The system does not merely organize behavior; it gradually recruits people capable of reproducing the organization that already exists. Individuals maintain the system, and the system conditions the individuals.
The Invisible Systems of Individual Life
The same analysis becomes considerably less comfortable when applied to individuals.
Organizations are easy to criticize because their dysfunction is external to us. Our own repeated results are more difficult to examine structurally because we tend to explain them through circumstances, intentions, isolated decisions, or failures of effort.
Someone repeatedly runs out of money before the end of the month, repeatedly becomes overwhelmed, repeatedly enters relationships with unavailable people, repeatedly fails to exercise, repeatedly becomes responsible for other people’s problems, or repeatedly cannot find time to meditate. Each individual occurrence can be explained plausibly. There was an unexpected expense, work became unusually demanding, this particular relationship was complicated, the week was unusually busy, or there simply was “not enough time” as if that varies.
At some point, however, repetition changes the appropriate level of analysis. The question is no longer merely why the result happened this time. The more important question becomes what organization makes this class of result so reliable.
There is no hidden supply of time waiting to be discovered. There is an organization of priorities, commitments, avoidance, rewards, fears, identities, habits, compensations, and competing demands that reliably allocates time as it does.
Likewise, chronic overwhelm is rarely produced only by an unfortunate sequence of unusually demanding weeks. It may be produced by an organization that repeatedly accepts obligations, avoids limits, derives identity from usefulness, postpones difficult choices, or externalizes the responsibility for prioritization.
The unwanted result is real, but so is the system producing it.
This is why the principle that applies to organizations applies equally to individuals:
Reliable results imply reliable organization.
Indeed, the more reliably an unwanted result appears, the more interested we should become in the invisible organization that produces it.
Intention Layered Over Structure
Most attempts at personal change begin at the wrong level because people respond to systemic results by increasing intention.
They resolve to become more disciplined, communicate better, meditate every morning, stop procrastinating, establish stronger boundaries, exercise regularly, spend less money, or make more time for themselves. These intentions may be entirely sincere, but sincerity does not alter structure.
When the underlying organization remains intact, the new intention must fight that organization repeatedly. This is sometimes referred to as “self-discipline.” The person may succeed temporarily through increased effort, but the effort is applied against a system whose default output has not changed. Eventually the system reasserts itself, the old result returns, and the person concludes that they need to try harder.
This is the personal equivalent of opening another ticket.
The failure is not necessarily insufficient motivation. The intervention may simply be occurring at the wrong level. New initiatives are layered over an invisible system that continues to generate the deeper result.
Development therefore requires a different class of question. Instead of asking only how to force a different behavior, we can ask what makes the existing behavior so reliable? What is the pattern accomplishing? What does it protect? Where does responsibility move when pressure appears? What reliably happens immediately before the familiar outcome?
What identity, relationship, emotional equilibrium, or compensatory structure would be disturbed if the pattern disappeared? There’s almost always some cost not willing to be paid.
These questions move inquiry beneath intention and toward organization.
A mature system does not require heroic effort every day merely to prevent itself from producing its default outcome. Development changes the default by reorganizing the conditions from which behavior emerges. The goal is not simply to exert greater force against an unchanged structure but to alter the structure sufficiently that different behavior becomes increasingly natural.
This is as true of individuals as it is of families, businesses, institutions, and telecommunications companies. Persistent dysfunction should therefore provoke curiosity before exhortation. When a result keeps appearing despite repeated declarations that nobody wants it, we should become less interested in promises to try harder and more interested in the architecture producing the result.
The system may not be failing at all.
It may be working exactly as designed.
Epilogue: As I schedule this for release, fiber has been down for just under three weeks, with no communicated schedule for a fix.