Beyond Endurance

These signs only diagnose anything as a set, and only when they cover the whole operating model. Track first response time for contracts and leave regulatory unmeasured, and you learn something about contracts. You learn nothing about whether stress is moving from one part of the function to another, which is the question that matters. Detection needs breadth, and legal runs across a wider spectrum than most departments.

WIP, intake rate, and cycle time aren't independent measures. Little's Law ties them together as three sides of one equation, where knowing two gives you the third. Track all three and the function can see whether its operating model is in balance.

Finance changed because the speed of business forced it to. Legal faces the same pressure now.

Section 5 — Friction as Signal

Be hard on the process, not the people.

Paper III made this point from the outside. What gets called friction is the network doing its job. The same holds from inside the operating model. The friction legal runs into in its own operations is signal, not failure.

Resilient systems need stressors to evolve. Friction from the business is a leading indicator, a read on where the system needs to adjust. Bottlenecks, delays, and overload are the diagnostic signals it gives off.

When operational friction surfaces, ask what the system is trying to teach you. Who is responsible comes later, if it comes up at all. Part of the answer is instrumentation. The harder part is cultural, moving from assigning blame to thinking in systems. Without that change, adaptive operations never take hold.

How are you supposed to free up time and resources for change when everyone is already overloaded? You go back to Drucker, Teller, and Little. Stop the activities that get in the way of the task, kill the projects that aren't producing, and pull down work-in-process to free up cycle time. You don't add capacity for change to the operating model. You create it by taking out what shouldn't have been there in the first place.

Section 6 — Compounding Returns on People Investments

"The purpose of an organization is to enable ordinary human beings to do extraordinary things.
No organization can depend on genius; the supply is always scarce and always unpredictable.
It is the test of an organization to make ordinary people perform better than they seem capable
of, to bring out whatever strength there is in its members, and to use each person's strength to
help all the other members perform."

— Peter Drucker, Management: Tasks, Responsibilities, Practices (1973)

Peter Drucker, the father of modern management, coined the term "knowledge worker" for employees whose value is their expertise rather than their physical labor. His argument was that management's job is to build the systems that get the most out of that expertise. Management theory before him was about the factory floor and manual labor, how to make a worker shovel coal faster. You cannot manage a knowledge worker that way. You build systems that put their intellect to work and clear the administrative friction out of the way.

In his 1999 book, Management Challenges for the 21st Century, Drucker put it this way: "one does not 'manage' people. The task is to lead people. And the goal is to make productive the specific strengths and knowledge of each individual." Knowledge workers differ from factory workers in a basic way. They own the "means of production."

Around the same time, McKinsey introduced the "War for Talent" model. It ran on pedigree, on drawing a sharp line between top performers and everyone else, and on a "star" culture that put individual genius above all of it. It gave intellectual cover to "rank-and-yank," the practice of forcing the bottom 10% out every year. Writing as Enron collapsed in 2001, Jeffrey Pfeffer named the mechanism in Fighting the War for Talent is Hazardous to Your Organization's Health. The model bred an arrogance that went around process, it rewarded information hoarding through zero-sum competition, and it shielded the designated stars from accountability. Enron, whose CEO Jeffrey Skilling was a former McKinsey partner, was the model's fullest expression and its most spectacular failure. Its stars got to skip compliance and legal protocols because the culture said you don't question the talent. The parallel to legal is direct. The profession lionizes the lone genius over team-based systems, and the "rockstar" attorney is the same star instinct Pfeffer named at Enron.

McKinsey's instinct was right. People matter. The frame was wrong. Their talent frame depends on finding and keeping the exceptional. Lean entirely on stars and institutional transparency disappears, the risk lives in their inboxes and their heads instead of being visible to the organization. Drucker's frame builds systems where ordinary people become effective.

Jeffrey Pfeffer and Robert Sutton frame the leadership task in their 2006 book Hard Facts, Dangerous Half-Truths, & Total Nonsense: "Perhaps the best way to view leadership is as the task of architecting organizational systems, teams, and cultures – as establishing the conditions and preconditions for others to succeed." The work of legal operations is to put fewer things in the way of the work lawyers are trained for, and to amplify what is already working.

The work is structural: decisions about scheduling, communication, status meetings, and which windows of the day are protected from the frenzy. Edward Hallowell described this in his 2005 Harvard Business Review article "Overloaded Circuits: Why Smart People Underperform" as a focus deficit produced by the environment, one that goes away when the environment changes. Gloria Mark's UC Irvine research found it takes 23 minutes and 15 seconds to resume a task after an interruption. ActivTrak’s 2026 State of the Workplace, built from 443 million hours of monitored work activity, put the average uninterrupted session at 13 minutes and 7 seconds by 2025. The two come from different measures, so the comparison is directional rather than exact, but the gap runs one way: focused stretches are now shorter than the recovery they require. The environment is designable. Without that design, lawyers absorb the cost privately.

Teresa Amabile and Steven Kramer studied nearly 12,000 daily diary entries from 238 knowledge workers and found that making visible progress on meaningful work is the strongest driver of a good day at work. When people can't see progress on the work they trained for, satisfaction erodes.

High-quality legal service takes uninterrupted time on high-value work, given to the right people. Everything else you put in their way erodes the service and the productivity of the people producing it.

Removing what burdens your people is one side of it. The other side, which Papers I and II drew from Batalden and the IHI Framework for Improving Joy in Work (2017), is finding out what matters to people, clearing the impediments with them rather than for them, and iterating on it steadily.

Compound the returns on those people investments and you start to counteract the negative flywheel of the doom loop. Investment in the right conditions builds capability, protects time, deepens engagement, and improves retention, and those compound into the institutional strength of a positive flywheel, the kind that produces extraordinary work.

Section 7 — Adaptive Operation Under Uncertainty

"It ain't what you don't know that gets you into trouble. It's what you know for sure that just ain't so."

— Misattributed to Mark Twain

For decades, the legal industry has “known for sure” that the only way to handle rising volume and risk is to lean on heroic individual endurance. That certainty is what is taking the function apart.

In Paper I we outlined how the legal department was caught in a doom loop, a negative flywheel of connected conditions that together create outcomes that exhaust even the hardiest of legal professionals. Despite the widespread acceptance of the problems individually, there has been no recognition of the connection and self-reinforcing nature of these conditions as a system.

Paper II widened the view to take in the whole enterprise. As long as the doom loop runs, internal clients will first criticize the department and then route around it, and legal has no way to change that story, or even defend itself, because it has no measures or data to point to.

With Paper III we stepped back further and drew the line from market dysfunction through the enterprise and down to legal. Short-termism, toxic incentives, frenetic operating conditions, and the gap between legal and the business all push more pressure onto legal, while management smooths over the short-term bumps and loads the risk into the tail. None of it started with legal, and most of it is not legal's to fix.

Paper IV brings us back to the individual, where people meet all of these forces by working harder and taking on more. Endurance has been the response, and it has been misread. Heroic systems move the load onto individuals and hide the structural problem, because the work still gets done. Robustness, the usual operational aim, isn't enough, because the operating conditions won't hold still long enough for any static model to work.

Across the four papers, the picture is a system that produces suffering, distorts reputations, pushes risk into the tail, falls short of the service and value the business expects, and resists fixing itself. Both sides, the enterprise and legal, have kept up the same fiction, that the individual is the unit of effectiveness. Knowledge work doesn't run that way. It needs shared infrastructure. Endurance was the only response on offer, and it wears people out without changing the conditions that cause the failure. What the department needs instead is a way to sense and respond. That means people, process, measures, and technology working together, so the department can operate under uncertainty without running on human endurance. Here is how you build it.

The reflex is to treat this as a project. Unfreeze the old way of working, change it, refreeze the new way, and move on. That is where the trouble starts, because there is no refreeze. The conditions don't stay put long enough for any fixed arrangement to stay right. Lock in this quarter's design and you've rebuilt the static system the doom loop runs on, already a revision behind reality the day it goes live.

It is also why the last few attempts faded. The reorganization, the platform rollout, the redesigned intake process were each designed, installed, and signed off, and each one met a system that kept moving underneath it. The gain was real for a while. Then the conditions shifted, the design stopped fitting them, and the same pressure showed up somewhere the project wasn't looking. They all failed for the same reason. Each was meant to be finished, and in this system nothing stays finished.

There's a reason you can't build the finished design in advance. You can't know how the people, the process, and the tools will behave together until they're actually running together. That interplay decides whether the design works, and it's the one thing no plan can see ahead of time. So the work is to run a version, watch how those pieces behave together, learn from what you see, adjust, and run the next version.

When you check into a hospital, the staff takes a handful of readings right away, things like temperature, heart rate, blood pressure, respiratory rate, oxygen, and pain level. No single one of them tells you much by itself. Put them together and you have a baseline for that patient, an early warning when something shifts, and a fast read on how urgently to step in and with what.

Medicine figured this out a long time ago. The corporate world, by and large, has not. As Jeffrey Pfeffer and Robert Sutton put it:

"In medicine, the evidence-based movement arose in response to thousands of deaths

and billions of wasted dollars that could have been averted by applying proven

practices. Similarly, in other fields, the growing pile of studies on the human and

financial costs of employee disengagement, management distrust, poor group

dynamics, faulty incentive schemes and other preventable damage suggests a need for

an evidence-based management movement."

— Jeffrey Pfeffer and Robert Sutton, "Trust the Evidence, Not Your Instincts," New York Times, September 3, 2011

Legal has its own vital signs, and they work the same way. They include time to first response on a client request, changes in the volume of requests and in who is sending them, cycle times, where the work backs up, how much each person is carrying, how the work turns out, client satisfaction, and attrition. Any one of them can mislead you. The set is what shows you where to look before something becomes a crisis.

None of this is easy in the early rounds. The useful data points can be hard to capture, and turning them into an accurate, complete picture is harder still.

The general counsel title is accurate. On most days, lawyers handle several areas of law, so the vital signs have to cover the full scope of their work, and the whole team's. A reading from one domain tells you nothing about the others. You wouldn't track a patient's health from their arm alone.

Most departments have systems and measures cobbled together over the years, and they tend to give a distorted or partial reading. Work and systems sit in silos, data takes time to surface and report, and there's little visibility into what has been done and what is underway right now.

Building work visibility is especially hard in legal, where operational structure can feel like an affront to professional expertise. That instinct isn't unique to lawyers. Surgery is at least as complex, and the stakes are higher, and it met the same resistance when it hit its own crisis of overload. It answered by building systems instead of asking surgeons to endure more. When Dr. Atul Gawande and the WHO introduced a 19-item surgical checklist, surgeons were skeptical at first, and many came around once they saw the results. Across eight hospitals in eight countries, complications fell from 11% to 7% and deaths from 1.5% to 0.8%. The checklist protected the surgeon's expertise instead of replacing it, taking the burden of routine recall off their plate so their attention could go to the surgery itself.

The hardest gap to close is the one with the outside. External signals are usually missing altogether. Even when a department does build feedback loops beyond its own walls, the team tends to reinforce its own confirmation bias, comparing notes internally and hearing mostly agreement. When people feel under siege, closing ranks around a shared story is a natural response. The second-order effect is that internal clients read that solidarity as the department being difficult or defensive.

What keeps the picture honest is unfiltered feedback from outside, and data that brings friction points to the surface. And those friction points are byproducts of the system, things you can fix, not failures of individuals.

Once you've captured useful, complete data, the hardest and most important work is making sense of it, because it can mean throwing out old mental models to make room for new ones.

Section 8 — Practice Is Not Delivery

"When the facts change, I change my mind. What do you do, sir?"

— Misattributed to John Maynard Keynes

Changing your mind sounds easy. Inside an in-house legal department, it rarely is. The trouble starts with a misread of where one kind of work ends and another begins, the line between the Practice Layer and the Operational Layer.

The Practice Layer is legal expertise applied. It's the demanding, bespoke work lawyers went to law school for, reading a regulatory shift, drafting a hard indemnity clause, negotiating an M&A deal, sizing up enterprise risk.

The Operational Layer is the delivery system around that work. It covers how a request comes in, gets triaged, routed, and tracked, and how the finished advice goes back out. The advice itself stays in the Practice Layer.

Back in 1983, the technology writer James Martin popularized CRUD, short for Create, Read, Update, and Delete, the four basic operations of stored data. A database stays accurate because it runs all four without playing favorites. It creates a record, reads it, updates it when the facts change, and deletes it when it goes stale. That's how software holds an accurate model of reality at any volume.

People don't run all four the same way. Create and Read come easily. We take in new information and pull up what we already know without much resistance. Update and Delete are where we balk, and the resistance has names. Updating means changing a belief when the facts change, and confirmation bias pulls the other way, toward whatever backs up the model we already hold. Deleting means letting go of something that has outlived its use, and loss aversion and the sunk-cost reflex keep us holding on well past the point where it helps. That's why changing your mind, which sounds simple, is so hard in practice. A database updates its model without a fight. We defend ours.

In the Practice Layer, rigidity is the point. The standard is zero defects. The method is to look backward, to precedent, to protect the future. Here, Dr. Larry Richard's findings on lawyers' high skepticism and autonomy are a real strength. You want a practitioner who is deeply skeptical of new information and slow to abandon a proven legal standard.

In the Operational Layer, rigidity is fatal. What it needs is to sense and respond continuously, which is the OODA loop. The method here is to look forward, to the vital signs, and adjust to volume and velocity. Here, that same skepticism toward new data, and that hard-won autonomy, will wreck the operation.

The structural failure of the modern legal department is that it runs the two layers together. Nothing in a lawyer's training tells them apart, so the approach that works in the Practice Layer, careful, precedent-bound, slow to change, gets applied to the Operational Layer by default. An intake form gets handled with the reverence owed to a court filing. The law and the mechanics of delivery never get separated, the orientation filter jams, and the friction lands on the individual lawyer.

From here the mechanics get easier. When the picture is accurate and shared, the right move is usually sitting there inside it, and a team that can see its own system doesn't have to reason its way to most decisions from scratch. What doesn't get easier is the discomfort, and it changes shape. It moves from seeing the system to acting on what you see before you can be sure it's right.

Section 9 — Build Through Small Experiments

The last two stages of the loop ask you to act before you can be sure. Deciding is forming a hypothesis about what will help. Acting is running it to find out. Both mean committing to a move when you can't know how it turns out, which cuts against the instincts the practice side of the work rewards.

So the rule for this part of the loop is simple. Treat it as a learning cycle, not a pass/fail gate. You clear or fail a gate, and a verdict comes with it. You run a cycle to learn, and a hypothesis that turns out wrong has still earned its keep by showing you something the picture couldn't. Put the rigor on the process, not on the people running it. Held that way, being wrong stops being a verdict and becomes information, and that is what makes an experimental approach possible at all for a profession trained on certainty.

Lawyers already form hypotheses. The skill is familiar. They do it in commercial negotiations, in litigation prep, in dealings with regulators, and more, each one a search for the right answer, the position that will hold up. What makes the Decide stage uncomfortable is the shift from the one right answer to the best answer you can reach right now, with the data you have. You can't know everything, and you can't predict how the parts will interact until they're running, so the answer stays open. Living with an open answer is uncomfortable, and structure is what makes it workable. Keep each test small enough that being wrong is cheap.

Acting is running that hypothesis as a small experiment. The experiments are operational. They change how the work moves, intake, triage, routing, handoffs, approvals, and they leave the legal judgment alone. That is what keeps practice consistency separate from operational flex. The substantive law stays as rigid as it should be, and only the delivery around it bends. Nothing a lawyer is trained to get right is on the table.

The experiments are designed and run by the people closest to the work, the ones who have hit the same operational problem across many cycles and worked it each time. That experience is the reason to put the design in their hands. They know which fixes hold and which come apart on the second pass, and they can adjust an experiment while it's running, which someone sitting above the work can't. This is a bottom-up build. The operating system gets assembled from a run of these small experiments, each one refined by the people running it, instead of drawn up whole and installed from the top.

That gives the leader's job a specific shape. It is to build the conditions the experiments need. The first is a safe zone, where a failed experiment counts as information and not as a verdict on the person who ran it. That safety lives in what a leader does the first time an experiment fails, far more than in anything announced ahead of time, and it holds only if it holds every time. One punished failure teaches the whole team that the safe zone was theater. The second condition is pace. You start small, let the early experiments build the muscle, and scale only as fast as the people and the process can absorb. Push harder than that and you've rebuilt the management pressure the whole system was meant to escape.

Where you point them is the last piece, and the order is by pain. Start with the most painful process, the one your people would name first if you asked them what wastes their time, and work down from there. That worst one is usually the biggest and most tangled, so you take it a slice at a time, a small experiment whose contained early win builds the muscle for the harder ground. This is where the diagnostic from the first paper pays off. The question of what makes people unproductive was always the map for where to start.

What you are building is a capacity. A department that observes its own operation, reads it honestly, acts on what it sees, and learns from what happens, again and again, without waiting for the next crisis to force the question. It takes two changes. The first is a way of seeing, separating the practice from its delivery, and you can reason your way to it. The second touches identity, where reasoning doesn't reach, so it has to be made safe instead. Neither one is easy. You will revert, reach for certainty under pressure, and run the loop badly before you run it well, and that is part of it, because the room you give a failed experiment is the same room the change itself needs. Learning to work this way is the very thing it describes. A cycle, run and corrected, never finished.

Nothing here freezes into a final state. The conditions will keep moving, the design will keep moving with them, and the setup never settles. What lasts is the capacity to keep adjusting. As that capacity sets in, the loops run faster and reach further, the natural result of a department that can see its own system and act on what it sees.

The suffering this Series has described from its first page was never the fault of the people inside it. The doom loop ran on them, on their endurance, their nights and weekends, their willingness to absorb what the system would not. A loop that senses and responds runs on something else, the system's own capacity to see and adjust. The work is still demanding, and there's no promise here of a lighter workload. What changes is the difference between a department that survives on its people and one that lets them practice. The conditions made the suffering. Different conditions can end it.

This is the fourth paper in the mot-r Foundation Series. The first three papers, The Quiet Crisis: Why Your Legal Team Is Struggling and What the Evidence Says You Can Do About It, Seeing the System: Why Your Legal Department's Problems Are Connected and What to Do About It, and The Wrong Verdict: Why Legal Gets a Reputation It Doesn't Deserve, are available at mot-r.com.

This paper was researched and written by Mike Tobias with drafting and editorial assistance from Claude, an AI assistant developed by Anthropic. All research, source evaluation, analytical judgments, and editorial decisions are the author's.

About mot-r

This whitepaper is part of a series produced by mot-r, a Customer-Aligned Enterprise Legal Management platform. The research in this series is independent, but it is not unrelated to what we build. The evidence-based frameworks, the organizational health principles, and the operational thinking described in these papers informed the original design of mot-r and continue to shape every subsequent improvement. We built the platform around this thinking because we believe it is right, and these papers exist so you can evaluate the thinking on its own merits, whether or not you ever use our product. To explore the ideas in this series further, visit the mot-r website at www.mot-r.com/resources.

Sources and Methodology Notes

The following sources are cited or relied upon in this paper. We have noted the methodology and any limitations for each, consistent with our evidence-based management commitment. Sources introduced at greater length in Papers I, II, and III are noted briefly here.

Primary Research

Haynes, A. B., Weiser, T. G., Berry, W. R., et al. “A Surgical Safety Checklist to Reduce Morbidity and Mortality in a Global Population.” New England Journal of Medicine 360, no. 5 (2009): 491–499. A prospective study across eight hospitals in eight countries, led by Atul Gawande’s group with the World Health Organization, measuring outcomes before and after introduction of a 19-item surgical checklist. The rate of any complication fell from 11.0% to 7.0% and in-hospital death from 1.5% to 0.8%. The design is a before-and-after observation without a randomized control group, so secular trends and a Hawthorne effect cannot be fully excluded; the consistency of the effect across very different sites is its main strength. Used here as an analogy for systems that protect expert judgment rather than asking professionals to endure more.

Amabile, Teresa, and Steven Kramer. The Progress Principle. Harvard Business Review Press, 2011. Analysis of roughly 12,000 daily diary entries from 238 employees across 26 project teams in seven companies. The central finding is that of all the events that lift inner work life, the most powerful is making progress in meaningful work. The design is a correlational diary study built on self-report, so it establishes association rather than cause. Used here for the link between visible progress and engagement.

Industry and Commercial Data (commercial interest noted)

ActivTrak. 2026 State of the Workplace: The Rise of Amplified Work. ActivTrak Productivity Lab, 2026. Aggregated behavioral data from more than 443 million hours of monitored work activity across 1,111 organizations and 163,638 employees, 2023 through 2025. Cited for the finding that the average uninterrupted work session fell to 13 minutes and 7 seconds in 2025. ActivTrak is a workforce-monitoring and productivity-analytics vendor, and the report is self-published from its own platform data; the sample is limited to organizations that run its software, and “focus session” is the vendor’s own measure. Treated as directional rather than authoritative, and not offered as a clean comparison with interruption-recovery research, which measures a different thing.

Axiom. “View from the Inside” survey series, conducted by Wakefield Research. A survey of in-house counsel across company sizes and industries. Cited for the finding that more than half of in-house lawyers are actively searching for or open to a new position. Axiom is an alternative legal services provider with a commercial interest in the space; the methodology is independently administered. Documented at greater length in Papers I and II.

Scholarly and Professional References

Drucker, Peter F. “Knowledge-Worker Productivity: The Biggest Challenge.” California Management Review 41, no. 2 (1999). Source of the principle that the first question in knowledge work is what the task is, and that removing what does not serve the task is a precondition for productivity.

Drucker, Peter F. Management: Tasks, Responsibilities, Practices. Harper & Row, 1973. Source of the epigraph that the purpose of an organization is to enable ordinary people to do extraordinary things. The line is widely attributed to this work.

Drucker, Peter F. Management Challenges for the 21st Century. HarperBusiness, 1999. Source of the observation that the task is to lead people rather than manage them, and that knowledge workers own their means of production.

Boyd, John R. “The Essence of Winning and Losing.” Briefing, 1996. The OODA loop as Boyd actually drew it, with Orient as the dominant element and implicit guidance and feedback running throughout. The two figures reproduced in this paper, the simplified four-step cycle and Boyd’s full sketch, are taken from Chet Richards, “Boyd’s OODA Loop” (2020), which sets the popular depiction against Boyd’s own. Boyd published little under his own name; his work survives mainly through briefings and the accounts of colleagues, so the record rests on secondary sources.

Taleb, Nassim Nicholas. The Black Swan (Random House, 2007) and Antifragile (Random House, 2012). Source of the fragile, robust, and antifragile distinction and the argument that suppressing visible variance relocates risk into the tail. Applied here to the difference between a static operating model and an adaptive one.

Little, John D. C. “A Proof for the Queuing Formula: L = λW.” Operations Research 9, no. 3 (1961): 383–387. Establishes that in a stable system the average work in process equals the arrival rate multiplied by the average time in the system. Used here to tie work-in-process, intake rate, and cycle time into one relationship.

McKinsey & Company. The War for Talent. The phrase comes from a 1997 McKinsey study and was developed in Ed Michaels, Helen Handfield-Jones, and Beth Axelrod, The War for Talent (Harvard Business School Press, 2001). Described in this paper as the model that elevated individual stars over systems, and presented critically rather than as endorsed practice.

Pfeffer, Jeffrey. “Fighting the War for Talent Is Hazardous to Your Organization’s Health.” Organizational Dynamics 29, no. 4 (2001): 248–259. Argues that a talent-war mindset breeds arrogance, zero-sum information hoarding, and unaccountable stars. Published the year Enron collapsed; Enron, whose chief executive Jeffrey Skilling came from McKinsey, is the usual illustration. Used here for the parallel to the “rockstar” attorney.

Pfeffer, Jeffrey, and Robert I. Sutton. Hard Facts, Dangerous Half-Truths, and Total Nonsense (Harvard Business School Press, 2006); and “Trust the Evidence, Not Your Instincts,” New York Times, September 3, 2011. Source of the view of leadership as building the conditions for others to succeed, and of the evidence-based commitment that runs through the Series.

Hope, Jeremy, and Robin Fraser. Beyond Budgeting: How Managers Can Break Free from the Annual Performance Trap. Harvard Business School Press, 2003. Names the dynamic in which managing to a fixed number produces behavior that hits the number at the cost of the organization’s health. Used here for the incentive problem built into lagging indicators.

Hallowell, Edward M. “Overloaded Circuits: Why Smart People Underperform.” Harvard Business Review, January 2005. Describes attention deficit trait, a focus deficit produced by the environment that resolves when conditions change. Used here for the claim that the attention environment is something an organization can design.

Mark, Gloria. University of California, Irvine. Research on attention and interruption, including “The Cost of Interrupted Work: More Speed and Stress” (2008). Source of the finding that it takes about 23 minutes, reported as 23 minutes and 15 seconds, to resume a task after an interruption. Paired in this paper with a separate, vendor-measured session length; the two are directional context rather than a single dataset.

Batalden, Paul, and the Institute for Healthcare Improvement. “Every system is perfectly designed to get the results it gets,” adapted by Batalden from Arthur Jones and popularized through the IHI; and the IHI Framework for Improving Joy in Work (Institute for Healthcare Improvement, 2017), source of the practice of identifying what matters to people, removing impediments through co-creation, and iterating. Documented at greater length in Papers I and II.

Richard, Larry. Lawyer personality research, drawn from validated psychometric profiling of more than 25,000 lawyers over three decades. Source of the finding that lawyers score below the general-population norm on resilience and high on skepticism and autonomy. Richard is a consultant to the legal profession. Documented at greater length in Papers I and III.

Martin, James. Credited with popularizing the CRUD acronym, for Create, Read, Update, and Delete, in his writing on database systems in the early 1980s. Used here as an analogy for the discipline of keeping an operational picture accurate by running all four operations rather than only the first two.

Lewin, Kurt. The unfreeze, change, refreeze model of change, conventionally credited to Lewin. Cummings, Bridgman, and Brown (“Unfreezing change as three steps: Rethinking Kurt Lewin’s legacy for change management,” Human Relations 69, no. 1, 2016) show the three-step version was assembled by later writers rather than stated by Lewin in that form. Referenced here as the model this paper argues against, since no refreeze is available under continuous change.

Duke, Annie. Quit: The Power of Knowing When to Walk Away. Portfolio, 2022. Argues that grit is overvalued and quitting stigmatized, which traps people and organizations in losing positions. Used here for why organizations resist stopping work that no longer serves the

task.

Teller, Astro. “The Unexpected Benefit of Celebrating Failure.” TED, 2016. The head of X, Alphabet’s moonshot factory, describes a culture in which teams that end unpromising projects are rewarded. This is a first-person account of an internal practice rather than independent research, used to show that termination can be designed as a rewarded behavior.

Cognitive biases. The resistance to revising a belief when the facts change (confirmation bias) and to abandoning a sunk investment (loss aversion and the sunk-cost effect) draw on the established behavioral-science literature, including Kahneman and Tversky’s work on loss aversion. Used here to explain why the update and delete operations are the ones people resist.

This paper was prepared with a commitment to source transparency. Where data is directional rather than definitive, we have said so. Where we have applied a framework or concept beyond its original domain, we have identified the inference as ours. We encourage readers to examine the underlying sources and draw their own conclusions.