No Plan Builds What Every Plan Needs

The capability that converts labor into throughput is thinning across physical industries simultaneously; it is already priced into deals as if it is not.

Names and identifying details have been changed.

Every capital plan in the physical economy prices a layer it never builds. The construction pro forma assumes a superintendent who can sequence the trades before they collide; the logistics underwriting assumes a dock supervisor who absorbs the variance the model failed to predict; the manufacturing model assumes a floor supervisor who has run the line long enough to see the break before it arrives. That layer converts labor into throughput. Every plan treats it as a fixed input: present at acquisition, available on demand, priced at zero because it was never a line item.

It is not a fixed input. It is a developed capability, built over years inside the operation, and it is thinning across every physical industry at the same time. You are not underwriting a future risk. You are buying assets that already carry the problem, priced as if they do not.

How the layer degrades will not show up in a staffing model. It shows up in a single shift, on a single phone call, long before it reaches a survey.

The Call and the Aftermath

A supervisor needed labor for the outbound dock, and he had the entire shift to arrange it. He did not.

We were short-staffed, and the convention when that happened was to pull drivers returning from city routes to work the dock for a few hours before going home. It bridged the gap, held down the overtime, and moved the freight with minimal intervention. It worked, and the supervisor knew it worked. He knew at one in the afternoon what the night would require, he knew he could open a seniority bid and have the labor scheduled long before the shift demanded it, and he knew the process because everyone knew it. Instead he spent the afternoon on his phone, avoiding the conversation he did not want to have with senior drivers who did not want the assignment.

By the time the driver walked in from the yard, he was hearing for the first time, at the window, that his night was not over. He had already called home to say he would make dinner. He meant it.

The driver refused. The assignment was enforceable; the supervisor had the authority and the policy supported the instruction. But the supervisor had produced the worst possible conditions for using that authority: no notice, no conversation, no dignity in the ask. I was on the phone with a supervisor who had manufactured the problem through a full shift of avoidance and a driver who was not only ambushed but disrespected, because he knew, as everyone knew, that a better way had been available all day.

I supported the instruction. I worked through my leadership team and directed the supervisor to notify the driver that he must comply or be stood down pending investigation.

I did this knowing the driver was right to be angry, knowing the investigation would end in a written warning and back pay, and knowing a final warning for the supervisor was waiting once I was back in the office. I did it because the alternative was worse. If a driver could refuse a supervisor and face no consequence because the supervisor had mishandled the ask, then every subsequent instruction from that supervisor, from me, from anyone in a coordination role, carried less weight. The operation runs on the credibility of that layer. When the credibility is in question, the work does not get sequenced, the dock backs up, the trailers do not close, and the freight does not move.

The supervisor took the badge and keys. The driver returned forty-eight hours later to his warning and his back pay. The supervisor went on a final written notice that same week; one foot out the door.

In the near term I was the villain. The station read it as protecting a supervisor who behaved badly and punishing a driver who reacted reasonably, and it took months of floor presence and individual conversations before that reading softened into something closer to the truth. The lesson was never about the supervisor or the driver. The driver absorbed the consequence. The supervisor absorbed the consequence. I absorbed the cost of resolving it correctly. The institution that built the conditions absorbed nothing, changed nothing, and will run the same sequence in the next operation it touches.

The field leader who acts in the enterprise’s interest takes the damage. The field leader who defers, deflects, or disappears absorbs nothing at all.

You learn not to try.

The Layer That Learned to Do Less

That call was not exceptional. It was the operating condition of field leadership in physical industries, made visible by one incident.

The dock supervisor, the operations manager, the terminal GM sit at the point where the enterprise’s legal exposure is greatest and its institutional support is thinnest. A complaint filed against a driver or a dock worker triggers a process with real standing: HR involvement, documented procedure, legal review, an infrastructure built to protect workers against management overreach, as it should be. A complaint filed against the manager triggers a faster, less formal process resolved disproportionately in favor of the complainant, because the manager is more expendable and less legally threatening than a sustained claim.

So the field leader who acts, who documents and confronts and disciplines and holds the line, is routinely more exposed than the one who does not. No policy states this. The corrective actions and the investigations that arrive whenever friction is created teach it more effectively than any directive could.

The capable field leader adapts: move slowly, act through intermediaries, document everything before any conversation, witness every interaction. The less capable one discovers that doing nothing produces no consequence, or at least more longevity. The institution reads both as equivalent occupants of the same role. The headcount holds. The capability degrades. I have sat across from supervisors who were once exceptional at this work and were now merely adequate, and the institution between us had never registered the difference.

This is not a bad hire or a training gap; it is a structural production, and the institution is not unaware of it. It has decided that changing the architecture costs more than living with what the architecture produces. That is a choice, and the consequence accumulates in every physical operation at once, in a form that appears in none of the reports the institution currently runs.

The Danger Now in the Data

For years the degradation stayed invisible because the headcount held. It is no longer invisible.

In construction, the AGC’s 2024 workforce survey found that 83 percent of firms seeking superintendents cannot fill the role. Engineering News-Record put roughly 34,000 superintendent positions unfilled nationally in May 2024 and classified the superintendent shortage as a deeper operational threat than the craft labor shortage beneath it. In October 2023 the industry launched its first standardized superintendent certification program, because no formal development pathway for the role had ever existed.

Logistics is the same shape at a later stage. Forty-three percent of transit and logistics workers are over 55, and 61 percent of operators report they cannot adequately staff transportation operations, the single most acute shortage point in the supply chain. The constraint is not the wage; it is that the people who can run these operations built that capability over years, and the next tier has not been in the role long enough to be ready.

Manufacturing and utilities carry the same signal. First-line production supervisors have ranked among the five most-demanded roles for five consecutive years against a decade requirement to reach the seat, and 76 percent of energy and utility employers report a skills gap inside their existing workforce as veteran lineworkers retire into a grid-modernization buildout that requires experienced field coordination to execute safely. Figures and sources in full below.

These are not four parallel problems. They are one problem: the field coordination layer thinning simultaneously in every sector that runs on sequenced physical work. The surveys have been public for years, and the institutions operating in these sectors have them.

The Limits of Capital

The standard response to a labor shortage is a wage increase, and the wage increase is correct as far as it goes. It moves people at the current moment. It does not increase the supply of people who have ten years of field coordination behind them, because that experience is gated by time in role: the pattern recognition that develops only through sustained exposure to sequencing failures, trade conflicts, and schedule compression across many cycles. Capital cannot accelerate time in role.

The deeper failure is that capital is being deployed to avoid answering for what was never built. Wages solve for now. They do not answer the question a capable person weighs before committing to a field coordination role: is there an organization here that will develop me, or one that will only deploy me.

The administrative layer of every physical industry has a legible answer. Rotational programs, management trainee tracks, credentialed pathways, career ladders with visible rungs. The degreed construction manager has a career architecture; the logistics analyst has a career architecture; the first-line supervisor on the dock or the job site has a title, a paycheck, and ten years of accumulated field judgment that the organization treats as a fixture rather than a product, and does not notice is gone until the sequencing fails.

That asymmetry was built on purpose on one side and left to chance on the other. The administrative layer was designed, with investment and institutional attention. The field coordination layer was allowed to form organically and was never re-examined once the organic pipeline began to thin. A deliberate choice to build one thing, sitting next to the absence of a choice to build the other, is not ambiguity. It is allocation.

The industry’s own account of this is that people do not want the work, and that account has the diagnosis backward. Supervisor attrition runs above 52 percent annually, with career-development failure cited as the primary driver by a margin of two to one over compensation. The people in these roles are not leaving because the pay is short; they are leaving because the role has no trajectory they can see and no one accountable for building one. Calling that a shortage is also a choice.

The Tools That Failed

Meanwhile the administrative layer grows, because bureaucracy has no natural end. Engineering teams build scheduling tools. Analytics teams produce arrival curves and labor-demand models. Process teams document workflows and run variance reports. The enterprise invests in instruments that make the field layer visible and manageable from a distance, and every one of those investments is a decision to fund the instrument rather than the capability it measures.

The field coordination layer runs at ratios that never kept pace: one leader for several thousand units processed, absorbing whatever variance the tools failed to predict, which is most of it. That ratio did not emerge from a labor market. It was set, and then left in place while the administrative headcount expanded around it. The person absorbing the ratio never set it; the person who set it has never absorbed it.

The tools deployed to close the gap cannot close it. A scheduling model that produces optimal staffing times cannot account for a billing office running at half throughput; a dock optimization system cannot account for a supervisor managing late arrivals, call-offs, and a damaged pallet in bay six at the same moment. The models are analytically correct. They were built without the coordination layer inside them, because putting it inside them would require admitting that it is not a fixed input. Capital can raise wages. It cannot manufacture sequence.

The Hidden Arithmetic

The dependencies are sequential. A construction project requires a superintendent; a utility connection requires a line crew supervisor; a logistics operation delivers the materials; a manufacturing facility produces the components. A field leadership gap at any one point extends every timeline downstream, and the losses compound rather than add.

A data center project in Texas draws from the same regional field leadership pool as grid expansion, housing, industrial buildout, and municipal work. Texas added 42,300 construction positions in 2024, more than any state, and already holds the highest concentration of first-line construction supervisors in the country; that same pool is now being asked to be everywhere at once. A project underwritten at eighteen months becomes a twenty-eight month project, not because any single constraint was catastrophic, but because every stage lost weeks at the field leadership layer and none of those weeks came back.

The arithmetic is unforgiving. A ten-month slip on a project financed at ten percent adds a holding cost the model never priced and pushes stabilized cash flow nearly a year to the right; on an asset underwritten to a mid-twenties IRR, that single delay can remove several hundred basis points of return before any other assumption moves. The delay never appeared in the pro forma because the layer that produced it was never in the pro forma. The layer is not a rounding error against the thesis. In a levered hold it is frequently the difference between the underwritten return and the realized one.

Then the loss gets attributed to labor market conditions, and the attribution is the mechanism by which no one is accountable. The labor market is not an actor and makes no decisions. The institutions that underdeveloped the coordination layer over the same decade they expanded the administrative overlay made decisions, and the schedule variance and cost overruns that every post-mortem in the sector files as structural and unavoidable are the result. They are structural. They are not unavoidable.

The Diagnostic

There is no dashboard for this layer, but there is a diagnostic, and it does not require access to the reports.

The break room tells you more than the operations summary. It is the liminal space in any physical operation, owned clearly by neither management nor the workforce, maintained by whoever takes responsibility for it between shifts. A functioning field leadership layer does not keep that space clean by standing over it; it produces a culture in which the people who use the space keep it themselves, because they take pride in the operation they belong to. The same reading is in the uniform, in whether it is worn like a marker of membership or like a thing issued to a number. Pride is the visible surface of organizational health, and its absence is one of the earliest legible signals of risk. When the layer thins, the pride goes first and the drift follows: the break room slides, the uniform slackens, and both are readable before any document is opened, in the same operations the dashboard still reports as within range.

The loudest manager on the floor is usually the least effective. Functioning field leadership produces quiet operations: work sequenced before the shift opens, people who know their task, problems solved where they originate. A manager doing the work himself is, once, a situation; every day, a planning failure. The intervention that looks like leadership is more often evidence that the system was never built to run without it.

Conversation content is the most telling data point. Where field leadership depth exists, people describe how the hub affects the inbound, how the inbound affects the city run, how the city run affects the outbound, and where a delay at one end will surface at the other; they carry the whole sequence and locate pressure inside it without being asked. Where the depth is gone, the conversation stays in the lane. The dock supervisor knows the dock. What the dock produces downstream has not been asked, and is not being answered.

The Team That Followed Incentives

One of the operations I took over had an inbound team that had given up. Not dramatically. Shifts ran, freight moved, more or less. What had worn them down was not the work; it was a run of managers who served themselves and their own political interests rather than the operation, and who dressed that up in the language of the enterprise. “Operational excellence” and “independence” arrived again and again as grand narrative, and often enough the narrative was cover: a record built to carry somewhere else, a cost pulled out of the floor and reported upward as a win. It was extraction framed as grand narrative. The team had been through enough cycles of new leadership, new systems, and new promises, each one the next thing that was going to change their future if only they believed in it and executed it, to reach a collective conclusion: nothing changes, and when the next thing fails, the people who tried carry the blame. They had tested that theory enough times to trust it. They punched in, punched out, and held variance just inside the range that did not trigger review.

The goal was survival, not success.

It took nine months of sustained individual work to break the adaptation they had built over years of institutional disappointment: floor presence late into shifts, individual conversations, adjustments small enough to be believed, and then the same again. The supervisors on that team were not incapable. They were rational. They had learned from direct experience that capability and ownership were liabilities in their environment, and the system had efficiently hardened that lesson into doctrine.

That team is not exceptional. It is what the architecture produces when it runs long enough. The headcount holds, the org chart is unchanged, and the operation keeps functioning at exactly the level required to avoid an intervention. The capability has been replaced by a performance of adequacy that is indistinguishable from adequacy in the data.

The Decision

A three-year exit horizon does not insulate against any of this. The gap does not take ten years to appear; it takes six to eighteen months, and it presents inside the hold period as schedule variance and cost overruns attributed to labor market conditions, material delays, and subcontractor performance. Those attributions are not wrong. They are incomplete, because the field coordination layer was never measured before the asset was priced and never developed after it was acquired. The operation was underwritten on the assumption that the layer existed. It did, at acquisition. By the time the schedule slips, it has been running on attrition for two years with no replacement strategy, because no one was accountable for building one and everyone assumed that paying more would be enough.

You are not underwriting a future risk. You are buying assets that already carry the problem, priced as if they do not. At the time you priced it, you could have checked. The diagnostic is one question: who inside this operation is being developed right now, not hired, developed, to replace the field coordination layer when it leaves.

If the answer is no one, the plan is assuming something nobody built, and nobody is planning to build it.

That is not a permanent condition; it is a choice, and the people who can rebuild the layer are already inside these operations. They showed up this morning. Some of them have stopped trying and some of them have not, and the architecture that produced the avoidance and the performance of adequacy was built by people making choices, which means it can be built differently by people making different ones. The physical economy still has to be constructed. The layer that makes construction possible still has to be developed rather than assumed. What remains is the decision.

Erik Ibe | Managing Principal, Induric

The Data

AGC 2024 Workforce Survey: 83 percent of firms seeking superintendents report difficulty filling the role; 81 percent for project managers and supervisors.

Engineering News-Record, May 2024: approximately 34,000 unfilled construction superintendent positions nationally; superintendent shortage classified as a deeper operational threat than the craft labor shortage.

Deloitte and The Manufacturing Institute, 2024: U.S. manufacturing projected to need 3.8 million additional workers by 2033, with up to 1.9 million potentially unfilled; first-line supervisors among the five most-demanded production roles for five consecutive years.

Bureau of Labor Statistics, Employment Projections 2024-2034: Construction Managers (SOC 11-9021) projected at 9 percent growth; First-Line Supervisors of Construction Trades (SOC 47-1011) projected at 5 to 7.6 percent growth.

U.S. federal appropriations, FY2024: $184.35 billion for higher education; $285 million for registered apprenticeships (Jobs for the Future / RAND, 2024).

NCCER, October 2023: Construction Superintendent Certification Program launched, the industry’s first standardized field leadership development pathway.

Talent Traction, April 2026: 43 percent of transit and logistics workers currently over 55; mid-level logistics managers being promoted before ready.

Descartes Systems Group, 2024: in a survey of 1,000 supply chain and logistics decision-makers, transportation operations was the area suffering most from workforce shortages at 61 percent, ahead of warehouse operations at 56 percent.

Manpower Group / CEWD, 2024: 76 percent of energy and utility employers experiencing a talent and skills gap within their existing workforce.

The Resource Company, 2025: project supervisor annual turnover rate 52.4 percent, career development failure cited as the primary driver by a margin of two to one over compensation.

Texas Workforce Commission / BLS OEWS 2023: Texas holds the highest total employment of first-line construction supervisors nationally; state added 42,300 construction positions in 2024.

Erik Ibe is Managing Principal of Induric, a Dallas-based industrial operating and acquisition platform.