Chapter 2 — The Construction Industry: Players, Project Types, and How the Business Works

The Hook: Eleven Jobs and a Cash Position

It is 6:52 on a Monday morning in November of Year 1, and I have dragged Dani Okonkwo out of the Northgate job trailer and into the Kestrel main office for the operations meeting.

Dani has been a field engineer for eight months. They have been living in submittals, layout, and concrete pours. They brought a notebook and, I am fairly sure, a set of questions about anchor bolts.

At 7:00 sharp, Nadia Haddad — Vice President of Operations, and the person who decides whether Kestrel bids a job, accelerates a job, or walks away from a job — puts eleven project names on the screen. These are the jobs over five million dollars. Everything smaller belongs to the small-projects group, which meets on Thursdays and which Nadia loves dearly and does not think about on Mondays.

KESTREL CONSTRUCTION GROUP — MONDAY OPERATIONS REVIEW
Jobs over $5M                                        Week of Nov 6, Year 1
──────────────────────────────────────────────────────────────────────────
 JOB                              VALUE      %CMP   BILL   CASH   FADE
 Northgate Outpatient Pavilion  $47.5M       41%    41%     OK     —
 Sable Ridge Logistics Center   $31.0M       15%     9%    (6)     —
 Rivermont Elementary #12       $22.4M       68%    74%     +6   (410)
 Fair Oaks Water Plant Exp.     $16.2M       74%    72%     OK     —
 Halyard Point Office Repo.      $9.6M       30%    31%     OK     —
 ...six more.......................................................
──────────────────────────────────────────────────────────────────────────
 BONDING: $150M agg / $60M single-project     WORKING CAPITAL: $21.4M
 CASH: $6.8M           LINE: $15.0M ($4.0M drawn)      BACKLOG: 4.2 months

Nadia goes around the table. She does not ask a single question about how to build anything.

"Sable Ridge. You're fifteen percent complete and nine percent billed. That's six points of underbilling on a thirty-one-million-dollar job. That is roughly one point nine million dollars of our money sitting in somebody else's building. Fix the schedule of values or fix the pay app, but fix one of them by the twenty-fifth."

"Rivermont Twelve. Curtis, you're billed six points ahead of complete and your projected fade is four hundred ten thousand. Walk me through the fade."

Curtis Boone, who is charming and quick and has never in his life been rattled in a meeting, says the fade is a timing issue and he expects to recover it in the masonry buyout.

Nadia writes something down and does not look up. "Northgate. Ray. Retention steps from ten to five at fifty percent complete. When does that hit, and what does it free up?"

I tell her: February, Year 2, and it releases about two million three.

Owen Baptiste, our CFO, then spends nine minutes on a single slide about cash. Not profit — cash. Which jobs are generating it, which are consuming it, and what happens to the company's cash position in March when three jobs mobilize in the same six weeks.

The meeting ends at 8:04. Nobody said the word "concrete." Nobody said "steel." Somebody said "curtain wall," but only because a deposit on it was moving $1.4 million out the door in December.

In the parking lot, Dani stares at their notebook, which has four words in it.

"Ray. That was a finance meeting."

"Yes."

"I thought we built buildings."

"We do. Kestrel builds buildings the way an airline flies planes. The planes are the visible part. The business is fuel hedging, gate slots, and financing. If Nadia gets the buildings right and the money wrong, we are out of business in eighteen months with a beautiful portfolio."

Dani writes down a fifth word.

Here is what this chapter gives you: the business you have actually joined. Why margins are so thin that a single late submittal can erase a quarter of a job's profit. Why almost nobody self-performs almost anything. Why the money takes two to three months to travel from the owner to the person who installed the drywall. Why the same firms make the same mistakes every cycle. And why the answer to "why is this industry like this?" is almost never "because people are dumb" and almost always "because of who owns which risk, and what it costs them."

🏃 Fast Track: If you have worked in the industry for five years, skim 2.1 through 2.3, then read 2.4 (the margin arithmetic), 2.6 (the tier structure and the payment chain), and 2.7 (backlog, bonding, and why growth kills contractors) closely. Do the 📋 Try it in 2.7 — most experienced field people have never run the numbers on their own company's bid capacity.

🔬 Deep Dive: The delivery methods sketched here get a full chapter each in Chapter 3 and Chapter 4. The bond, lien, and insurance framework is Chapter 5. The payment mechanics are Chapter 32, and the company-level finance — WIP schedules, over/under billing, bonding capacity — is Chapter 34. If you are on the owner's side of the table, Appendix E is written for you.

Everyone and every company in this book — Kestrel, Northgate, Meridian, Rivermont Elementary #12, all of them — is a composite drawn from real projects, with the names and numbers changed. The behavior is real. The names are not.

2.1 How Big, How Fragmented, and Why That Explains Almost Everything

Construction is one of the largest single sectors of the United States economy. The value of construction put in place — the industry's term for the dollar value of work actually installed in a given period — runs well over a trillion dollars a year. I am deliberately not giving you a precise figure, because it moves every month and any number I print here will be wrong by the time you read it. The U.S. Census Bureau publishes the series monthly, broken down into residential, nonresidential, public, and private. Go look at the current number before any interview, any market analysis, and any conversation with a lender. It takes four minutes.

Two more magnitudes worth carrying around, both hedged the same way:

• Construction employs several percent of the American workforce. The Bureau of Labor Statistics tracks it, and the number swings hard with the cycle — more on that in 2.7.
• Construction is one of the most fragmented major industries in the country. There are hundreds of thousands of construction establishments in the U.S., and the overwhelming majority are small — most have fewer than ten employees. Even the very largest contractors hold only a low single-digit share of the national market. Engineering News-Record (ENR) publishes the standard annual rankings — the Top 400 Contractors, the Top 500 Design Firms — and the useful thing about those lists is not the order, it is the revenue at the bottom of them. The four-hundredth-largest contractor in the country is a real company with real jobs, which tells you how flat the distribution is.

Compare that to almost any other trillion-dollar industry. Automobiles: a handful of global manufacturers. Airlines: four carriers move most domestic passengers. Semiconductors: you can name the players on one hand. Construction: a hundred thousand players, and the biggest one cannot dictate anything to anybody.

The shape of the industry

Here is the structural picture. This is a description of firm types, not a statistical distribution — I am not going to invent percentages.

Why fragmentation is the root cause of so much else

Hold on to one sentence, because a surprising amount of this book is downstream of it:

Every construction project is a temporary organization, assembled out of firms that have often never worked together, that will disband the moment the job ends.

Think about what that does.

It kills the learning curve. A manufacturer builds the same product ten thousand times and gets better every thousand. You build one hospital, learn twenty expensive lessons, and then the team scatters. The lessons live in individual people's heads, not in the organization — which is why Chapter 40 treats lessons-learned as a deliverable and not a formality.

It makes technology adoption brutally slow. A tool that only pays off when every trade uses it has to be sold twenty separate times, to twenty companies with twenty different balance sheets and no shared incentive. That is the honest reason construction technology diffuses slowly — not because field people are Luddites. Most field people I know will adopt anything that makes their Tuesday easier. They will not adopt something that makes the general contractor's Tuesday easier at their own expense. Chapter 39 takes this apart.

It puts an information gap at every contract boundary. The owner knows why they are building. The architect knows the design intent. The GC knows the schedule. The sub knows the crew. The crew knows the actual condition of the wall. None of them share a database, an incentive, or a boss. Every interface is a place where information is lost and risk is quietly handed to whoever is least able to argue.

That last one is the whole job. Construction management, at bottom, is the practice of moving accurate information across contract boundaries fast enough to make decisions before they get expensive.

🔄 Check your understanding. In one sentence: why does a genuinely good productivity technology spread through construction more slowly than through manufacturing?

Answer

Because a manufacturer can decide once, for its whole production system, while a construction project's production system is assembled from dozens of independent firms — so the technology has to be sold, funded, and learned separately by each of them, and the firm that pays for it is often not the firm that captures the benefit.

2.2 The Four Sectors

The industry sorts itself into four sectors. They share vocabulary and almost nothing else. The contracts differ, the owners differ, the margins differ, the risks differ, and the career paths differ. People do move between them, but the move is harder than newcomers expect, because what you actually learned in one sector was a risk model, and the risk model does not travel.

Residential

Single-family homes, townhouses, condominiums, and apartments. Split it further:

• Single-family — custom homes (one owner, one house, high touch, high change) and production housing (a builder builds the same four floor plans ninety times a year and lives or dies on cycle time). Kestrel does not do this. Harbor Ridge, the 34-lot subdivision run by Tessa Bright Homes with superintendent Colton Reyes on a 92-calendar-day house cycle, is our residential window, and it gets a full chapter in Chapter 37.
• Multifamily — apartments and condos. Financially this behaves like commercial development, because the owner is usually a developer with a pro forma, a construction loan, and an exit. Physically, above a certain height it stops being residential construction and becomes commercial construction with bedrooms in it.

Residential is where most contractors are, where licensing barriers are lowest, and where the customer is often an individual spending the largest sum of their life. That last fact drives more of residential CM than any technical consideration.

Commercial and institutional

Offices, retail, hotels, schools, universities, healthcare, government buildings, laboratories, churches, courthouses. This is the biggest bucket by firm count in the commercial world and the one most CM graduates enter. Northgate lives here — a 132,000 square foot outpatient pavilion for Meridian Health System, delivered CM at Risk with a Guaranteed Maximum Price of $47,500,000 over 565 calendar days.

Healthcare deserves its own note, because it is the most demanding subset of this sector: infection-control requirements during construction, medical-gas certification, redundant power, imaging equipment with structural and shielding implications, and an owner who cannot close the building next door while you work. Northgate's tight north property line, adjacent to an active clinic that stays open, is not a plot device. It is the normal condition of healthcare construction, and it is why the job's site logistics get an entire chapter of attention later.

Industrial

Refineries, chemical plants, power generation, manufacturing facilities, data centers, semiconductor fabs, food processing. The building is usually a shed around a process; the process is the project. Industrial work is dominated by a smaller set of specialized contractors, is heavily engineered, is frequently cost-reimbursable, and often runs on shift work with enormous craft workforces. Schedules are frequently driven by an outage window or a production start date that a corporate board has already promised to investors, which makes the schedule pressure ferocious and the money enormous.

You will not see much industrial work in this book, but you should know it exists and that its economics are genuinely different. If someone tells you a job "ran two hundred million over," they are usually talking about industrial or heavy civil.

Heavy civil and infrastructure

Highways, bridges, tunnels, water and wastewater treatment, dams, ports, transit, transmission, and energy. Almost always public owners, almost always competitively bid, almost always unit price contracts — you bid a price per cubic yard, per linear foot, per ton, and get paid for the measured quantity actually installed. Cottonwood Creek Bridge Replacement — $18,700,000 for a state DOT, 210 working days, superintendent Del Ferraro, project engineer Ingrid Sørensen — is our window into it, in Chapter 38.

Heavy civil is capital-intensive: you own the excavators, the cranes, the batch plant. That changes everything about how you bid, because your equipment is a fixed cost that has to be fed.

The comparison

Where the book's projects sit

Note the pattern in that table. Four of the five projects the book uses are institutional or public. That is not an accident of my career; it is where a large share of construction management jobs are, and it is where the rules are most explicit — which makes it the best place to learn them.

2.3 Who Is Buying: Public Versus Private Owners

Before you learn anything else about an owner, learn whether they are spending their own money or the public's. Everything else follows.

How each one buys work

A private owner can hire whomever they want, on whatever terms they can negotiate. They can select a contractor based on a relationship, on a proposal, on a qualifications-based interview, on the strength of a preconstruction team, or on a coin flip. They can negotiate a GMP. They can bring the contractor in during design. Meridian Health System selected Kestrel for Northgate through a qualifications-based process and then negotiated the GMP — no sealed bids, no public opening, no legal obligation to take the lowest number.

A public owner generally cannot. Public procurement law — which varies by state, by agency, and by funding source, and which changes — is built around the idea that taxpayers are entitled to an open, documented, non-discriminatory process. In practice, on most publicly funded construction, that means:

• The project is publicly advertised.
• Bids are sealed and opened publicly at a stated time, often read aloud.
• Award goes to the lowest responsive, responsible bidder — "responsive" meaning the bid complied with the requirements, "responsible" meaning the bidder is qualified, licensed, bondable, and not debarred.
• Bid bonds, payment bonds, and performance bonds are required.
• Prevailing wage requirements often apply.
• Everything is a public record, including your bid.

Alternative delivery on public work — design-build, CM at Risk, progressive design-build — has become far more common than it was twenty years ago, but it is generally available only where a statute specifically authorizes it, and the authorizing statutes vary enormously by state and by agency type. Never assume. Read the procurement code for that jurisdiction, and read it again when you cross a state line.

What that does to the price

Two mechanisms drive public work's price in opposite directions, and you need both in your head.

Prevailing wage pushes labor cost up. On federally funded construction, the Davis-Bacon Act requires payment of locally prevailing wages and fringe benefits, determined by the Department of Labor by classification and county. Most states have their own analogous statutes — often called "little Davis-Bacon" acts — with their own thresholds, their own wage determinations, and their own certified-payroll requirements. Some states have none at all. Some have repealed theirs and some have reinstated them. This is a jurisdictional question with a jurisdictional answer, every single time.

Practically, on a job like Rivermont Elementary #12 or Willow Street, prevailing wage means: you pay published rates by classification, you file certified payroll reports on a schedule, you keep apprentice-to-journeyman ratios inside the allowed range, and you carry real administrative cost to do all of it. It also means your open-shop labor cost advantage largely evaporates, which is exactly why Kestrel is union on public work and open shop on most private work. We are not making a philosophical statement. We are responding to the price structure the law creates.

Low-bid competition pushes contract price down. When the only selection criterion that varies is price, and five qualified bidders are chasing one job, the winner is frequently the bidder who made the most optimistic assumption or the largest arithmetic error. There is a name for this in economics — the winner's curse — and construction lives in it. Chapter 15 is largely about not being that bidder.

⚖️ What the contract says. On most public projects you cannot file a mechanic's lien. You cannot foreclose on a courthouse. In place of lien rights, public work substitutes a payment bond — a surety's promise to pay subcontractors and suppliers who go unpaid. On federal projects the Miller Act requires payment and performance bonds above a contract threshold; nearly every state has a "Little Miller Act" for state and local work, with different thresholds and, critically, different notice and claim deadlines. Those deadlines are short, they are strict, and missing one can extinguish an otherwise valid claim. If you are a subcontractor or a supplier on public work, the two dates you must know are the deadline to give notice and the deadline to sue on the bond — and you must get them from that jurisdiction's statute, not from a book. Chapter 5 walks the framework; your attorney supplies the dates.

How they behave when things go wrong

This is the difference that will affect your Tuesday afternoon.

That last row is the one to tattoo somewhere. On a private negotiated job, the reason your reasonable request gets granted is often that the owner wants to work with you again. On a public hard-bid job, there is no "again" — the next project goes to whoever is low. Your entire leverage is the contract and the documentation. This is exactly why Curtis Boone's habit of managing by force of personality works fine for eleven months on Rivermont Elementary #12 and then stops working all at once.

🔄 Check your understanding. Name three ways a state DOT replacing a bridge will behave differently from a developer building a 96-unit apartment building — and which of the three will change how you spend your Tuesday afternoon.

Answer

Any three of: the DOT must advertise and award to the low responsive responsible bidder while the developer picks whoever they like; the DOT's contract terms are fixed and published while the developer's are negotiable; the DOT's change approvals run through a procedural chain and an appropriated budget while the developer can decide fast from a contingency they control; prevailing wage and payment/performance bonds are statutory on the DOT job and optional on the developer's. The Tuesday-afternoon difference is decision speed on changes — on the DOT job you must document, notice, and wait, and if you build first and paper it later you may never get paid.

2.4 The Players and Their Economics

Every party on a construction project is running a different business with a different revenue model, a different cost structure, and a different definition of a good day. Almost every conflict you will mediate for the rest of your career is a collision between two of these business models — not a collision between two bad people.

Learn what each party sells and how they get paid, and their behavior stops being mysterious.

Owners and developers

The owner is buying an asset, not a construction project. That distinction is everything.

An institutional owner — Meridian Health System, a school district, a university — is buying capacity. Meridian is not building Northgate because they want a building; they are building it because their existing clinics are full and their leased interim clinic space expires on October 1 of Year 2. The building is a means. The date is the point.

A developer is buying a spread. They build to a pro forma — a projection of what the finished asset will earn and what it will be worth — and their profit is the difference between what the asset is worth when stabilized and what it cost to create. Two numbers govern them:

• Cost of capital. Every day of construction is a day of interest on a construction loan and a day their equity is tied up earning nothing.
• Time to revenue. Every day the building is not open is a day of rent, or patient volume, or sales, that they will never get back. Not deferred — gone.

Read that second one again, because contractors chronically misunderstand it. A six-month delay does not push a developer's income six months later. It permanently deletes six months of income from the life of the deal. case-study-02.md in this chapter works the entire arithmetic on a real-shaped multifamily pro forma, including the part where liquidated damages recover only about a third of what the delay actually costs the owner. Read it before you next tell an owner that a two-week slip is "not a big deal."

Architects and engineers

Design professionals typically earn a fee expressed as a percentage of construction cost — commonly somewhere in the mid single digits to low teens depending on building type, scope of services, and complexity; healthcare and laboratories run high, warehouses run low. Sometimes it is a fixed fee derived the same way. The percentage matters less than three structural facts about their business:

Their fee is fixed early and their effort is not. The fee was negotiated against an assumed scope and an assumed level of documentation. Every RFI (Request for Information) you send, every submittal you route, every substitution you propose consumes hours the fee does not fund. Dale Whitcomb at Halvorsen + Pike is not slow on submittals because he does not care. He is slow because he has four other projects and a fee that assumed a normal RFI volume. When Kestrel manages the submittal log well, Dale is fast. When we dump forty submittals on him in one week, he is slow, and it is our fault. That is Chapter 25.

They are judged by a standard of care, not a warranty. A designer does not typically guarantee a perfect set of drawings. The legal standard is generally what a reasonably prudent professional in that discipline, in that locality, would have done. Drawings will have errors. Some of those errors become change orders; some become claims; some are absorbed. Understanding this keeps you from the fantasy that you can make an architect pay for every coordination gap, and from the opposite fantasy that you must absorb all of them.

Their liability is long and their insurance is specific. Professional liability coverage is different from the general liability that contractors carry, and design errors are not covered by a contractor's policy. This shapes how designers write, how they answer RFIs, and why some answers come back as questions.

General contractors and construction managers

Now the part that governs your career.

🧩 Productive struggle. Kestrel's CM fee on Northgate is 4.0 percent of the $45,120,000 subtotal — $1,804,800 on a $47,500,000 contract. Before you read another word: how much of that do you think Kestrel actually keeps as profit? Write down a number and one sentence of reasoning. Take three minutes. Almost everybody gets this wrong the first time, and the direction of the error tells you something about how you have been taught to think about business.

Here is the answer, worked.

The fee is not the profit. The fee has to carry Kestrel's share of the things that exist whether or not Northgate does: the estimating department that bid eleven jobs to win three, Owen Baptiste's accounting group, the IT systems, the corporate insurance, Bea Salgado's safety department, the lease on the office, the marketing that got us shortlisted, and the salaries of everybody who works on projects that never happen. Allocate that fairly across revenue and it eats over half the fee.

A net margin somewhere in the range of two to four percent of revenue is typical for a general contractor, and one to three is common. Industry financial benchmarking studies — the AGC and the Construction Financial Management Association publish them periodically — are where to get current figures, and they will confirm this order of magnitude. Specialty trade contractors typically do somewhat better as a percentage, because a larger share of their revenue is their own labor rather than pass-through subcontract value.

Two percent is a grocery-store margin. Now watch what that does.

💰 Money check: what one late submittal costs.

Northgate's projected net profit: $957,300 — call it $950,000.

The anchor-bolt and embed submittal sat in Kestrel's office for 11 days before it went to Caldwell Structural, and Caldwell took its full 14-day contractual review. Ironbridge Steel missed its mill rolling slot. The next opening was five weeks out. Steel erection start slipped from August 4 to August 27, Year 1 — 23 calendar days — and steel erection was on the critical path.

Northgate's total daily exposure to slipping substantial completion is $10,650 per calendar day: $5,150 of extended general conditions plus $5,500 of liquidated damages.

23 CD × $10,650/CD = $244,950

$244,950 ÷ $957,300 = 25.6%

One submittal that sat on a desk for eleven days consumed roughly a quarter of the entire projected profit on the largest job in the company.

And now the part that should keep you up at night. At Kestrel's 2.3 percent net margin, how much new revenue does the company have to win and build to replace $244,950 of lost profit?

$244,950 ÷ 0.023 = $10,650,000

To earn back what one late submittal cost, Kestrel has to go find, bid, win, and safely build a $10.65 million project. The arithmetic has a grim symmetry to it — $10,650 a day out the door, $10,650,000 of revenue to put it back.

This is theme 2 in its purest form: the schedule and the budget are the same conversation. Anyone on your job who talks about days without talking about dollars is doing half the job.

That single calculation is why this industry behaves the way it does. It explains the obsession with change orders. It explains why contractors document everything. It explains why a good project manager gets uncomfortable when a submittal log goes untended for a week. And it explains why a firm that is excellent at building and mediocre at contracts, schedule, and paperwork will lose money forever without ever knowing why.

💡 Aha moment. A general contractor's product is not a building. It is a promise about a building — a price and a date, sold before the thing exists, and backed by the contractor's balance sheet. Everything in the GC's business is an attempt to make that promise safely: the estimate is a priced forecast, the schedule is a modeled forecast, the contingency is a hedge, the bond is insurance on the promise, and the fee is what you get paid for carrying the difference between the promise and reality.

Specialty trade contractors

Here is a number that reorganizes most people's mental picture of a job site: on Northgate, roughly 80 percent of the $40,000,000 direct cost of work is subcontracted. Kestrel self-performs concrete, rough carpentry, and general trades. Everything else — steel, curtain wall, precast, roofing, mechanical, electrical, plumbing, fire protection, elevators, drywall, flooring, casework, paint, landscaping — belongs to somebody else's company.

Which means: the general contractor performs a minority of the work and carries a majority of the risk.

Specialty trade contractors are where the actual labor and the actual production risk live. Cardinal Mechanical, run by Sofia Marchetti, is a real business with a real payroll, real fabrication shops, and real cash-flow problems. Halcyon Electric under Devlin Achebe is the same. When either of them has a bad quarter, your schedule has a bad quarter, and there is nothing in your contract that will fix it fast.

🔍 Why this works: why almost everyone subcontracts.

The naive explanation is that GCs subcontract to shed risk. That is a benefit, but it is not the mechanism.

The mechanism is demand lumpiness. On Northgate, the structural steel erection crew is needed for roughly fourteen weeks. The curtain wall crew is needed for about twenty. The roofers are needed for six. If Kestrel employed all of those crews directly, we would need a portfolio of jobs whose steel phases, glazing phases, and roofing phases lined up perfectly and continuously — forever. They never will. We would spend a fortune carrying idle crews between phases, and we would still be short-handed at every peak.

A specialty contractor solves that by aggregating demand across many general contractors. Ironbridge Steel keeps its erection crews busy because it works for eleven GCs, not one. Its utilization curve is smooth because it is the sum of many lumpy curves. That is real economic value, and it is why the structure exists.

Now the cost. Every subcontract converts a management problem into a contract boundary, and every boundary is a place where information stops, incentives diverge, and scope falls into a gap. The industry traded coordination cost for utilization efficiency. That trade is the reason Chapter 16 exists (scope gaps live between subcontracts) and the reason Chapter 19 exists (you do not manage the work; you manage the people who contracted to do the work).

The single most useful thing you can learn about a subcontractor before you sign them is not their price. It is their capacity and their cash. A sub whose backlog just tripled will put your job at the bottom of their crew list. A sub whose receivables are stretched will slow down, then slow down more, then disappear.

🏗️ From the field. I was 27 and running my first job of any size — a three-story office building — when the drywall sub stopped showing up. Forty percent complete, good crews, good foreman, then nothing. I did what a 27-year-old does: I called and demanded manpower. I wrote a letter about manpower. I put manpower on the agenda.

It was never a manpower problem. It was a payroll problem. Their biggest customer — a different GC across town, not us — was ninety days slow. They could not make Friday payroll on three jobs, so they made it on two, and we were the two-hundred-fifty-thousand-dollar job, not the two-million-dollar job.

Nobody had taught me to read a subcontractor as a business. I lost five weeks and a good subcontractor, and the whole thing was visible eight weeks earlier in the way their pay applications had started arriving on the first of the month instead of the twenty-fifth. Somebody was watching cash very carefully over there, and it should have told me something.

Now, on every job, I know which of my subs is thinly capitalized, and I pay them fast and I watch them close. It is not charity. It is schedule protection, and it is the cheapest schedule protection available.

2.5 The Supporting Cast: Suppliers, Sureties, Insurers, Lenders, and the AHJ

Five more parties, each with a business model you should be able to state in a sentence.

Suppliers and fabricators sell material, and their currency is lead time. A fabricator's real product is a slot in a production queue. Miss the slot and you do not lose the days you were late — you lose the days until the next slot, which is exactly what happened to Ironbridge Steel on Northgate: eleven days plus fourteen days of review turned into five weeks of mill availability and twenty-three days of schedule. Suppliers also carry escalation risk on quoted materials; quotations frequently expire in 30 to 90 days, and price protection beyond that is something you negotiate and often pay for.

Sureties issue bonds. Learn this sentence and never forget it: a surety underwrites the contractor, not the project. A bond is not insurance in the ordinary sense. Insurance is a pooled transfer of risk that expects losses. A surety bond is a three-party credit instrument in which the surety guarantees the contractor's performance to the owner, and if the surety pays a claim, it has a right of indemnity against the contractor — and typically against the contractor's owners personally, who signed a general indemnity agreement, often pledging their homes. The surety expects zero losses. That is why they underwrite your balance sheet like a bank and why bonding capacity is the hard ceiling on a contractor's growth. More in 2.7 and in Chapter 34.

Insurers price risk they expect to pay on. General liability, builder's risk, auto, umbrella, and workers' compensation — where the experience modification rate (EMR) takes your claims history and multiplies your premium up or down. A contractor with a poor safety record pays more for the same work than a contractor with a good one, on every job, forever, and many owners will not let a contractor above a certain EMR onto the site at all. Safety is not just an ethical obligation and not just a moral one. It is a line on your cost structure and a gate on your ability to bid. Chapter 24.

Lenders fund construction with a loan drawn in installments against verified progress, usually inspected by the lender's own consultant and converted to permanent financing at completion. A lender's dominant fear is a half-finished building, which secures nothing. That is why lenders care about lien waivers, why they hold retention, why they will not fund a draw with an unresolved lien — and why an owner who seems irrationally rigid about paperwork is often simply relaying their lender's requirements.

AHJs — Authorities Having Jurisdiction. The building department, the fire marshal, the health department, the elevator inspector, the environmental agency, the utility, the DOT for a lane closure. Frank Petrosyan is the building official for the Rivermont jurisdiction; he is not a party to your contract, has no interest in your schedule, and can stop your job with a piece of paper. An AHJ enforces the code that that jurisdiction adopted — often the International Building Code, NFPA 70 (the National Electrical Code), NFPA 101, the ADA Standards for Accessible Design, and an energy code, but always with local amendments. The amendments are where people get hurt. Read them.

The whole cast on one page

🔄 Check your understanding. A GC's fee on a job is 4 percent and its expected net profit on the same job is 2 percent. Where did the other 2 percent go — and what does that imply about a contractor who suddenly doubles their revenue?

Answer

It went to home-office overhead: estimating, accounting, safety, IT, insurance, marketing, rent, and the cost of pursuing jobs the company did not win — allocated across all revenue. The implication is subtle and dangerous: because overhead is largely fixed in the short run, adding revenue can look highly profitable at the margin, which is exactly the reasoning that leads contractors to chase volume. It works until the added volume requires added overhead — more staff, more systems, more supervision — and then the fixed cost steps up all at once while the thin margin has no room to absorb it. See 2.7 and case-study-01.md.

2.6 The Tier Structure: How Risk Flows Down and Money Flows Up

Now the diagram that explains the rest of the book.

📊 Diagram (described): the contractual tier structure of a typical building project.

Picture a pyramid with the owner at the top. The owner holds exactly two contracts of consequence: one with the designer and one with the general contractor. These two are peers — the architect does not work for the GC and the GC does not work for the architect, which surprises newcomers constantly and explains a great deal of friction.

Below the general contractor, the structure fans out: fifteen to forty subcontracts, one per trade package. Below several of those subcontractors, it fans out again into sub-subcontractors — the steel subcontractor's erector and detailer, the mechanical subcontractor's insulation and controls specialists. At the bottom sit suppliers and fabricators, who typically have no contract with anyone above their immediate customer.

Risk flows down the pyramid through flow-down clauses, which bind each subcontractor to the terms the tier above accepted. Money flows up the pyramid through the payment chain, slowly, and gets a little smaller at every step.

              LENDER ──── funds draws ────┐
        (construction loan)               │
                                          ▼
    DESIGNER ◄──── design contract ────► OWNER ◄──── AHJ (permit, inspect, occupy)
  (Architect of Record)                  (developer / institution / agency)
        │                                   │
        │ consultant agreements             │  PRIME CONTRACT
        ▼                                   │  ── risk flows DOWN ──▼
  Structural (Caldwell)                     │  ── money flows UP  ──▲
  MEP (Trellis)                             ▼
  Civil / Landscape / Specialty   GENERAL CONTRACTOR  ◄──── SURETY (bonds)
                                  (Kestrel)           ◄──── INSURERS
                                          │
   ┌──────────┬───────────┬───────────────┼────────────┬─────────────┐
   ▼          ▼           ▼               ▼            ▼             ▼
Concrete    Steel      Mechanical     Electrical     Drywall    Curtain wall
(SELF-    (Ironbridge) (Cardinal)     (Halcyon)                 + Precast
 PERFORM)     │            │              │             │
              ▼            ▼              ▼             ▼
          Erector      Sheet metal    Fire alarm     Framing
          Detailer     Insulation     Low voltage    Finishing
          Coatings     Controls            ── SUB-SUBCONTRACTORS ──
              │            │
              ▼            ▼
           Mill /      Equipment
         Fabricator     vendors     ── SUPPLIERS: no contract with the Owner ──

Three consequences fall directly out of that picture.

1. Nobody at the bottom has a relationship with anybody at the top. The drywall finisher's helper who is actually doing the work is four contract levels from the person paying for the building. Information that has to travel that far arrives late, distorted, or not at all. This is why we hold coordination meetings, why the daily report matters, and why the last-planner conversation in Chapter 27 is about talking to the people who actually control the work.

2. Risk gets pushed down until it lands on somebody who cannot price it. The owner shifts risk to the GC through the prime contract. The GC shifts it to the subs through flow-down clauses. The sub shifts what it can to its sub-subs and suppliers. Each transfer is legal and each is rational for the transferor. The problem is that risk keeps traveling downward until it reaches a party with no leverage to refuse it and no balance sheet to absorb it — a twelve-person subcontractor who signs a no-damages-for-delay clause because they need the work. When that party fails, the risk does not vanish. It ricochets straight back up to you as a schedule problem, and you now own it with none of the money you thought you had transferred. That is theme 1 in one paragraph: when you cannot say who owns a risk and what it costs them, you have found the thing that will hurt you.

3. The GC sits in the middle and finances the whole thing. Which brings us to the money.

Why the money moves so slowly

Follow one dollar of drywall labor from the moment the worker earns it to the moment their employer sees it.

 DAY 1–30        DAY 25         DAY 25–40        DAY 55–60        DAY 60–70
┌──────────┐   ┌──────────┐   ┌───────────┐   ┌───────────┐   ┌────────────┐
│ SUB does │   │ Sub bills│   │ GC assem- │   │ OWNER     │   │ GC pays    │
│ the work.│──▶│ GC by the│──▶│ bles pay  │──▶│ reviews,  │──▶│ subs, less │
│ Pays its │   │ 20th–25th│   │ app, cert-│   │ certifies,│   │ retention, │
│ crew     │   │ of month │   │ ifies, and│   │ funds the │   │ against    │
│ WEEKLY.  │   │          │   │ submits by│   │ draw in   │   │ signed lien│
│          │   │          │   │ the 25th  │   │ ~30 days  │   │ waivers    │
└──────────┘   └──────────┘   └───────────┘   └───────────┘   └────────────┘
     │                                                              │
     └──── the subcontractor's own money is out for 45 to 90 days ──┘
                     └─── plus 5% to 10% retention, held for MONTHS ───┘

Four things are happening at once, and each of them alone would be survivable.

The billing cycle is monthly, not continuous. Work performed on the 2nd of the month is not billed until the 25th. Twenty-three days of float lost before anybody even reviews anything. Northgate's terms are canonical and typical: pay application by the 25th, owner pays in 30 days.

Review and certification take time. The architect certifies progress. On some jobs the lender's inspector visits. Disagreements about percent complete are resolved by argument, not arithmetic, and argument takes days.

Retention is withheld. On Northgate, 10 percent until the job is 50 percent complete, then 5 percent. On Willow Street, 5 percent throughout. Retention is money you have earned, that the owner agrees you earned, that you will not see until closeout — sometimes a year or more later. Retention practice varies enormously: many states cap retention on public work, some require it to be held in escrow or to bear interest, some require reduction at substantial completion, and the rules change. Check the statute for your jurisdiction and read your contract; do not assume the number in this book applies to you.

Payment flows down the same tier structure, one step at a time. The owner pays the GC. The GC pays the subs. The subs pay the sub-subs and suppliers. Every step adds days, and every step adds a conditional.

Add it up. The drywall subcontractor paid its crew every Friday starting on day 3 and sees the money for that work somewhere between day 60 and day 90 — minus retention that will sit for another six to twelve months. The subcontractor is the project's real bank, and the general contractor is its second bank. This is the mechanism behind the industry's chronic cash stress, and it is why Chapter 32 and Chapter 34 exist as separate chapters: the first is about the job's cash, the second about the company's.

⚖️ What the contract says: "pay-when-paid" versus "pay-if-paid."

These two clauses look nearly identical on the page and are completely different in effect.

A pay-when-paid clause is generally read as a timing provision: the GC may wait a reasonable time for the owner's payment, but the GC eventually owes the subcontractor regardless.

A pay-if-paid clause attempts to make the owner's payment a condition precedent to the GC's obligation — meaning if the owner never pays, the subcontractor never gets paid, and the risk of the owner's insolvency has been transferred down the tier.

Whether a pay-if-paid clause is enforceable varies substantially by state. Some jurisdictions void them as against public policy. Some enforce them only when the risk-shifting intent is stated in unmistakable language. Some enforce them broadly. The law also changes, and courts distinguish between contract rights and rights under a payment bond or a lien.

This is not something to resolve by reading a textbook, including this one. Find the clause in your subcontract, find the same clause in your prime contract, and take both to a construction attorney licensed in the state where the project sits. What you should carry away is the concept: somebody in this structure is going to bear the risk that the owner does not pay, and the contract decides who. Find that clause before you sign, not after. Chapter 5.

🔄 Check your understanding. Your drywall subcontractor says they are "financing the job." Translate that into cash terms — what specifically are they financing, and for how long?

Answer

They pay their crews weekly beginning the first week of work, plus materials on supplier terms, but they do not bill until the 20th–25th of the month, the GC does not submit until the 25th, and the owner pays roughly 30 days later — so they are out of pocket for roughly 45 to 90 days of labor and material cost on a rolling basis. On top of that, 5–10 percent of everything they have earned is held as retention until closeout, often six to twelve months past the work. They are lending the project their working capital, interest-free, for the entire job.

2.7 Cyclicality, Backlog, Bonding, and Why Growth Kills Contractors

The cycle

Construction is one of the most cyclical industries there is. It is a derived demand business — nobody wants a building; they want the thing the building lets them do. When credit tightens, when interest rates rise, when a corporate board defers capital spending, when a state's tax receipts fall, construction demand does not soften gently. It stops. And when the cycle turns back up, it does not recover gently either.

The pattern repeats with depressing regularity:

1. Boom. Everyone is busy. Prices rise. Contractors are selective. Subcontractors will not return your calls. Margins improve, and contractors hire, buy equipment, and expand — because they extrapolate.
2. Peak. Capacity is fully deployed. Escalation is running hot. Owners complain about pricing. Some projects get shelved.
3. Contraction. Work dries up faster than overhead does. Now you have the staff, the equipment, and the office lease of a boom-era company and the revenue of a bust-era one.
4. Trough. Firms bid work they should not, at prices they cannot build for, because a job at zero margin still contributes something toward fixed overhead — and because the alternative is laying off the team they spent five years building.
5. Recovery — and here is where firms actually die. The backlog you bid at trough prices gets built in the recovery, at recovery-era labor rates and recovery-era material prices. You carried 3 percent escalation and got 11. Every job in the book is now underwater simultaneously.

Read step 5 twice. Contractors do not usually fail in the recession. They fail on the way out of it. The recession is when the fatal decisions are made; the recovery is when the bill arrives.

Backlog: the number that runs the company

Backlog is the total remaining value of work under contract but not yet performed. It is the single most-watched number in a contracting company, and it is watched in units of time: backlog divided by monthly revenue capacity.

Kestrel does about $410,000,000 a year, which is roughly $34,200,000 a month of revenue capacity. Most contractors this size want to carry somewhere between nine and eighteen months of backlog. Below that, the machine starts to starve; above it, you are turning away good work and straining your people. Kestrel is currently at 4.2 months. That is thin, and Nadia knows it, and everyone in that Monday meeting knows it. A hungry contractor makes bad decisions, and the first bad decision is always the same one.

Buying work

Buying work means bidding at or below your true cost, on purpose, to capture revenue — betting that you will make it back later through change orders, through squeezing your subcontractors, or through some efficiency you have not yet identified.

Curtis Boone bought Rivermont Elementary School #12.

Curtis was $780,000 below the next bidder — 3.4 percent — and $1,700,000 below the independent estimate — 7.1 percent. On a hard-bid public job, that is not winning. That is buying.

Being low is not itself proof of error; sometimes you genuinely are cheaper, and I will come back to that. But look at the shape of the spread. The second, third, and independent estimates cluster within about 4 percent of each other. Kestrel is the outlier. When four independent parties agree and you are alone on the other side, the base-rate assumption is not that you are brilliant.

Curtis's recovery plan was two-part, and both parts are worth naming because you will be tempted by both.

Part one was change orders. The plan was to find every ambiguity in the drawings and price it. That plan works right up until the moment the owner realizes it is a plan, at which point every legitimate change you submit for the rest of the job is treated as an attack. Curtis's job ends in a claim he cannot substantiate against an owner who no longer believes anything he writes. Chapter 33 shows exactly how that unwinds.

Part two was bid shopping — taking the subcontractor numbers he used to win, and then going back after the award to push them lower, playing subs against each other with the prices in hand. It is legal in most private contexts, it is restricted or prohibited on many public projects, and it is the fastest way to teach every good subcontractor in your market to pad your numbers or stop bidding you altogether. The dollar you extract this year, you pay back with interest for a decade. Chapter 16 draws the line precisely and shows what falls on each side of it.

Buying work can succeed, and honesty requires naming when:

• You have a genuine structural cost advantage the competition does not: you self-perform a major trade, you own the batch plant, you are already mobilized on the adjacent site, you built the identical building last year.
• Your marginal cost really is below your average cost, because your overhead is already fully absorbed by other work, and the job is short enough and simple enough that the probability of a real loss is small.

If neither is true, you are not buying work. You are placing a bet against your own client with your company's equity, at odds you did not calculate.

Barriers to entry — and the real one

The nominal barriers to entering construction are low. That is why there are hundreds of thousands of firms. But the barriers to entering your part of construction rise sharply with project size:

Kestrel's surety program is $150,000,000 aggregate and $60,000,000 single-project. Those two numbers are a hard governor on the company. The aggregate limits the total uncompleted bonded work Kestrel may carry at one time. The single-project limit caps any one job. Owen Baptiste spends real effort every year on the surety relationship — reviewed financials, a work-in-progress schedule, a business plan, an explanation of every job with fade — because the difference between a $60M single-project limit and a $75M one is the difference between competing for a class of work and watching it go by.

📋 Try it: can Kestrel bid the $58 million job?

A developer client has invited Kestrel to bid a $58,000,000 mixed-use project. Bids are due next month. Nadia asks you — you are the field engineer who sat in on the Monday meeting — to draft the capacity analysis before she decides.

Kestrel's company numbers:

The backlog — five largest active jobs, November of Year 1:

Everything else — the other six jobs on Nadia's board plus the small-projects group: $71,800,000 of remaining value, of which $33,600,000 is bonded.

Your task. Answer Nadia's question — can we bid it? — by working through four constraints in order: bonding, backlog, cash, and staff. Show your arithmetic. Then give her a one-paragraph recommendation. Give yourself twenty minutes before you open the answer.

Why growth kills more contractors than recession

Everything above converges on one of the most reliable patterns in this industry, and it is the subject of case-study-01.md in this chapter: contractors fail while growing.

Four mechanisms, all operating at once:

Growth consumes cash. Every new job requires front-end funding before the first payment arrives and permanently parks retention on the balance sheet. A company growing 40 percent a year is continuously funding the front end of new work out of the back end of old work. It looks like operating cash flow. It is not.

Growth dilutes supervision. A superintendent takes fifteen years to make. You cannot buy them in a quarter. Doubling revenue means either doubling the number of superintendents — which you cannot do — or doubling what each one carries. The second option is invisible on the income statement right up until two jobs go bad in the same month.

Growth outruns the systems. The cost-coding, the buyout process, the schedule updating, the safety program, and the monthly forecasting that worked at $80,000,000 do not work at $240,000,000. They fail quietly, and the first symptom is that nobody knows how a job is really doing until it is over.

Growth stretches the bonding ratios, and the surety notices. A surety sizes your program against your working capital and net worth. Grow revenue faster than equity and your ratios deteriorate. When a bad job finally shows up in the WIP, the surety does not just decline the next bond — it reduces the program you already have, and everything you have bid but not yet bonded evaporates.

🔄 Check your understanding. Without looking back: what is "buying work," and under what two conditions can it actually be a rational decision rather than a bet?

Answer

Buying work is deliberately bidding at or below true cost to capture revenue, betting on recovery through change orders, subcontractor squeeze, or unidentified efficiency. It can be rational when (1) you have a genuine structural cost advantage the competition lacks — self-performance, owned equipment, existing mobilization, a repeat building — or (2) your marginal cost is truly below your average cost because overhead is already absorbed by other work, and the job is short and simple enough that the probability of a real loss is low. Absent both, it is a wager against your own client, placed with the company's equity.

2.8 The Labor Picture

Every bar on your schedule is a claim that a specific number of trained human beings will show up on a specific day, understand the plan, and cooperate with three other trades working in the same room. That is theme 6, and in the next decade it is going to be the binding constraint on this entire industry.

Here is the honest, hedged picture.

The craft workforce is aging. The average age of skilled construction craftworkers has been drifting upward for years — the Bureau of Labor Statistics publishes the demographics, and the direction is not in dispute. A great many highly skilled people are within a decade of retiring, and they carry knowledge that lives nowhere else. Nobody has written down how Jamal Foster reads a formwork joint or how Margo Deacon knows, from the sound of a job at seven in the morning, that something is wrong.

The pipeline is thinner than the demand. Two generations of policy and culture pushed young people toward four-year degrees and away from the trades, and many high schools eliminated shop programs entirely. Apprenticeship capacity — through union joint apprenticeship and training committees and through open-shop programs run by associations like the AGC and ABC — is real and it works, but it has not been scaled to the size of the gap. Industry associations have been sounding this alarm for years, and the shortage tends to worsen exactly when construction demand is strongest, which is the least convenient possible timing.

Immigration is a significant part of the workforce in many markets. In much of the country, a substantial share of craft labor is foreign-born. Immigration policy is therefore construction labor policy, whatever anyone's politics, and policy volatility shows up directly in crew availability and wage rates. State that neutrally to yourself and plan for it.

Union and open shop are two different labor supply systems, not two ideologies. Under a union agreement, you get trained, credentialed craft delivered through the hall, with negotiated wage and benefit rates, defined jurisdictional boundaries, and portable training you did not pay to build. Open shop, you own the entire problem: recruiting, screening, training, retention, and career pathing — and you also own the flexibility and, in many markets, a lower direct wage rate. Kestrel runs union on public work (where prevailing wage largely erases the rate advantage anyway) and open shop on most private work. Both models produce excellent work and both produce bad work; the model is not the variable, the training system is.

Why this is a schedule problem before it is a cost problem. A labor shortage does not usually announce itself as a wage increase. It announces itself as a subcontractor who bid four crews and can staff two. Your durations were built on crew sizes and productivity rates; when the crew is half the assumed size, the duration doubles, and you find out in week three of a fourteen-week activity. Money you can forecast. Absent people you cannot — you can only replan around them, and replanning costs float you have already spent.

This is why Chapter 20 sits where it does, and why it treats crew composition and man-hour budgeting as core project-controls skills rather than an HR topic.

2.9 The Industry's Chronic Problems, Stated Honestly

I am not going to sell you an industry. You will make up your own mind, and you should make it up with the real numbers.

Construction has five chronic problems. They are genuine, they are old, and they are not going to be fixed by a slogan or an app. But every one of them is also an opportunity, because in an industry with a two percent margin, a manager who is materially better than average is worth an enormous amount of money to somebody.

1. Productivity growth has lagged badly. Analyses over several decades have consistently found that construction labor productivity has grown far more slowly than manufacturing's — some studies find it essentially flat over long periods. The measurement is genuinely contested (construction output is hard to define when every product is unique, and quality and code requirements have risen substantially), but the direction is not seriously disputed by anyone who has stood on a job site and watched how much of a craftworker's day is spent waiting, walking, looking for material, or redoing something.

2. Relationships are adversarial by default. The contract structure in 2.6 places every party's interests in partial opposition and then hands them the tools to fight about it. A great many projects are run by people who genuinely like each other and still end in a claim, because the structure rewards it.

3. The safety record is unacceptable. Construction accounts for roughly one in five workplace fatalities in the United States, while employing a considerably smaller share of the workforce. That ratio has been stubbornly persistent for decades. The Bureau of Labor Statistics publishes the Census of Fatal Occupational Injuries; OSHA's framing of the leading causes as the Focus Four — falls, struck-by, caught-in/between, and electrocution — is the standard organizing framework, and 29 CFR Part 1926 is the construction standard itself.

⚠️ Safety alert. Read the tier diagram in 2.6 one more time, this time as a safety diagram. Schedule pressure enters at the top — an owner's opening date, a lender's deadline — and travels down through the prime contract to the general contractor, through flow-down clauses to the subcontractor, through the subcontractor's foreman to a crew of four people on a scaffold. Each transfer strips out context and adds urgency. What arrives at the bottom is not a business decision. It is a man being told to make up two days on the north elevation.

On Northgate, in week 34, a mason tender named Milo Serrano stepped onto a scaffold plank that a different trade had lifted overnight to run conduit and had not re-secured. The plank shifted. He caught himself on the top rail. Nobody was hurt.

The investigation found three failures, and the third one is the one nobody wanted to write down: the crew was running behind after the steel acceleration, under an unwritten expectation to make it up. That expectation was not in any contract. It came down the tiers anyway. Schedule pressure is a hazard exactly like an unguarded edge, and it is one that a project manager creates. Chapter 24 takes this apart properly.

4. Fragmentation blocks improvement. Covered in 2.1: no firm is large enough to set a standard, R&D is chronically underfunded relative to the sector's size, and every project rebuilds its organization from scratch.

5. Profitability is thin and fragile. Two to four percent net for a general contractor. One bad job can consume a year of company profit — you saw the arithmetic in 2.4, where a single late submittal took a quarter of Northgate's profit and would require $10.65 million of new revenue to replace.

What this book does about each of them

This is not a book about complaining. Here is the map.

Notice that the answer to every one of them is the same shape: see the problem earlier, write it down, and price it. That is the job. It is not glamorous and it is worth a great deal of money.

🔄 Check your understanding. Why is a labor shortage a schedule problem before it becomes a cost problem?

Answer

Because schedule durations are built from assumed crew sizes and productivity rates. A shortage shows up first as a subcontractor who committed four crews and can field two — the duration stretches while the unit cost may not move at all. You can forecast a wage increase; you cannot forecast people who are not there. You can only replan around them, and replanning consumes float you have already spent elsewhere.

Project Checkpoint: Market and Player Analysis for Willow Street

In Chapter 1 you opened your Willow Street Project Notebook with a project charter — your role, your stakeholders, and a written definition of what success means. Keep it open; this checkpoint sits on top of it and makes it specific.

Your project: the Willow Street Community Center — $6,800,000, 24,000 square feet, gymnasium, two multipurpose rooms, commercial kitchen, offices, and locker rooms for the City of Rivermont Parks & Recreation. Design-bid-build, lump sum, 425 calendar days, LDs of $1,200 per calendar day, 5 percent retention, 100 percent payment and performance bonds, prevailing wage. Full package: Appendix K.

Deliverable: a market and project-type analysis, in three parts. Two to three pages.

Part 1 — Classify the job. One paragraph each: What sector and sub-sector is this? What type of owner, and what does that mean for how they buy, how fast they decide, and how they behave when something goes wrong? What is the delivery environment — how is the contract awarded, what is fixed, what is negotiable, and where does risk sit before you have seen a drawing? Name at least three consequences of the public-owner classification that will change how you run the job — think prevailing wage and certified payroll, the absence of lien rights and the presence of a payment bond, and how a $30,000 change actually gets approved.

Part 2 — Map every player. A table with a row for every party who will touch this job, and do not stop at the obvious ones: the City as owner; the Parks & Recreation staff who will actually operate the building (not the capital projects office); the architect and their structural, MEP, and civil consultants; the general contractor; the trade subcontractors you can predict from the program; suppliers with real lead times; the surety and the insurers; the AHJs — building department, fire marshal, health department for the commercial kitchen, water utility for that main relocation; and the neighbors and park users, who have no contract and considerable influence.

Part 3 — One line per player: what they want, and how it conflicts with what you want. This is the part that matters, and it must be specific. Not "the architect wants a good building," but: "The architect wants the exposed heavy-timber connections detailed as drawn; I want a connection the framer can build in the field without a two-week RFI cycle in the middle of the enclosure sequence." Not "the city wants to save money," but: "The capital projects office wants no change orders, because each one needs a council agenda item — so a legitimate differing-site-condition claim on the water main will move slowly, and I must give notice early and in writing."

Aim for fifteen to twenty-five rows. Then circle the three conflicts most likely to cost you money or time. Those three are the seed of the risk register you build in Chapter 6, and two of them will resurface in your bid/no-bid analysis in Chapter 15.

Next checkpoint: In Chapter 3 you will write a delivery-method comparison memo — why design-bid-build was chosen for Willow Street, what CM at Risk or design-build would have changed, and what you would have recommended to the City if they had asked you first. The player map you build today is the evidence for that memo, because delivery method is fundamentally a decision about which of these parties absorbs which unknown.

Chapter Summary

A reference card for the business you have joined.

The industry in six facts

The five questions to ask about any project, before anything else

1. Who is the owner, and are they spending their own money or the public's? That answers how the job was bought, how fast decisions move, whether prevailing wage applies, and whether you have a lien right or a payment bond.
2. What sector is this, and what does that sector's risk model look like? Unit price and differing site conditions on civil; design coordination and schedule on commercial; process and startup on industrial; cycle time and sales price on residential.
3. Where does the money come from, and how does it flow down the tiers? Find the lender, the retention terms, the payment cycle, and the pay-if-paid clause before you sign anything.
4. Who owns each significant risk, and can that party actually absorb it? Risk pushed onto a party that cannot carry it is not transferred. It is deferred, and it comes back as a schedule problem you own.
5. What is the daily cost of being late? On Northgate it is $10,650. Know your number on day one, because it converts every schedule conversation into a money conversation, which is the only kind that gets decided.

The three arithmetic facts to carry in your head

• Northgate's fee is $1,804,800 and its projected net profit is $957,300 — the fee is not the profit; home-office overhead takes over half of it.
• 23 calendar days × $10,650 per day = $244,950, roughly a quarter of that profit, from one submittal that sat eleven days on a desk.
• Replacing $244,950 of profit at a 2.3 percent net margin requires $10,650,000 of new revenue. Preventing a loss is worth vastly more than winning more work.

The one sentence

Construction is a low-margin, high-risk, fragmented business in which every project assembles a temporary organization out of firms with divergent incentives, pushes risk downward until it lands on someone who cannot price it, and pays for the work two to three months after it is performed — which is precisely why the manager who sees the problem four weeks early, writes it down, and prices it is the most valuable person on the job.

What's Next

You now know who the players are and what each of them wants. Chapter 3 asks the next question: how do the owner, the designer, and the contractor get arranged? Design-bid-build, design-build, CM at Risk, and integrated project delivery are four different answers, and each one moves the unknowns to a different party at a different price. Then Chapter 4 puts a dollar sign on that arrangement — lump sum, GMP, cost-plus, unit price — and shows you that the delivery method and the contract type are really one decision, and that the decision is the price.