Chapter 19 — Subcontractor Management: Selection, Coordination, Quality, and Getting 20 Trades to Work Together

The Hook: Forty Minutes on Level Three

Month eleven. Tuesday, January 20 of Year 2. Northgate is dried-in on the north half of levels two and three — the by-area enclosure resequence we bought during the steel acceleration — and the interiors have started loading up behind it. I am in the trailer with a cost report open when Dani Okonkwo comes through the door without knocking, which is how I know it is bad.

"Zone 3C is stopped," they say. "All four crews. Nobody's working."

Zone 3C is eighteen hundred square feet on the third floor: eleven exam rooms and a corridor. Here is what was standing in it at 7:05 that morning.

Six of Sightline Interiors' framers, finished. Their metal-stud framing in 3C was done Friday and there was nothing else framed-ready in that wing.

Four electricians from Halcyon Electric, where the schedule said six. Two of Halcyon's people had been pulled to a data-center job across town that was three weeks behind and screaming louder than we were.

Four insulators from Thermaline Insulation, who showed up exactly when the six-week look-ahead told them to show up, and who could not install one batt of insulation because the electrical rough-in had not been inspected.

And eight board hangers under Roddy Vance, Sightline's drywall foreman, who had nothing to hang, and who told Dani — using words I will paraphrase — that if there was no wall to close by Thursday he was moving his crew to a hotel job across Rivermont and Kestrel would see him again in three weeks.

Twenty-two people. One room. Zero production.

Margo Deacon, our general superintendent, was on level four when I called her. She was in 3C ninety seconds later and she did not raise her voice once.

Minutes 0 to 8. She walked the zone with all four foremen — in the room, not in the trailer — and made them show her the actual constraint instead of describing it. It turned out that eight of the eleven exam rooms had complete electrical rough-in. Three did not. Halcyon had requested the rough-in inspection as one package for the whole zone, so eight finished rooms were being held hostage by three unfinished ones.

Minutes 8 to 14. She called Frank Petrosyan's office at the building department and asked for a partial rough inspection Wednesday morning for rooms 3C-01 through 3C-08, with the area taped off and the room numbers listed on the request. His office allows that if you demarcate it and you do not waste their trip. That one phone call unlocked eight of the eleven rooms.

Minutes 14 to 22. She moved everybody who was standing still. Thermaline's four went to Zone 3A in the west corridor, which had passed inspection the previous Thursday and had been sitting untouched because Thermaline was scheduled here. Sightline's six framers went up to Zone 4B and started layout and track two weeks early — work that was always going to happen, just not that morning. Vance's eight split: four followed Thermaline into 3A to hang one side, four went down to the level-two corridors to hang the soffits we had deferred in December.

Minutes 22 to 31. She called Yolanda Prieto, who owns Halcyon Electric. Not Devlin Achebe, Halcyon's project manager, who had already told Margo three times that he was doing what he could with the bodies he had. She called the owner, and she did not threaten anybody. I will show you exactly what she said in §19.6, because the script matters more than the escalation.

Minutes 31 to 40. She wrote it down. Text to me at 7:44. My email to Devlin and Yolanda went out at 2:10 that afternoon: the facts, the commitments, the dates, the resequence, and one neutral sentence about impact. Dani's daily report carried the crew counts by trade and three photographs.

Wednesday's partial inspection passed at 9:40 a.m. Halcyon had seven people Thursday and eight Friday. The three remaining rooms were roughed Thursday and inspected the following Monday. Vance did not move his crew.

Cost of the forty minutes: about $310 in supervision time and one phone call. Cost of letting it run three days, which is what would have happened without her: somewhere between $18,700 of burned craft labor — mostly Sightline's money, not ours — and $50,000-plus once you add the schedule and the risk that Vance actually left. I work that arithmetic in §19.6 and again, in full, in Case Study 1.

That is this chapter. Not the forty minutes. The why underneath them — why Margo reached for a resequence and a phone call instead of an order, and why the version of this job where she reaches for the subcontract first is a job that ends in a claim.

🏃 Fast Track: If you have run field operations for a few years, skim §19.1 and §19.3, and read §19.2 (the threshold concept), §19.5 (manpower as a leading indicator and the escalation ladder — the sample notice language is the part people get wrong), and §19.9 (the view from the other side). If you have never written a cure notice, do not skim §19.5.

🔬 Deep Dive: Buyout, scope sheets, and flow-down are Chapter 16. The bonds and legal machinery behind a default are Chapter 5. Making work ready so coordination never gets to this point is Chapter 27. Proving a disruption claim is Chapter 33.

19.1 What a General Contractor Actually Is

Let me start by taking something away from you.

You are not a builder. Not on a job like this one. On the Northgate Outpatient Pavilion, Kestrel Construction Group self-performs concrete, rough carpentry, and general trades — roughly $5,900,000 of a $40,000,000 direct cost of work — plus about $1,300,000 of direct material and equipment purchases. Everything else, $32,800,000 across 34 subcontracts, is performed by companies that do not work for you, do not report to you, do not care about your bonus, and have a completely different idea than you do about what the word "priority" means.

That is 82 percent of the work. On a typical commercial building project the number generally lands somewhere between 80 and 90 percent, though it varies enormously by market, project type, and how much the general contractor self-performs. I have run a job that was 100 percent subcontracted and a job that was 55 percent. The principle does not change: most of your project is being built by strangers with their own profit-and-loss statement.

Here is what Kestrel's ten largest Northgate subcontracts look like. Study the concentration.

(Every company and person in this book is a Tier-3 illustrative composite. The numbers are internally consistent and realistic; they are not a real project.)

Ten companies hold roughly seventy percent of the work. If any one of the top four fails, the job is in serious trouble no matter how good your bond is. That is the first thing to understand about your role.

19.1.1 Three jobs in one title

You are a coordinator. Somebody has to decide who is in which room on which day, in what order, with what left behind for the next trade. Nobody else on the project can do this. The architect does not know your crew sizes. The owner does not know your sequence. Each subcontractor knows only their own scope and their own backlog. The sequence exists in exactly one head and in exactly one document, and both of them are yours.

You are an integrator. Thirty-four scopes of work were bought separately, priced separately, and drawn by different consultants, and they have to become one building that passes inspection and holds water. The seam between two subcontracts is where buildings leak, fail inspection, and generate change orders. You are the only party contractually obligated to care about the seam. This is the field consequence of the buyout lesson in Chapter 16: scope gaps live between subcontracts, and eight months later they surface as a fight about who owns a sealant joint.

You are a risk aggregator. This is the one nobody teaches. You signed a contract with Meridian Health System promising a building for $47,500,000 in 565 calendar days. You then bought thirty-four pieces of that promise from thirty-four companies of varying financial strength, and you kept the difference between what you promised and what you bought. Every one of those thirty-four is a small probability of failure, and you are holding all of them at once.

📊 Diagram (described): the contractual chain, and where the risk piles up.

The Northgate contract structure is a fan, not a chain, and the general contractor sits at the narrow end holding every strand. Meridian Health System has exactly one construction contract — with Kestrel. Kestrel has thirty-four. Each of those subcontractors has its own suppliers and lower-tier subs; Sightline alone buys board and metal stud from Ridgeline Gypsum Supply and hires an independent taping crew. Roughly a hundred second- and third-tier companies have money riding on this job, and Meridian has a legal relationship with none of them.

``` MERIDIAN HEALTH SYSTEM (owner) | one contract: $47.5M GMP | KESTREL CONSTRUCTION GROUP <-- ALL risk aggregates HERE self-perform $5.9M | 34 subcontracts $32.8M ___|___ | | | | | | Cardinal Ironbridge Halcyon Aperture Sightline (29 more) $6.40M $4.80M $4.60M $3.40M $2.10M | | | | | suppliers, lower-tier subs, rental houses, taping crews (~100 companies with money riding on this job)

Meridian's exposure if Sightline goes broke: none, contractually. Kestrel's exposure if Sightline goes broke: all of it. ```

Read the diagram twice. The owner bought a single point of accountability. That single point is you.

19.1.2 The honest framing of the relationship

I want you to hold two ideas at the same time, because young project managers usually pick one and get hurt.

Idea one: the subcontractor is not your employee. They have their own business, their own backlog, their own cash-flow problem, their own bonding capacity, and their own opinion about where your job ranks among the six jobs they are running. Halcyon Electric has about 240 people. Northgate is roughly 14 percent of Halcyon's year. When their data-center job started losing money, the two electricians they pulled off my floor were a rational business decision made by a person who was not being lazy or dishonest. He was triaging.

Idea two: you are accountable for their performance to somebody who does not care about idea one. Pri Sethi, Meridian's owner's representative, is not interested in Halcyon's data center. Her board asked her when the clinic opens. "My electrician is short-handed" is not an answer that survives contact with a hospital operations committee.

Those two ideas live in permanent tension, and the whole craft of this chapter is managing it. Here is the resolution, and it is the sentence I would tattoo on a field engineer:

Your job is to be the customer they want to perform for — and to have the contractual leverage for when they don't.

Both halves. A general contractor who is only pleasant gets rolled. A general contractor who is only contractual gets the minimum performance the contract can be made to require, which on a construction project is remarkably little and remarkably slow.

🔄 Check your understanding. Kestrel's top four Northgate subcontracts total $19,200,000. Why does that concentration matter more to a general contractor than the same dollar value spread across twenty subcontractors?

Answer

Because risk does not average out when it is concentrated. Twenty $960,000 subcontracts are twenty small, largely independent failure probabilities, and losing any one is survivable — plenty of firms can pick up a $960,000 scope on short notice. One $6,400,000 healthcare mechanical subcontract is a different animal: very few firms in the market can bond it, staff it, or absorb the mid-project learning curve; the replacement premium is brutal; and the schedule impact of a changeover runs to months rather than weeks. Concentration also removes your practical leverage — you cannot credibly supplement or replace a subcontractor whose scope nobody else can perform on your timeline. That is why the prequalification bar rises with package size, and why the top four names on this list get more of your attention every week than the bottom twenty combined.

19.2 🚪 Threshold Concept: You Do Not Manage the Work

🚪 Threshold concept. You do not manage the work. You manage the people who contracted to do the work. Your leverage is the subcontract, the schedule, and the coordination — it is not authority. A general contractor who tries to command subcontractors instead of enabling them and holding them accountable will fail at both.

This is the gateway. Readers who cross it manage differently for the rest of their careers. Readers who do not spend twenty years being frustrated by the same thing every Tuesday.

Before you understand this, the job looks like a chain of command. There is a schedule; the schedule says Halcyon is supposed to have six electricians in Zone 3C; Halcyon has four; therefore somebody needs to tell Halcyon to bring six. You escalate. You raise your voice in the coordination meeting. You send an email with the schedule attached and the words "per the subcontract" in it. When that does not work — and it does not work — you conclude that this particular subcontractor is unreasonable, and you start building a file.

I ran a job like that when I was twenty-nine. I was right about almost every fact and I lost anyway, because being right is not a management technique.

After you understand this, the job looks like an economy. Every subcontractor on your site is making a continuous, rational, self-interested decision about where to put their people tomorrow. Your competitors for those people are the six other jobs that company is running. You cannot order those people onto your floor. What you can do is change the inputs to their decision so your floor becomes the obvious place to send them — and, when that fails, use the contract you both signed to make non-performance more expensive than performance.

The word that changes is not "authority." It is leverage.

19.2.1 The four sources of leverage, in order of usefulness

I am giving these to you in order deliberately, because the order is the lesson.

Number one is the big one, and it is almost entirely within your control. A subcontractor makes money on your job by getting a crew into a released area and keeping them producing until they leave. Everything that interrupts that — a wall that is not framed, a hoist that is not available, an RFI (request for information) that has been open eleven days, a decision the owner has not made, a pallet of material in the way, an area you released and then took back — costs them money directly. A general contractor who runs a clean, released, decided job is literally more profitable to work for than one who does not, and subcontractors know exactly which kind you are within about three weeks of mobilization. This is the field payoff of everything in Chapter 18: a well-run site removes the reasons people stand still.

Number two is the one contractors squander for no reason at all. More on that in §19.8.

Number three is where most of your actual weekly work lives. A schedule everybody believes is a coordination tool. A schedule everybody knows is fiction is worse than no schedule, because it teaches your subcontractors to ignore all of your dates — including the true ones.

Number four is real, and it is last for a reason. The enforcement clauses in your subcontract are genuine and you must know them cold. But notice what they cost. The moment you send a formal cure notice, you convert a working relationship into a documented adversarial position, and everybody on both sides starts writing letters instead of hanging drywall. Enforcement is slow, expensive, and it does not put a single electrician on your floor tomorrow.

💡 Aha moment. A general contractor who reaches for #4 first has usually already failed at #1 through #3. When I am asked to review a job that is in trouble and I find a thick file of cure notices, I do not immediately conclude the subcontractors were bad. I go look at the RFI log, the released-area log, and the payment history — and about two-thirds of the time I find the general contractor's own fingerprints all over the cause.

19.2.2 The honest exception

Now the caveat, because the paragraph above gets misread by people who do not like conflict.

Sometimes #4 is the only thing left. Some subcontractors are underwater on your job because they bought it at a number that never worked. Some are failing for reasons that have nothing to do with you — a founder who died, a bank that pulled a line of credit, a project three states away that ate the company. Some are simply badly run. In those situations being nicer does not help, and a general contractor who keeps hoping while the schedule bleeds is not being kind. He is quietly transferring his subcontractor's problem onto his owner, his surety, and every other trade on the job.

The discipline is this: be slow to start the ladder and fast to climb it. Do not send a notice you have not earned. But once you have earned it, send it that week, follow it exactly, and do not skip a rung. §19.5 is the ladder.

🔄 Check your understanding. Your fire-protection subcontractor has been short-handed for three weeks. Before you send anything in writing, which three questions must you be able to answer about your own performance?

Answer

(1) Did I release the work? Are the areas actually ready — framed, roughed by the trades ahead of them, free of stored material, accessible, lit, heated? (2) Did I answer their questions? Any open RFIs, unreturned submittals, or undecided owner items blocking their fabrication or installation? (3) Did I pay them? Are their applications processed on time, at the right amounts, with no unexplained deductions and no slow-walked retention? If the answer to any of these is no, your notice will be answered with a one-page letter and an attachment that makes you look careless, and you will have spent your credibility for nothing. Fix your side first. It usually takes a day, and it is usually the actual cause.

19.3 Selection: Getting the Right Subcontractors on the Job

Most subcontractor problems are selection problems that arrived on a truck eleven months later.

Chapter 16 covered prequalification as a procurement exercise: financial statements, bonding capacity, safety record, licensing, references, and a scored matrix. I am not going to repeat it. I want to give you the part that only shows up in the field, and it is the part that decides whether your year is pleasant.

19.3.1 The four things prequalification does not tell you

Does the company put its A-team on your job? Every subcontractor of any size has a best foreman, a competent middle, and one crew they hope you never meet. Prequalification measures the company. The field experiences the crew. A firm with a spotless balance sheet that sends you its C-team will hurt you more than a marginal firm that sends its best people because your job is the one they want.

What is the foreman's quality? This is the single highest-correlation predictor of how a trade performs on your project, and it appears nowhere in a prequalification package. A good foreman plans two weeks out, tells you the truth about manpower, catches his own defects, and protects the trade behind him. A bad one runs the crew he was given at the wall in front of him and lets you find out about problems when they are already three rooms deep.

Does the office answer the phone? Submittals, RFIs, change-order pricing, certified payroll, insurance certificates, lien waivers, and closeout documents all come from a project manager or a coordinator sitting in an office somewhere. A firm with excellent field crews and a two-person office that returns calls in four days will cost you real money in delay and rework.

What is your job's share of their year, and what is its rank? A $2,100,000 subcontract with a firm doing $60,000,000 a year is 3.5 percent of their business and will lose every internal fight. The same subcontract with a firm doing $9,000,000 a year is 23 percent of their business and gets the owner's personal attention — but it may also strain their working capital, their bonding capacity, and their supervision bench. There is no universally right answer. There is only knowing which situation you are in and managing accordingly.

19.3.2 Reference questions that produce real answers

The standard reference call is worthless. "Would you use them again?" gets you "yes" ninety-eight percent of the time because nobody wants to knife a competitor in writing, and because the person answering has thirty seconds of interest in your problem.

Ask questions that require a specific answer instead of a judgment.

Two more that I always ask and that almost nobody else does:

"Who did they use for their taping crew / their erection crew / their controls sub?" Most trades subcontract part of their own scope. You are not just buying Sightline; you are buying whoever Sightline hires. Knowing that in advance lets you prequalify the tier below.

"If you had it to do again, what would you have put in the subcontract?" This is the question that gets you the truth. People who will not criticize a competitor will happily tell you what they wish they had written down.

🏗️ From the field. I once took a mechanical reference from a project manager who answered every question with a version of "they were fine." On the last question — what would you have put in the subcontract — he paused for eight seconds and said, "A requirement that their coordination drawings be signed by the same person every month." That sentence told me the whole story: their detailing was passed between people, nothing was owned, and coordination drifted. We wrote that requirement into the subcontract and it saved us a month.

19.4 The Kickoff: The Highest-Return Hour on the Job

Every subcontractor gets a pre-mobilization meeting before their first crew walks through the gate. On Northgate we held thirty-four of them. They run sixty to ninety minutes, they happen two to four weeks before mobilization, and I consider that hour the highest-return hour in the entire relationship.

Here is the reason, stated as coldly as I can. Expectations you set at the start are free. Expectations you set in month eleven cost money, because by month eleven every expectation is competing against a habit the subcontractor has already formed, a crew that has already been assigned, and a price that has already been bid.

Who attends: the subcontractor's project manager, the general foreman or foreman who will actually run the work, and — this is not negotiable — somebody with authority to commit manpower. From Kestrel: the project manager, the superintendent for that area, the project engineer, and the safety manager. If the subcontractor sends only a salesperson, reschedule. That is not rudeness. That is the first expectation you are setting.

19.4.1 The agenda

Item 9 deserves a sentence of its own. Tell every subcontractor, in the kickoff, the name of the only Kestrel person who can direct extra work — and tell them that if anyone else directs it, they will not be paid for it. On Northgate that person is me, and in my absence Margo. This single sentence is the cheapest insurance you will ever buy, and the story of what happens without it is the canonical CO #14 mess: an assistant superintendent under schedule pressure let a concrete crew build a deeper depressed slab on a verbal go-ahead, and Kestrel ate $43,650 of a $186,400 cost because nobody had written down who could say go.

19.4.2 The artifact that comes out of it

The kickoff produces a two-page subcontractor expectation memo, signed by both parties, distributed within 48 hours, and filed with the subcontract. Not minutes — a memo, written in the form of commitments. It contains the manpower curve, the submittal dates, the mockup dates, the named competent persons, the named change-order authority, the notice sheet, and the meeting attendance requirement.

When a subcontractor falls behind eight months later, that memo is the first document you attach to anything. It is the difference between "we expected more people" and "on April 14 your project manager committed in writing to 18 electricians in weeks 45 through 52 and has averaged 12.5."

⚖️ What the contract says. The kickoff memo does not amend the subcontract and you should never pretend it does. Its power is evidentiary, not contractual: it establishes what the parties understood, contemporaneously, before anybody had a reason to lie. Most standard subcontract forms already obligate the subcontractor to furnish sufficient properly skilled workers, materials, and equipment to maintain progress in accordance with the schedule, and to participate in coordination. The memo makes that abstract duty specific and dated. Read your own form for the actual language — obligations vary between AIA, ConsensusDocs, and the custom forms most large general contractors use, and a custom form written by your company's counsel is usually more demanding than either published form.

🔄 Check your understanding. Why is an expectation set at the kickoff meeting cheaper to enforce than the identical expectation set in month eleven?

Answer

Because by month eleven the expectation is competing against three things that did not exist at kickoff: a habit the subcontractor's crews have already formed, a crew assignment already made on the basis of the old understanding, and a price already bid without the new requirement in it. Change any of those mid-job and the subcontractor has a legitimate argument that you are asking for something they did not price — which converts a management instruction into a potential change order. At kickoff, none of that has happened yet: the requirement is simply part of the deal, absorbed at zero marginal cost. Zone 3C is the illustration. Halcyon's practice of requesting inspections one whole zone at a time was a stable, knowable behavior that a single kickoff question would have surfaced, and eleven months later it cost a morning and nearly cost a drywall crew.

19.5 Manpower Is the Leading Indicator

Here is the most useful thing in this chapter, and it takes about four minutes a week.

Percent complete is a lagging indicator. By the time a trade's percent complete is visibly behind, they have already been behind for a month and the recovery window has closed. Manpower is a leading indicator. A trade that is 20 percent short on manpower today is 20 percent short on progress three weeks from now, and today is when you can still do something about it cheaply.

Your daily reports already collect crew counts by trade — Chapter 26 covers the discipline of that. Roll them up weekly against the curve each subcontractor committed to in the kickoff, and you have an early-warning system that costs you almost nothing.

19.5.1 Reading the curve

This is the Northgate interior build-out, weeks 45 through 52 of the job — roughly early January through late February of Year 2. P is the promised weekly headcount from each subcontractor's kickoff manpower commitment. A is the actual average headcount from the daily reports.

Two trades are visibly short: Halcyon at 70 percent and Sightline's board crew at 76 percent. A project manager who has not learned to read this table sends two letters. A project manager who has learned to read it sends one, because only one of those two shortfalls is a cause. The other is a symptom.

Sightline cannot man up the board crew until the walls are closed. Closing the walls requires Halcyon's rough-in and Frank Petrosyan's inspection. Sightline's 76 percent is not Sightline's fault; it is Halcyon's shortfall arriving downstream, one trade later. Send Sightline a letter about it and you will get back a letter that makes you look like you do not understand your own schedule — plus, quietly, a permanently worse relationship with the trade that is going to close out your building.

🔍 Why this works. Manpower is a leading indicator because construction production is, over a period of weeks, roughly proportional to labor hours applied. Progress this month was purchased by hours worked last month. A crew-count shortfall is therefore visible in the daily report weeks before it is visible in the percent-complete column of the schedule update, which is exactly the window in which corrective action is still cheap. The second mechanism is subtler and more important: manpower is self-reported by the subcontractor in advance. When a firm commits to a curve and then does not staff to it, you have not just detected a schedule problem — you have detected a commitment problem, and commitment problems predict every other kind of failure the company is going to have on your job.

19.5.2 Setting early-warning thresholds before you need them

Decide your thresholds at buyout, not in the moment. Otherwise every conversation becomes an argument about whether the shortfall is "really" a problem.

Use a four-week rolling average, not a single week. One bad week is weather, a funeral, or a truck that broke. Four weeks is a decision somebody made.

19.5.3 The escalation ladder

When a subcontractor falls behind and the cause is genuinely theirs, there is a sequence. Skipping a rung does not accelerate anything. It destroys your position later, because every enforcement remedy in your subcontract is conditioned on notice and an opportunity to cure, and a remedy you exercised without them is a remedy you will have to give back.

Note what rung 2 does. Escalating to the owner of the firm is not a threat and it is not going over somebody's head to punish them. It is recognizing that a project manager cannot conjure electricians out of a company that has committed them elsewhere. The manpower decision is made above the project manager's pay grade, so the manpower conversation belongs above his pay grade too. Tell his PM you are making that call, before you make it. Every time.

19.5.4 Sample notice language

I am giving you three. They are illustrative starting points, not legal forms — your subcontract's notice clause governs the required content, the delivery method, the recipient, and the number of days, and those vary between contract forms and between jurisdictions. Read yours before you send anything, and have your first few reviewed by counsel.

Rung 3 — written notice of the schedule requirement. Neutral, factual, no adjectives.

Re: Northgate Outpatient Pavilion — Subcontract 16-A, Electrical — Schedule and Manpower

Devlin: This letter confirms our conversations of January 20 and January 27 regarding manpower in the level 3 and level 4 interior areas.

Halcyon's pre-mobilization manpower commitment dated April 14, Year 1 provided for an average of 18 workers in weeks 45 through 52. Kestrel's daily reports for that period record a four-week rolling average of 12.5 workers, or 70 percent of the committed level. Halcyon's remaining electrical rough-in scope in these areas is approximately 5,600 labor hours against a required completion of the end of week 56.

Kestrel confirms that all areas identified on the attached released-area log have been released to Halcyon, that no Halcyon RFIs are currently open beyond their response dates, and that all Halcyon payment applications through Application 22 have been processed and paid within contract terms.

Please provide a written recovery plan no later than 5:00 p.m. on Friday, February 6, identifying by week the manpower Halcyon will apply, the sequence Halcyon proposes, and the date by which the level 3 and level 4 electrical rough-in will be complete and inspected. Kestrel will make the additional released areas described in the attachment available to support that plan.

This letter is provided under Article 4 of the Subcontract and is not a notice of default.

Read that last line twice. Say what a letter is not. It costs you nothing and it keeps the temperature down while still creating the record.

Notice also the middle paragraph. Before you assert their failure, you affirmatively state your own performance, with references. That paragraph is doing two jobs: it is honest, and it removes their best defense before they think of it.

Rung 5 — cure notice. Now the tone changes, and every word tracks the contract.

Re: Northgate Outpatient Pavilion — Subcontract 16-A — NOTICE OF FAILURE TO PROSECUTE THE WORK

Pursuant to Article 9.1 of the Subcontract, Kestrel Construction Group hereby notifies Halcyon Electric that it has failed to supply sufficient properly skilled workers to prosecute the Work in accordance with the Project Schedule, as described in Kestrel's letter of January 29 and Halcyon's recovery plan of February 6, which Halcyon has not met.

Halcyon shall have seventy-two (72) hours from receipt of this notice to commence and thereafter diligently continue correction of this failure. If Halcyon fails to do so, Kestrel may, without prejudice to any other remedy, supplement Halcyon's forces or provide such labor and materials as may be necessary, and deduct the cost thereof, including reasonable overhead, from amounts due or to become due to Halcyon under the Subcontract.

A copy of this notice is provided to Halcyon's surety.

Notice to the surety (only where the subcontract is bonded). Send it early. Sureties dislike surprises and a surety brought in at week two of a problem has options that a surety brought in at week twelve does not.

Re: Payment and Performance Bond No. [bond number] — Halcyon Electric — Northgate Outpatient Pavilion

This letter is provided as notice of a potential claim under the referenced performance bond. Enclosed are Kestrel's letters of January 29 and February 12, the subcontractor's recovery plan of February 6, and manpower records for weeks 45 through 54. Kestrel has not declared a default and is continuing to work with the subcontractor toward recovery. Kestrel will provide the surety with updated information every two weeks and will provide immediate notice of any change in status.

That letter is deliberately mild, and it is one of the most valuable letters in construction. It preserves your bond claim, it puts a sophisticated third party with money at risk into the conversation, and it very frequently produces a phone call from the surety to the subcontractor's owner that solves the problem without anybody declaring anything.

⚠️ Safety alert. When a subcontractor is behind and knows you are watching, the temptation is to make it up with headcount, hours, and speed. That is precisely the condition that produced the scaffold near-miss on this job. In week 34, after the steel acceleration, a frame scaffold on the north elevation was modified overnight by a trade that had not erected it; a plank was lifted for conduit and not re-secured; a mason tender named Emiliano Serrano stepped onto it at 7:20 a.m. and went down onto the platform. Nobody was hurt, and the investigation found three failures — a stale competent-person inspection tag, a scaffold modified by a trade that did not erect it with no re-inspection, and a crew running behind under an unwritten "make it up" pressure. Finding number three is the one nobody wanted to write down. When you send a recovery-plan demand, say in the letter that recovery will be achieved without compromising safety requirements, and mean it. Then watch that trade's near-miss reports for the next month. Schedule pressure is a hazard exactly like an unguarded edge. Chapter 24 develops this as a system property rather than a rule.

19.5.5 Supplementation and backcharge

Supplementation means you bring in additional labor — your own forces or another subcontractor — to work alongside the deficient subcontractor and charge them the cost. Backcharge means you deduct a cost you incurred because of their failure from money you owe them.

Both are ordinary tools. Both are also the single most common source of subcontract disputes I see, and the reason is always the same: the general contractor did the work first and built the paperwork afterward.

The contractual prerequisites, in order:

1. A right in the subcontract. Find the clause. Read the exact conditions.
2. Notice, in the form the clause requires. Certified mail if it says certified mail. To the address in the subcontract, not the field trailer.
3. A genuine opportunity to cure, for the full period stated. Seventy-two hours means seventy-two hours, including the weekend if the clause does not exclude it.
4. Documented failure to cure at the end of that period — dated, specific, photographed.
5. Contemporaneous cost records for what you actually spent: signed time-and-material tickets by the day, invoices, rental agreements, delivery tickets.
6. A written notice of the backcharge, itemized, sent when it is incurred — not discovered on the final pay application.

💰 Money check: what an unsubstantiated backcharge is worth.

Kestrel supplements Halcyon with a temporary crew for three weeks: 6 electricians × 40 hours × 3 weeks = 720 labor hours at a supplemental-crew billing rate of $94.00 per hour.

• 720 MH × $94.00/MH = $67,680 direct
• Plus Kestrel's contractual 10% overhead on supplementation: $6,768
• Total backcharge asserted: $74,448

Now the scenarios.

Scenario	What Kestrel can prove	Realistic recovery
Full documentation — notice, cure period, daily signed tickets, photos	$74,448	$74,448
Notice given, cure period observed, but tickets unsigned and reconstructed from payroll	$74,448 asserted | ~$45,000–55,000 after negotiation
No written cure notice; supplementation started on a verbal	$74,448 asserted	$0, and Halcyon may have a claim against you for interfering with their work

A backcharge you cannot substantiate is not a backcharge. It is a gift to the other side, wrapped in your own letterhead. It also does something worse than lose you the money: it hands the subcontractor a grievance they will use to color every disputed item for the rest of the job.

The parallel is exact, and I want you to see it. CO #14 on this project — the MRI slab built on a verbal directive — cost Kestrel $43,650 of unrecovered cost on a $186,400 expenditure, because only $121,000 could be substantiated with contemporaneous records. Backcharges fail for precisely the same reason changes fail. The price is set by what you can document, not by what it cost you. That is theme 5 of this book and it does not care which direction the money is flowing.

19.5.6 When a subcontractor is actually failing

Default is not a surprise. It is a set of signals you either read or ignore, usually over four to eight weeks. Learn them.

Any two of those in the same month is a conversation. Any four is a notice to the surety.

What a takeover actually looks like. If you declare default and call the performance bond, the surety has options — it may finance the defaulted subcontractor, tender a replacement contractor, take over and complete the work itself, or pay you the cost of completion up to the penal sum. Which options it has, and in what order, depends on the bond form. What none of the options do is give you back your schedule. Expect thirty to sixty days between your default declaration and productive work by a replacement, and expect the replacement to be slower than the original for another two to four weeks while they learn a building somebody else started. The bond pays money. It does not pay days. Chapter 5 covers the bond mechanics; Case Study 2 walks a real default from the first warning sign to the final invoice.

🧩 Productive struggle. Your drywall subcontractor is three weeks behind. Your subcontract gives you the right to supplement their forces after 72 hours' written notice, and you have a supplemental crew available Monday at $94 per labor hour. Before you read on, spend three minutes and write down the three things you must have in hand before you send that notice, and the one question you must answer honestly about yourself first.

The answer

The three things.

(1) Proof of released work. A dated log showing which areas were released to them and when, demonstrating that they had the physical opportunity to perform. Without this, their defense writes itself.

(2) Contemporaneous documentation of the shortfall. Daily reports with crew counts by day, over enough weeks to establish a trend rather than a bad week, plus the manpower commitment they made at kickoff so there is a baseline to be short against.

(3) The notice clause itself, read carefully. To whom, by what delivery method, how many hours, what the notice must contain, and whether the cure period runs on calendar or business days. A notice emailed to the field foreman when the clause requires certified mail to the corporate office is not notice.

The question about yourself: did I cause this? Did I release the areas? Did I answer their RFIs? Did I approve their submittals? Did I pay them on time and release retention when earned? Did another trade block them? Did I change the sequence on them twice? If any answer is uncomfortable, you are not sending a cure notice this week — you are fixing your own side, in writing, and then starting the ladder from rung 1 with clean hands. A cure notice sent by a general contractor whose own performance is defective does not just fail. It converts a schedule problem into a claim, and the claim runs against you.

19.5.7 📋 Try it: read the curve and pick the trade

📋 Try it. It is Monday of week 49. You are the project manager. Below is the manpower record for weeks 41 through 48 for six trades on the Northgate interior build-out, together with each trade's remaining scope in labor hours and the schedule's required completion week. Assume a 40-hour work week and that crews will continue at their recent four-week average unless something changes.

Manpower: promised (P) vs. actual (A), weeks 41–48

Trade		W41	W42	W43	W44	W45	W46	W47	W48
Sightline — framing	P	12	14	16	16	16	14	12	10
	A	11	14	15	16	16	15	13	11
Halcyon — electrical rough	P	14	16	18	20	20	20	18	16
	A	12	13	13	14	13	12	12	11
Cardinal — HVAC & plumbing rough	P	16	18	20	20	18	18	16	14
	A	15	18	19	21	19	18	17	15
Vanguard — fire protection	P	6	8	8	8	8	6	6	6
	A	4	6	7	8	8	7	6	6
Thermaline — insulation	P	0	0	4	6	6	6	6	6
	A	0	0	3	5	6	6	5	5
Sightline — board, tape, finish	P	0	0	0	8	12	16	18	20
	A	0	0	0	6	9	12	14	15

Remaining scope and required dates, measured at the end of week 48

Trade	Remaining labor hours	Required complete	Weeks available
Sightline — framing	2,600	end of W54	6
Halcyon — electrical rough	5,600	end of W56	8
Cardinal — HVAC & plumbing rough	4,900	end of W56	8
Vanguard — fire protection	2,000	end of W57	9
Thermaline — insulation	1,700	end of W58	10
Sightline — board, tape, finish	11,800	end of W62	14

The interior sequence carries 15 working days of total float between "level 3 and 4 rough-in complete and inspected" and the point at which it becomes critical to the September 18, Year 2 substantial completion date. Kestrel's total daily exposure to slipping substantial completion is $10,650 per calendar day.

Answer four questions. (1) Which trade will cause the next critical-path problem? (2) When, and by how much? (3) What is the dollar exposure if nothing changes? (4) What is the correct escalation step this week — and what is the correct step for the other trade that looks bad in this data?

Worked answer

Step 1 — compute the required crew for each trade.

Required crew = remaining labor hours ÷ (weeks available × 40 hours per week).

Trade	Remaining MH	Weeks avail.	Required crew	Recent 4-wk actual (W45–48)	Weeks needed at actual	Projected finish	Variance
Sightline — framing	2,600	6	2,600 ÷ 240 = 10.8	(16+15+13+11)/4 = 13.8	2,600 ÷ 552 = 4.7	W52.7	1.3 wk early
Halcyon — elec. rough	5,600	8	5,600 ÷ 320 = 17.5	(13+12+12+11)/4 = 12.0	5,600 ÷ 480 = 11.7	W59.7	3.7 wk LATE
Cardinal — mech. rough	4,900	8	4,900 ÷ 320 = 15.3	(19+18+17+15)/4 = 17.3	4,900 ÷ 692 = 7.1	W55.1	0.9 wk early
Vanguard — fire prot.	2,000	9	2,000 ÷ 360 = 5.6	(8+7+6+6)/4 = 6.8	2,000 ÷ 272 = 7.4	W55.4	1.6 wk early
Thermaline — insul.	1,700	10	1,700 ÷ 400 = 4.3	(6+6+5+5)/4 = 5.5	1,700 ÷ 220 = 7.7	W55.7	2.3 wk early
Sightline — board	11,800	14	11,800 ÷ 560 = 21.1	(9+12+14+15)/4 = 12.5	11,800 ÷ 500 = 23.6	W71.6	9.6 wk LATE

(1) Which trade? Halcyon Electric. Sightline's board crew projects worse on paper, but that projection is meaningless as a standalone number, because the board crew cannot man up until walls are closed, and walls cannot close until Halcyon's rough-in is complete and inspected. Sightline's shortfall is a symptom; Halcyon's is the cause. Sightline's kickoff curve continues climbing to 26 in week 52, so the crew they need is within the capacity they already committed to — they just have nowhere to put it. Note also that Sightline's framing crew is running at 101 percent of plan, which is strong evidence the company is willing and able to staff this job.

(2) When and by how much? Halcyon needs 17.5 electricians and is running 12.0. At that rate their remaining 5,600 hours take 11.7 weeks instead of 8, finishing around week 60 against a required week 56 — a slip of about 3.7 weeks, or roughly 18.5 working days.

(3) Dollar exposure. The sequence carries 15 working days of float. An 18.5-working-day slip consumes all 15 and pushes approximately 3.5 working days onto the critical path, which is about 5 calendar days.

5 CD × $10,650/CD = $53,250

And that is the optimistic number. It assumes nothing else goes wrong for the remaining thirteen months, and it ignores the disruption cost to every trade downstream of Halcyon, which is real and which nobody has priced yet.

(4) The correct escalation step this week.

For Halcyon: not a cure notice. Rung 3 — a written notice of the schedule requirement with a demand for a written recovery plan within five business days, delivered per the subcontract's notice clause, copied to Yolanda Prieto as the firm's owner because the manpower decision is being made above Devlin's level. Before it goes out, confirm and state in the letter that Kestrel has released the areas, closed the RFIs, and paid on time. Rungs 1 and 2 have already occurred and are recorded in the daily reports and in the January 20 email, which is what entitles you to move to rung 3 now rather than starting over.

For Sightline's board crew: nothing in writing about their manpower. The correct action is a phone call telling Roddy Vance what you are doing about Halcyon, giving him a realistic released-area forecast by week so he can plan his own staffing, and asking what he needs in order to ramp to 22 once the walls open up. Then hold the released-area forecast you gave him. That call costs you ten minutes and is worth more than any letter you could write.

The meta-lesson. Two trades looked bad. One letter was warranted. Reading the dependency, not just the variance, is the entire skill.

19.6 Coordination: The Daily Craft of It

Escalation is the exception. Coordination is the job.

19.6.1 The coordination meeting

Every Wednesday at 1:00 p.m. in the Northgate trailer, ninety minutes, and it is the most important recurring event on the project. Chapter 26 covers meeting structure across the whole project; this is the field trades' meeting and it has its own rules.

Who attends. The foreman or general foreman of every trade working in the next six weeks — the person who assigns people, not the person who sells work. Their project manager may attend; the foreman must. Margo runs it. I attend and mostly keep my mouth shut. Dani takes the minutes and issues them within 24 hours.

The agenda. Fixed, every week, in this order:

The discipline that makes it work: commitments with names and dates. Not "we should get the third floor roughed in this month." "Devlin's crew will complete electrical rough-in in rooms 3C-09, -10, and -11 by Thursday, February 5, with six electricians." Name, scope, quantity, date.

And "we'll try" is not an answer. When somebody says "we'll try," Margo asks the same three questions every time, in the same order:

1. "What would have to be true for you to say yes?"
2. "Which of those is on me?"
3. "So what date can you commit to?"

That is not a rhetorical trick. It is a genuine offer to remove the obstacle, and about half the time the answer to question two is something Kestrel can fix in a day — a released area, a hoist window, an answered RFI, a decision. The other half of the time it produces an honest date instead of a polite lie, which is worth more than the optimistic date would have been.

🔍 Why this works. Three mechanisms, and they compound. First, a promise made in front of the other trades whose work depends on it carries a social cost that a promise made privately to the superintendent does not — the people who will be hurt by the miss are in the room. Second, writing "who, by when" forces the promiser to check their own capacity rather than agree in order to be agreeable; most missed commitments in construction are not deception, they are arithmetic nobody did. Third, reading last week's commitments aloud at the top of every meeting creates a short, reliable feedback loop, and feedback loops are what convert intention into behavior. The corollary is brutal and absolute: if you skip item 2 for two consecutive weeks, the whole discipline evaporates, because everyone learns that commitments are not checked. I have watched it die in fourteen days. This is the same mechanism the Last Planner System formalizes, which is Chapter 27.

What Margo actually said to Yolanda Prieto. I promised you this script in the hook. Here it is, close to verbatim, because the structure of it is teachable.

"Yolanda, it's Margo at Northgate. Devlin's doing everything he can and this isn't about him — I told him I was calling you. Here's the situation. Your kickoff curve had eighteen people in weeks forty-five through fifty-two. We've averaged twelve and a half. I've got your work released, your RFIs are closed, and Lorena's paid every application inside terms.

Here's what happens next if nothing changes, and I'd rather tell you than surprise you. In about two weeks I have to send Devlin a written schedule notice. That goes in your file with us, and Kestrel's prequal database has a long memory. I don't want to send it and I don't think you want to get it.

So: what would it take to get eight bodies on level three by Thursday? And what can I do for you? I can give you two more clean released zones so your crews never wait on us, and I'll ask Lorena to turn your January application in fourteen days instead of thirty if the manpower shows up."

Count what is in ninety seconds. A fact, not an accusation. An explicit statement that she has done her own job first. A consequence stated as information rather than a threat. A question instead of a demand. And an offer — two things Kestrel gives, in exchange for something Kestrel needs. That is what leverage sounds like when it is being used well.

💰 Money check: forty minutes versus three days.

Return to Zone 3C. Twenty-two workers, four crews, nothing to do. Suppose Margo does not intervene and the logjam runs three days. Use an illustrative fully burdened craft cost of $62 per labor hour for these interior trades — a blended composite; your market and your labor agreement will differ, sometimes by a lot.

Crew	People	MH over 3 days	Est. productive elsewhere	Hours lost
Thermaline — insulation	4	96	0% (fully idle)	96
Sightline — board	8	192	30%	134
Sightline — framing	6	144	50%	72
Halcyon — electrical	4	96	100% (they were working)	0
Total	22	528		302

302 MH × $62.00/MH = $18,724, call it $18,700 of craft labor burned.

Whose money is that? Not Kestrel's — at least not on the day it happens. It is Sightline's and Thermaline's. Which is exactly why an inexperienced project manager shrugs at it. So do the second calculation.

Sightline's Northgate subcontract is $2,100,000. At a 6 percent margin, their entire profit on this job is $126,000. The board and framing share of that three-day loss is 206 MH × $62 = $12,772 — about ten percent of everything Sightline expects to earn on Northgate, gone in three days, through no fault of theirs. They will come after it. They will come after it as a disruption claim, as thicker markup on the next change order, as a slower response when I need a favor in month eighteen, or as a quiet decision by their estimator to add a point to their number the next time Kestrel invites them to bid.

And Kestrel's own exposure. If those three days consume float and touch the critical path: 3 CD × $10,650/CD = $31,950. Add the real risk in the room — Roddy Vance taking eight board hangers to another job for three weeks, which costs Sightline almost nothing and costs Kestrel roughly fifteen working days of level-three interior progress. That one is the expensive one, and it does not appear on any cost report until it is far too late.

Margo's fix: five people × 40 minutes ≈ 3.3 MH of supervision at about $85/MH ≈ $280, plus twenty minutes of my time writing the email. Call it $310.

$310 against $18,700 to $50,000-plus. That ratio is the entire economic argument for a superintendent who walks the building every morning.

19.6.2 Trade sequencing and the clean handoff

The interior sequence you learned in Chapter 9 is now something you have to manage, area by area, twenty times over. This is the Northgate interior sequence for a typical partition zone.

  AREA-BY-AREA INTERIOR SEQUENCE  (one zone, ~1,800 SF)

  [1] Layout, track, stud framing .............. Sightline framing
  [2] In-wall rough: conduit, boxes, piping,
      medical gas, low voltage, fire alarm ..... Halcyon / Cardinal / Vanguard
  [3] In-wall blocking ......................... Kestrel self-perform carpentry
  [4] AHJ ROUGH-IN INSPECTION  <-- HARD GATE ... Petrosyan (building dept.)
  [5] Batt insulation .......................... Thermaline
  [6] Board one side, then second side ......... Sightline board
  [7] Tape, finish, prime ...................... Sightline finish
  [8] Overhead: duct, pipe, conduit, sprinkler . Cardinal / Halcyon / Vanguard
  [9] Ceiling grid ............................. Sightline acoustical
 [10] Above-ceiling inspection <-- HARD GATE ... Petrosyan
 [11] Ceiling tile, devices, trim, flooring,
      paint, casework, hardware ............... six more trades

  Two hard gates. Everything upstream of a gate is worthless
  until the gate opens. This is why Margo split the inspection.

The clean handoff. A handoff is clean when the trade receiving the area can start work the moment they walk in — nothing to move, nothing to fix, nothing to ask about. A handoff is dirty when the receiving trade spends the first day of their schedule doing the previous trade's cleanup and calling you about defects.

Dirty handoffs are the most under-managed cost on a commercial project because they never appear as a line item. They appear as everybody being slightly slower than planned, forever.

The fix is a handoff inspection, and it takes ten minutes:

1. The releasing trade's foreman, the receiving trade's foreman, and a Kestrel superintendent walk the area together.
2. The releasing foreman confirms his work is complete and his debris and material are gone.
3. The receiving foreman either accepts the area or identifies specifically what is missing.
4. Kestrel logs it: area, date, released by, received by, exceptions.
5. The exceptions have a name and a date. They are not "punch."

Do this and two things happen. The receiving trade cannot later claim they inherited a mess. And the releasing trade knows in advance that somebody is going to look, which changes how they finish. That second effect is worth more than the first.

🔄 Check your understanding. Why is a documented area handoff more valuable to a general contractor than to either of the two subcontractors involved?

Answer

Because the general contractor is the only party who will be blamed by both. When the receiving trade is late, they will say they inherited an incomplete area; when the releasing trade is backcharged, they will say the area was fine when they left it. Neither has a record, so the argument is unresolvable and lands on you — and unresolvable arguments get settled by splitting the difference, which means you pay half of a cost that belonged entirely to one of them. The handoff log converts an argument into a fact. It also does something quieter and more valuable: it fixes responsibility for the seam, and the seam is where the buildings in Chapter 16's scope-gap lesson come apart.

19.6.3 Trade stacking and congestion

Trade stacking is putting more crews into a space than the space was planned to hold. It happens for three reasons: you are accelerating, you are recovering, or you never really planned the area loading in the first place.

I need to be careful here, and I want you to see me being careful.

Productivity falls when you stack trades. That is not controversial — every superintendent knows it and every disruption claim asserts it. But the magnitude is genuinely uncertain, it varies by trade, by space, by crew, and by what specifically is interfering with what. You will see tables in estimating guides and claims literature that quote loss percentages for congestion, and the ranges they give commonly run from roughly ten percent for mild crowding to thirty percent or more for severe stacking. I have seen worse than that in a bad week and better than that in a well-run one.

I am not going to hand you a precise table, because a precise table would be a fabricated precision that nobody actually has, and because if you take a number out of a book into a negotiation, the first competent question you get will destroy it.

Here is what I will give you instead: the mechanism, and the honest way to price it on your own job.

The mechanism. Congestion costs production through five specific channels, and naming them is what lets you argue about them.

The honest way to price it: the measured mile. Instead of quoting a table, measure your own job. Find a period when the same crew did the same kind of work without the impact, and compare it to a period when they did it with the impact.

Northgate, Sightline's board hangers:

Productivity loss = (21.6 − 15.2) ÷ 21.6 = 29.6 percent, call it 30 percent.

Now price it against the remaining scope. Sightline has roughly 84,000 SF of board remaining on the affected floors.

• At the unimpacted rate: 84,000 ÷ 21.6 = 3,889 MH
• At the impacted rate: 84,000 ÷ 15.2 = 5,526 MH
• Difference: 1,637 MH × $62.00/MH = $101,494, call it $101,000

That is a defensible number, and notice why: it came from this job's own records rather than from a published table. It is also a number you can attack, and you should know how, because the other side will. The two periods must be genuinely comparable — same crew, same kind of work, same position on the learning curve, same weather, same material availability. If Zone 2C had taller walls, more penetrations, or a green crew, the comparison is contaminated and an experienced scheduling consultant will say so in about four minutes. Chapter 33 develops this properly.

Tie this back to the steel acceleration. When Kestrel spent $168,000 to recover 17 of the 23 days lost to the anchor-bolt submittal, we bought a second erection crew, premium Saturday time, and a resequenced enclosure by area. The dollar comparison at the time looked like this:

On those numbers acceleration saved $13,050, which is nearly a wash — the real driver was that Meridian's leased interim clinic space expires October 1, Year 2, and six days late is survivable while twenty-three is not.

But look at what the $168,000 estimate did not contain. It contained no line for the downstream disruption to the interior trades who got compressed into a shorter window. It contained no line for the deck-edge rework event. It contained no line for the near-miss spike in weeks 34 through 36 that produced the scaffold investigation. The acceleration estimate priced the acceleration and not its consequences, which is the single most common error in acceleration decisions, and I made it on this job. When Nadia Haddad asks you to price an acceleration, price the stacking too, and say out loud that the number is uncertain.

19.6.4 Interference and the duty to coordinate

A subcontractor delayed by another subcontractor almost never sues the other subcontractor. They come to you.

They come to you because you are the one they have a contract with, because you are the one who wrote the schedule, and because you are the one with the money. And they have a real argument, because in most standard subcontract arrangements the general contractor takes on an obligation — sometimes express, sometimes implied — to schedule and coordinate the work and to provide reasonable access.

⚖️ What the contract says. Three provisions decide who pays when Trade A blocks Trade B, and you need to find all three in your own form before you need them.

The coordination obligation. Standard subcontract forms generally obligate the general contractor to schedule and coordinate the trades and to make the site reasonably available. What that obligation guarantees is limited — it is generally read as a duty to coordinate reasonably and in good faith, not a warranty that no trade will ever be delayed by another. But a general contractor who actively interferes, who fails to sequence at all, or who repeatedly releases and un-releases areas is in different territory than one who ran a disciplined coordination process and still had a bad month.

The delay clause and its flow-down. Most subcontracts flow down the prime contract's delay provisions, and many contain a no-damage-for-delay clause: the subcontractor's sole remedy for delay is an extension of time, not money. Enforceability is highly jurisdiction-dependent, and courts in a number of states recognize exceptions — commonly for active interference, bad faith, delays not contemplated by the parties, or abandonment of the contract. A no-damage-for-delay clause is not a force field, and a general contractor who treats it as one will eventually meet a judge who disagrees.

The claims-against-other-subcontractors provision. Many subcontracts require a delayed subcontractor to look to the offending subcontractor for its damages, sometimes with an obligation on the general contractor to enforce the offending subcontract or to assign its rights. This clause is the reason it matters enormously that your handoff log, daily reports, and coordination minutes actually identify which trade caused what. Without that record, you cannot push the claim down; you can only absorb it.

All three vary by contract form and by jurisdiction. Read yours, and bring the specific language to counsel before you rely on it.

The practical management posture that follows is unglamorous and effective:

Document causation in real time. The daily report entry that says "Thermaline 4 workers idle in 3C — electrical rough-in incomplete, inspection not called" is worth more than any letter you can write four months later. It costs Dani thirty seconds. Contemporaneous records are worth roughly ten times reconstructed ones, and this is the daily, boring form of that principle.

Do not adjudicate it in the field. When Sofia Marchetti tells you Halcyon's conduit is in her duct space, your job is to solve the physical problem today and record who caused it, not to declare a winner. Declaring winners in the trailer at 7 a.m. is how you acquire two enemies instead of one problem.

⚠️ Safety alert: the multi-employer worksite. On a site with twenty subcontractors, more than one employer can be cited for the same hazard. OSHA's multi-employer citation policy recognizes four roles, and an employer can occupy more than one:

• The creating employer — whoever caused the hazardous condition.
• The exposing employer — whoever has employees exposed to it.
• The correcting employer — whoever was engaged to correct or maintain that condition.
• The controlling employer — whoever has general supervisory authority over the worksite, including the authority to correct hazards or require others to correct them.

As general contractor, you are ordinarily the controlling employer. The standard applied to a controlling employer is reasonable care, which is generally understood to be less than what is required of the employer whose own people are exposed — but it is not zero. Reasonable care generally means conducting periodic inspections at a frequency appropriate to the hazards present, having an effective system for identifying and correcting what you find, and actually enforcing compliance against subcontractors who do not correct.

The practical consequence: "my subcontractor did that work, not me" is not a defense. Your subcontract's safety clause, your right to remove a noncompliant worker, and your right to stop work all exist because you are already exposed; exercising them does not create liability you did not already have. The scaffold on the north elevation in week 34 is the textbook version of this — a scaffold erected by one trade, modified by a second, with no re-inspection, and a third trade's employee stepping onto it. Every one of those employers had a role, and so did Kestrel. Chapter 24 takes the doctrine and the program apart in full; Appendix F has the competent-person duties.

19.7 Quality Is Set by What You Accept the First Time

Three mechanisms, in order of leverage.

Mockups and benchmark installations. Build it once, look at it together, and agree what "acceptable" means before there are ten thousand square feet of it. On Northgate we mocked up a curtain-wall unit with its perimeter conditions, an exterior precast joint, an exam-room wall assembly with its lead shielding at the imaging suite, and a typical patient-corridor ceiling with its devices. Each mockup cost real money and each one saved a multiple of it. The precast joint mockup alone changed the sealant backer detail before 21,000 SF of panel arrived on site.

First-work inspection — "do one, and we all look at it." For every repetitive assembly, the first installation stops and gets inspected by the subcontractor's foreman, Kestrel, and where relevant the architect. This is the cheapest quality tool in construction and the most frequently skipped, because it feels like it slows the start. It does slow the start, by about half a day, and it routinely prevents the discovery of a systematic defect in room forty-one.

The punch process, which is a closeout activity and belongs to Chapter 23 and Chapter 40. I will say only this about it here: a punch list is evidence that your first two mechanisms failed. A long punch list is not thoroughness. It is a bill for inspections you did not do in month eleven.

Now the rule that actually governs subcontractor quality, and it is about you rather than about them:

Quality is set by what the general contractor accepts the first time.

The first time a subcontractor installs something below the standard and you let it go — because they are behind, because you are behind, because it is Friday, because arguing about it will cost you a day — you have just published the actual standard for the rest of the project. Every foreman on that crew now knows what Kestrel accepts. It will not go back up.

This works in the other direction too, and faster than you would expect. The first time you make a trade tear out and redo a poor installation, the rework costs somebody a day and the message reaches every foreman on the site by lunch. I have never regretted a first-week rejection. I have regretted a great many first-week accommodations.

🏗️ From the field. On a hospital job years ago I accepted a mechanical hanger detail in the first mechanical room that was not what the specification called for. It worked. It was safe. It was not what was drawn. Nine months later the commissioning agent flagged the same detail in eleven locations, the engineer of record would not accept it, and the correction cost the mechanical subcontractor about $30,000 in a building that was already occupied on two floors. Every dollar of it traced back to my shrug in month three. The subcontractor was furious, and he was right to be — I had told him it was fine, and then a document told him it was not.

🔄 Check your understanding. A subcontractor installs the first twelve of four hundred identical assemblies slightly below the specified standard. It works, it is safe, and rejecting it will cost you a day you do not have. What are you actually deciding?

Answer

You are deciding the standard for the remaining 388, and you are publishing that decision to every foreman on the site by lunch. That is the real transaction — not twelve assemblies but four hundred, plus the precedent that governs every other trade watching to see what Kestrel accepts. There is a second cost that is invisible on the day: acceptance in the field does not bind the architect, the engineer of record, or the commissioning agent, so you have not resolved the nonconformance — you have only deferred it to the most expensive possible moment, when the building is partly occupied and the correction is somebody's $30,000 problem. If the deviation is genuinely acceptable, the correct action is not a shrug; it is an RFI or a request for substitution, decided in writing by the party with authority to decide it.

19.8 Paying Subcontractors Is a Management Tool

I have watched more subcontractor performance problems get created in an accounting office than on a job site.

Here is the chain on Northgate. Subcontractors submit applications by the 20th. Lorena Vasquez and I review them against actual progress by the 23rd. Kestrel's application to Meridian goes in by the 25th. Meridian pays in 30 days. Retention is 10 percent until 50 percent complete, then 5 percent. Subcontractors are paid within a defined number of days after Kestrel receives payment, per the subcontract's payment clause. Chapter 32 works the whole mechanism, including the schedule of values, lien waivers, and the cash-flow curve.

The management question is not "how does the chain work." It is: what does your payment behavior teach your subcontractors?

19.8.1 Slow payment costs more than it saves

A general contractor who holds subcontractor money an extra thirty days is earning float. Let us price both sides honestly.

💰 Money check: the true cost of holding subcontractor money.

Suppose Kestrel systematically pays subcontractors 30 days later than it could. Across the interior trades, average outstanding payable subject to that delay is roughly $2,400,000.

What Kestrel gains: $2,400,000 held for 30 extra days. At a 7% annual cost of capital:

$2,400,000 × 7% × (30 ÷ 365) = $13,808

What Kestrel pays for it. Every one of those subcontractors finances the gap on a line of credit, and every one of them prices that into the next bid. If just three interior trades add 1 percent to a combined $9,000,000 of future Kestrel work to cover the aggravation, that is $90,000. If one drywall foreman moves his crew to a general contractor who pays on time, and Kestrel loses fifteen working days of interior progress, the schedule exposure at $10,650 per calendar day for the roughly 21 calendar days that represents is $223,650.

$13,808 earned. Six figures of exposure bought. This is not a close call, and I have never once seen it calculated by the people who make the decision.

19.8.2 Retention, joint checks, and the ethics of it

Retention exists to secure completion, and every subcontractor on your job is financing it out of their own working capital. On a $2,100,000 subcontract, 10 percent retention is $210,000 of Sightline's money sitting in Meridian's account. When the job passes 50 percent and retention drops to 5 percent, release the difference promptly. It costs Kestrel nothing — the reduction flows through from the prime contract — and failing to pass it through immediately is one of the fastest ways to convince a subcontractor that you are playing games with their money. Some jurisdictions regulate retention amounts and release timing by statute, and prompt-payment acts often set outside limits on how long you may hold subcontractor funds after being paid. Those rules vary substantially by state and by whether the owner is public or private. Verify yours.

Joint checks — a check made payable jointly to your subcontractor and one of their suppliers — are a legitimate tool when a lower tier is at risk of not being paid. Use them when you have a preliminary notice from a supplier and a reason to worry, and get a written joint-check agreement that says exactly what the payment covers and what it does not. Be careful: a joint check can, in some circumstances and some jurisdictions, be argued to create obligations or expectations you did not intend. It is a tool with an edge on it.

Now the ethics, and I am going to be blunt about three practices because every one of them is common and every one of them is wrong.

Withholding payment as leverage on an unrelated dispute. You are in a fight with Aperture Glazing over a $60,000 change order at the entry canopy. Their February application for completed, undisputed curtain-wall work is $340,000. Holding the $340,000 to get leverage on the $60,000 is not negotiation. It is using money you owe for work in place — money that flows through to their glaziers' paychecks and their supplier's invoice — as a weapon in an argument about something else. It may also violate your prompt-payment statute and your own subcontract. Withhold what is genuinely in dispute, itemized in writing, and pay everything else.

"We'll take care of you later." A subcontractor performs extra work on your verbal assurance that it will be handled at the end. Sometimes the person saying it means it. Often the person saying it will be on another job by then, and the file will contain no directive, no price, and no tickets. If you would not put it in an email today, you are not going to honor it in fourteen months. Say no, or write it down. CO #14 is what "we'll take care of it" looks like eight weeks later: $186,400 spent, $121,000 substantiated, $142,750 settled, $43,650 eaten — and Meridian's owner's representative had to live with it too.

Requiring uncompensated acceleration. Directing a subcontractor to add crews, work Saturdays, or double-shift to recover a delay they did not cause, and refusing to pay for it, is constructive acceleration, and it is one of the more reliable ways to convert a schedule problem into a claim you will lose. If the delay is theirs, the subcontract's recovery obligation applies and you should enforce it. If it is yours, or the owner's, or nobody's, and you want it recovered faster, that is a change and it costs money. The line is not subtle. Ask one question: who caused the delay I am asking them to absorb?

🔄 Check your understanding. Your electrical subcontractor is two weeks behind because you released three areas late. You need those two weeks back and you direct them to work the next four Saturdays. They comply and submit a change order for premium time. Do you owe it?

Answer

Almost certainly yes, and the reason is causation rather than generosity. The delay you are asking them to overcome was caused by your own late release of areas, not by their performance. Directing recovery for a delay you caused, without compensation, is constructive acceleration — you have effectively ordered a change and declined to pay for it. The right analysis has three steps: (1) establish who caused the delay, from the records, before you direct anything; (2) if it is you or the owner, price the acceleration as a change and — if the owner caused it — pass it through with notice; (3) if it is genuinely theirs, cite the subcontract's recovery obligation in writing and require the plan at their cost. What you must never do is direct the overtime first and decide the causation question afterward, because by then the only evidence of causation is whatever you happened to write down, and if you were the cause you probably did not write much.

19.9 The View from the Other Side of the Table

Everything up to here has been written from inside the general contractor's trailer. That is a distorted vantage point and I want to correct it, because the correction is the most useful thing in this chapter.

Two years ago I asked Sofia Marchetti, Cardinal Mechanical's project manager and the loudest voice in my coordination meeting, to write down what a good general contractor looks like from her trailer. She sent me a page and a half. I have used it in every new-hire orientation since. Here is the substance of it, in her voice.

1. Tell me the truth about the schedule. A general contractor who publishes dates everybody knows are fiction has taught me to ignore all of his dates, including the real ones. Then when a date genuinely matters, he has no way to tell me. I would rather have a schedule that says I start four weeks late than one that says I start on time and doesn't.

2. Give me clean released areas and don't take them back. I would rather have three floors sequentially than six floors half-available. Every time you release an area and then pull it back because another trade is not done, you have cost me a mobilization, a foreman's planning day, and about a day and a half of my crew's rhythm. I cannot bill you for any of that and you will never see it on a cost report.

3. Answer my RFI. Every day an RFI sits open, my detailer is either guessing or idle. If he guesses and he is wrong, we fabricate the wrong thing and both of us pay for it. Eleven days is not a response time. It is a decision to make me guess.

4. Pay me on time and don't play games with retention. I have about $1.9 million of my line of credit tied up in your job at peak. My cost of that money is real. When you hold my money to win an argument about something else, you are not being tough. You are borrowing from me at a rate I did not agree to.

5. Don't make me pay for your coordination failure. If you sequenced me into a ceiling where the duct cannot fit because the sprinkler main went in first, that is not a change order I should have to fight you for over six weeks. Own it in the meeting the day we find it.

6. Decide. The worst general contractor is not the mean one. It is the one who will not decide anything. A wrong decision on Tuesday costs me less than a right decision on the following Monday, almost every time.

7. Talk to me before you write to me. A written notice with no phone call first tells me you have already given up on the relationship and started building a file. I will respond accordingly, and both of us will spend the rest of the job with lawyers in our heads.

8. Know my business. I have 240 people and eleven active jobs. Yours is 14 percent of my year. I am not going to tell you it is my favorite job. I am going to tell you that when my best foreman comes free in March, he goes to whichever job is easiest to staff and least likely to make him look bad. Make that job yours.

Read item 8 again, because it is the economics of the entire chapter in four sentences.

19.9.1 The best subcontractors choose their general contractors

This is the part that new project managers do not believe until they see it.

In a healthy market, a good mechanical or electrical contractor turns down a large majority of the invitations it receives. It has finite estimating capacity, finite bonding capacity, and finite supervision, and it spends all three on the jobs it expects to make money on. The bid you receive is not a market price. It is a price adjusted for who you are.

I asked Sofia once what our reputation was worth to Cardinal in dollars. She thought about it and said: "On your jobs I don't carry a change-order fight allowance. On some general contractors I carry three percent."

Three percent of Cardinal's $6,400,000 Northgate subcontract is $192,000.

Now — she may have been being kind to a customer, and I would not put that sentence in a claim. But every estimator I know carries some version of that number, under some name, and nobody writes it on the bid form. It shows up as a slightly richer labor rate, a slightly fatter allowance, a slightly more conservative production factor. You will never see the line item. You will only ever see the total, and you will assume it is the market.

Here is the practical version, and it is the argument you use when your operations vice president asks why you are spending money on a coordination process:

None of that shows up on a cost report. All of it shows up in the bids you get and the crews you are assigned, which is where the money actually is.

🔄 Check your understanding. Kestrel and a competitor invite the same six drywall subcontractors to bid the same job. Kestrel's average bid comes back 3 percent lower. Nothing about the drawings, the schedule, or the specifications is different. What is the most likely explanation, and where does that 3 percent live on the bid form?

Answer

The most likely explanation is risk pricing based on the general contractor's reputation. Each bidder has priced the same scope but carried a different allowance for the friction they expect: how long RFIs will take, how clean the released areas will be, how hard it will be to get a legitimate change approved, how fast they will be paid, and how likely they are to be blamed for someone else's delay. It lives nowhere identifiable on the bid form — it is buried in the labor productivity factor, the general-conditions line, the contingency, or simply in the fee. That is why you cannot negotiate it out at buyout; it was baked in before the number was written, and the only way to remove it is to have earned a different reputation over years. Your operating behavior is priced into every bid you receive, permanently, and you never get to see the line item.

19.10 Building a Deeper Bench: Small, Emerging, and Diverse Subcontractors

Meridian Health System is a non-profit health system with a community-benefit commitment, and its CM at Risk agreement with Kestrel carries a voluntary 18 percent participation target for certified small, minority-owned, women-owned, and veteran-owned businesses, with quarterly reporting. It is a reporting requirement rather than a hard contractual one. On public work the requirements are usually harder: Rivermont Elementary School #12 carried a hard 12 percent requirement with a documented good-faith-effort standard, and the Willow Street Community Center will carry something similar. The specific programs, certification categories, thresholds, and good-faith-effort standards vary enormously by jurisdiction and by funding source, and they change. Read the ones that apply to your project; do not assume the last job's rules.

I want to skip the part where I tell you this is a worthy goal, which you already know, and go straight to the part nobody explains: why it is structurally hard, and what a general contractor can actually do that helps.

19.10.1 The real barriers

These are not attitude problems. They are capital problems, and they are largely mechanical.

Look at the pattern. Almost every barrier is a size test used as a proxy for a competence test. That proxy is convenient, it is defensible, and on a large complex package it is frequently correct. It is also the mechanism by which a bid list stays the same for twenty years.

19.10.2 What a general contractor can actually do

Five things, in rough order of effectiveness. Each has a cost, and I am going to state each cost honestly, because programs that pretend to be free get cancelled in the first bad quarter.

Unbundle packages. This is the highest-leverage move and it is purely a buyout decision. On Northgate, Kestrel split what would have been a single $560,000 doors, frames, and hardware package into a $340,000 hollow-metal-and-hardware package and a $220,000 wood-door-and-installation package. A nine-person firm bid and won the second one and could not have bid the first. Cost to Kestrel: one more subcontract to administer, one more seam to manage, and a scope-gap risk at the split that has to be written carefully into both scope sheets. That is real work, and it is Chapter 16's lesson applied deliberately.

Pay faster. Offer 14-day terms to firms below a revenue threshold, or pay twice a month. Cost to Kestrel: float, calculated exactly the way §19.8.1 calculates it — a real but small number. Benefit: it directly attacks the working-capital barrier, which is the barrier that actually kills firms.

Reduce or restructure the bond-back requirement. Options include lowering the bond-back threshold's applicability for smaller scopes, using a subcontractor default insurance program in place of individual bonds, requiring a funds-control or joint-check arrangement instead, or accepting a letter of credit. Cost to Kestrel: every one of these moves risk from the surety onto Kestrel's balance sheet, and Owen Baptiste, our chief financial officer, will want to know exactly how much. Be honest with yourself: this is not free, and it is the option most likely to hurt you if you do it carelessly.

Prequalify on the right things. Replace "three similar projects" with a demonstrated-capacity test: crew size and supervision available for this scope, references from work of any size, a site visit to their shop, an interview with the foreman who would run it. Cost to Kestrel: prequalification takes longer and requires judgment instead of a threshold.

Mentor and pair. Put a smaller firm on a shared scope with an established one, or bring them onto a smaller package first with the explicit understanding that it is a trial for a larger one. Cost to Kestrel: supervision time, and the honesty to say so when it does not work out.

19.10.3 The business case, stated without sentiment

I am going to make the argument in Kestrel's self-interest, because arguments that depend on virtue do not survive a bad year and arguments that depend on economics do.

A general contractor with four qualified drywall subcontractors in its database has less price leverage and less resilience than one with eleven. When one of the four defaults, or is booked, or is angry about last year's job, your options collapse and your price goes up. Every bid list you can legitimately lengthen is worth money on every future job.

The firms you build at $220,000 are the firms bidding your $2,200,000 packages in six years. Every established subcontractor on your bench was once a small firm somebody took a chance on. If nobody in your market does that, your bench ages out and does not get replaced, which is a real and observable problem in an industry with a well-documented workforce shortage.

And a bench you built is a bench that answers your call. The firm whose first commercial package came from Kestrel remembers it. That is not sentiment; it is the same reputational economics as §19.9.1, running in the other direction.

🏗️ From the field. The nine-person firm that took the Northgate wood-door package delivered the cleanest hardware installation on the job and closed their punch in eleven days, which no other trade came close to. They also needed help — Dani spent about four hours walking the firm's owner through our pay-application backup requirements and lien-waiver forms, twice, because nobody had ever explained the paperwork to him. Four hours. That is what the barrier actually looked like up close: not capability, not effort, but a forms process nobody had taught him and that we had assumed everyone knew.

🔄 Check your understanding. Almost every barrier facing a small subcontractor in the table above is a size test used as a proxy for a competence test. Name two situations in which that proxy is genuinely correct, and one specific substitute you would use when it is not.

Answer

The proxy is correct when scale itself is the risk. First, on a large, schedule-critical package — a $6,400,000 healthcare mechanical scope — where the firm must simultaneously field multiple crews, staff a detailing department, carry months of payroll, and absorb a bad month without failing. Size is not a proxy there; it is the requirement. Second, where the work carries a long warranty or performance obligation and the owner needs the firm to still exist in eight years. A substitute for the size test, where it does not apply: a demonstrated-capacity assessment — the specific crew and foreman who would run this scope, a shop visit, references from work of any size, and a look at their current backlog against their supervision bench. It takes judgment and about four more hours than reading a threshold, which is precisely why most companies keep using the threshold.

Spaced Review

Cover the answers. Try to produce each one from memory before you read on — the retrieval is the part that works.

1. From Chapter 16 — where do scope gaps live, and what is flow-down?

Recall first. — Scope gaps live between subcontracts, not inside them. Every subcontractor prices what its scope sheet says and nothing more, so the work nobody's scope sheet claimed — the sealant joint between the precast panel and the curtain wall, the blocking behind the wall-mounted equipment, the roof curb that the mechanical sub thought the roofer had and the roofer thought the mechanical sub had — belongs, by default, to you. Flow-down is the provision binding each subcontractor to the applicable terms of the prime contract, so the obligations Kestrel owes Meridian pass down the chain. Now connect it to this chapter: §19.6.2's handoff inspection and §19.6.4's claims-against-other-subcontractors clause are both field mechanisms for managing gaps you failed to close at buyout. The cheapest place to close a scope gap is a scope sheet. The second cheapest is a coordination meeting. The most expensive is a change order in month twenty.

2. From Chapter 18 — what is the superintendent actually doing when they plan site logistics?

Recall first. — Removing the reasons people stand still. Laydown, access, hoisting, gate hours, delivery windows, temporary power and light, and cleanup are not housekeeping. They are the physical infrastructure of production, and every one of them exists so that a crew that arrives at 7:00 a.m. is producing at 7:10. Now see it through this chapter's lens: that is leverage source number one from §19.2.1. A general contractor who runs excellent logistics is measurably more profitable to work for, which means better crews, better bids, and better cooperation — not because anybody is grateful, but because it is arithmetic on their side of the table too.

3. The deep callback — from Chapter 10: what sets the interior schedule?

Recall first. — MEP coordination, not structure. Above the ceiling is where the project is won or lost. Everything in this chapter is that threshold concept arriving in the field with names attached. Zone 3C stopped because of a hard inspection gate on electrical rough-in. Halcyon's manpower shortfall in the Try-it drill drives level 3 and 4 rough-in, which drives board, which drives ceilings, which drives finishes, which drives the September 18 substantial completion date. The three largest subcontracts on the job — Cardinal at $6,400,000, Ironbridge at $4,800,000, and Halcyon at $4,600,000 — include the two trades that will determine when this building opens. If you are going to spend your attention unevenly, and you are, spend it there.

Project Checkpoint: The Willow Street Subcontractor Management Plan

Your notebook already holds a buyout log with scope sheets for five trades (Ch 16), a preconstruction plan (Ch 17), and a site logistics plan with laydown, access, and lift planning (Ch 18). That logistics plan established the physical conditions under which your subcontractors will work. This deliverable establishes the management conditions.

Build a Subcontractor Management Plan for the Willow Street Community Center — $6,800,000, 24,000 SF, 425 calendar days, liquidated damages $1,200 per calendar day, prevailing wage, municipal owner. Source data is in Appendix K; blank forms are in Appendix D. Five parts.

Part 1 — Prequalification criteria and scoring. A one-page scoring sheet with weighted categories: financial capacity and bonding, safety record, relevant experience, current backlog and available supervision, reference results, and participation status under the City's MBE/WBE program. Assign weights totalling 100 and state your minimum passing score. Then add the paragraph that matters most: which criteria you will apply differently to packages under $250,000, and why — this is §19.10 made concrete on your own job.

Part 2 — Kickoff meeting agenda. Adapt the twelve-item agenda in §19.4.1 to Willow Street, and for each item state the artifact that leaves the room. Include, by name, the single person authorized to direct extra work. Attach the two-page expectation memo template that the meeting produces.

Part 3 — Manpower loading plan by trade by month. A table: trades down the left, the 14 contract months across the top, committed headcount in each cell, with a peak row and a total-labor-hours column. Cover at minimum concrete, masonry, structural steel, rough carpentry, roofing, mechanical, electrical, fire protection, drywall and framing, and finishes. Below the table, write your early-warning thresholds — green, yellow, orange, red — using a four-week rolling average, plus one line stating which trades are on your critical path and therefore get a tighter threshold.

Part 4 — Coordination meeting structure. Day, time, duration, required attendees by role, the fixed agenda, who takes minutes, and the distribution deadline. Write the two sentences describing how last week's commitments get read aloud and recorded. Include your area-handoff log format: area, date, released by, received by, exceptions with names and dates.

Part 5 — Written escalation procedure with sample notices. The six-rung ladder from §19.5.3, adapted to your subcontract's actual notice clause — recipient, delivery method, cure period, and required content. Then draft two letters in full for a hypothetical framing subcontractor running at 68 percent of committed manpower in months 6 through 8: a rung-3 schedule notice and a rung-5 cure notice. Each must affirmatively state your own performance — released areas, closed RFIs, payments within terms — before asserting theirs.

Next chapter goes inside the crews themselves. In Chapter 20 you will build the labor plan behind the manpower table you just made: crew composition, the man-hour budget, productivity targets, the union and open-shop distinction, and the certified-payroll process that prevailing wage requires on this job. The headcounts in Part 3 are about to acquire craft classifications, wage rates, and a burden factor.

Chapter Summary

A reference card, not a recap.

The threshold, restated. You do not manage the work. You manage the people who contracted to do the work. Your leverage is the subcontract, the schedule, and the coordination — not authority.

The four sources of leverage, in order.

The escalation ladder — never skip a rung.

  1. Field conversation with the foreman            (same week)
  2. Call the PM, then the firm's owner             (+1 week)
  3. WRITTEN notice of schedule requirement
     + demand for a recovery plan                   (+1-2 weeks)
  4. Recovery plan accepted/rejected IN WRITING     (5 business days)
  5. FORMAL CURE NOTICE per the subcontract
     (copy the surety if bonded)                    (after 3 & 4 fail)
  6. Supplement / backcharge / terminate            (after cure period)

  Skipping a rung does not save time.
  It voids the remedy at the end of the ladder.

Manpower thresholds (four-week rolling average against the committed curve): ≥95% green · 85–94% yellow, field conversation · 75–84% orange, written notice and recovery plan · <75% red, formal notice and evaluate supplementation. Any critical-path trade below 90% for three straight weeks is red regardless.

Before you send any notice, answer three questions about yourself. Did I release the work? Did I answer their questions? Did I pay them? If any answer is no, fix that first — in writing — and start the ladder over with clean hands.

Five things that are true about money on this chapter's subject.

• A backcharge you cannot substantiate is a gift to the other side. Notice, cure period, and contemporaneous tickets, in that order, or do not bother.
• The bond pays the money. It does not pay the schedule. Expect 30–60 days from default to productive replacement work, plus a relearning curve.
• Holding subcontractor money to earn float buys you thousands and costs you six figures.
• Trade-stacking loss is real and its magnitude is genuinely uncertain. Measure your own job's unimpacted period and compare — do not quote a table.
• Quality is set by what you accept the first time. A long punch list is a bill for the inspections you skipped in month eleven.

The three ethical lines, named. Withholding undisputed payment as leverage on an unrelated dispute. "We'll take care of you later" with nothing in writing. Directing uncompensated acceleration for a delay you or the owner caused. Each is common, each is wrong, and each costs more than it pays.

And the sentence to keep. Your job is to be the customer they want to perform for, and to have the contractual leverage for when they don't. Both halves, in that order.

What's Next

The manpower table in §19.5 treated crews as headcount — a number in a cell. Chapter 20 opens that cell up: who is actually in the crew, what each craft costs loaded, how productivity is measured and lost, what changes when the job is union versus open shop, and what prevailing wage and certified payroll require of you on public work. After that, Chapter 21 does the same for the machines, and Chapter 23 takes the mockup-and-first-work discipline from §19.7 and builds it into a full inspection and test plan.