Chapter 16 — Procurement and Buyout: Subcontracts, Purchase Orders, Long-Lead Items, and the Supply Chain

The Hook: $312,000 of Savings, and Two of Them Were Imaginary

Week three after notice to proceed. March 20, Year 1. The Northgate trailer still smelled like new carpet and the copier was not yet on the network, so Dani Okonkwo was walking paper down the hall like it was 1994.

They put a one-page summary on my desk and stood there with the specific stillness of a person who thinks they have brought good news.

"Buyout summary through today. Nine packages awarded. We're three hundred and twelve thousand under the estimate."

I looked at it for a while. Dani waited, and then stopped waiting, and then said, "That's — good? That's good, right?"

Here is the sheet:

Parentheses mean favorable — under the estimate. Half the cost of work bought in three weeks, and the number is green. On a bad day I would have signed it and gone to lunch.

Instead I asked Dani for two things: Solano Interior Systems' proposal for package 09-1, and Halcyon Electric's proposal for 26-1. Not the bid tab. The actual proposals, with the exclusion pages.

Solano's exclusion page, item 11: "In-wall blocking and backing by others."

Halcyon's exclusion page, item 6: "Disconnect switches and final connections to mechanical equipment by mechanical contractor."

Then I opened Cardinal Mechanical's proposal, which we had not yet awarded, and turned to their exclusion page. Item 4: "All electrical work, including power wiring, disconnects, and final connections, by electrical contractor."

I turned the monitor around so Dani could see both pages at once.

"Count the disconnects on that job for me. Rooftop units, air handlers, pumps, exhaust fans, chillers, the medical-gas compressors. Every piece of mechanical equipment that needs a local disconnect and a final connection."

They came back forty minutes later. Sixty-eight pieces of equipment. Twenty-two fire and smoke dampers needing power and control wiring.

"So who's buying that?"

Dani looked at the two exclusion pages. Then at the buyout sheet. Then at me.

"...Nobody."

"Nobody. Now do the blocking."

The blocking was worse, because it was quieter. Solano excluded in-wall blocking and backing — the plywood and 16-gauge steel strap that goes inside a metal stud wall so that later somebody can hang a grab bar, a headwall, a ceiling-lift track, a monitor arm, a fifty-pound wall cabinet. Kestrel self-performs rough carpentry, so a reasonable person would assume our own carpenters had it. I pulled Tomás Reyes' estimate detail for package 06-1. Our carpentry estimate covered wood blocking at roof curbs, at door frames in wood-framed openings, and at the café millwork. It did not cover in-wall backing in metal stud partitions, because the estimator who priced it had assumed — reasonably, in a different world — that the drywall contractor carried it.

Three thousand nine hundred linear feet of backing. Two hundred and fourteen heavy equipment backing plates.

So: of the $312,000 in "savings," the three headline items were drywall at $114,000 under, electrical at $82,000 under, and earthwork at $68,000 under. Two of those three were not savings. They were scope that nobody had bought, showing up on my report as profit.

Here is the sentence this chapter exists to put in your head, and I want it in there before you read another paragraph:

Buyout savings are only real if the scope is complete. A variance column that shows green because you failed to buy something is not a saving. It is a deferred loss in a nicer font, and it will find you in month seven, when the wall is already closed, the subcontractor is already on site, and you have no leverage at all.

🏃 Fast Track: If you have run a buyout before, skim §16.1 and §16.2 and go straight to §16.4 (scope sheets and the seam checklist), §16.6 (subcontract clauses, pay-if-paid, bonds versus subcontractor default insurance), and §16.8 (back-scheduling long leads). The Northgate steel back-schedule in §16.8 is the one to read even if you read nothing else.

🔬 Deep Dive: For the contract language behind flow-down, notice, indemnity, and pay-if-paid, go to Appendix G. For the buyout log, scope sheet, and award recommendation memo as blank forms you can use tomorrow, see Appendix D. For how the bought budget becomes the control budget, see Chapter 28.

16.1 What Buyout Actually Is (and What It Produces)

Buyout is the process of converting an estimate into a set of executed, enforceable commitments — subcontracts and purchase orders — and then reporting the difference between what you estimated and what you actually committed.

It runs roughly from award to about ninety days into the job. On a hard-bid lump-sum project it starts the day you are the apparent low bidder. On a CM at Risk job like Northgate, where the guaranteed maximum price (GMP) was built out of real subcontractor bids during preconstruction, buyout starts even earlier — you are mostly executing prices you already have, which is why nine packages could be awarded in three weeks. That speed is a feature and a trap. Fast execution of an incomplete scope is just a faster way to be wrong.

Buyout produces three things.

1. Executed agreements. Northgate's cost of work breaks into 32 bid packages. Two of those — cast-in-place concrete and rough carpentry — Kestrel self-performs, so they become internal work orders, not contracts. The other thirty resolve into 38 executed subcontracts, because several packages split in two (flooring became a resilient-and-carpet subcontract plus a tile-and-terrazzo subcontract; the Division 01 site-services package became three separate agreements), plus 26 purchase orders for material we buy directly.

2. A buyout log. This is the living document of the process. Not a spreadsheet you build once and email. A log, updated the day something moves.

3. The bought budget. When the last package is awarded, every line in your estimate has been replaced by a real number that a real company has signed for. That set of real numbers — not the estimate — is what you manage for the next eighteen months. In Chapter 28 we will call it the control budget. It is born here.

The Northgate buyout log

Here is the full log, reconciled to the canonical $40,000,000 cost of work inside the $47,500,000 GMP. Award dates are shown for the nine packages awarded as of the hook scene; the rest carry a status.

Two design decisions in that table are worth stealing.

Every package has a named human being. Not "PM." Not "the team." A name. Unowned packages are the ones that get awarded in month five with a scope sheet nobody wrote.

The status column has only four values — Bidding, Scope review, Awarded, Committed. The moment you let people type "in progress," you have lost the ability to see the job at a glance, and a buyout log you cannot read in ten seconds is a buyout log nobody reads.

🔄 Check your understanding. The log shows $40,000,000 of estimate, but the GMP is $47,500,000. What is the other $7,500,000, and why is none of it in the buyout log?

16.2 Packaging the Work: How Many Contracts, and Where Do You Cut?

Before you buy anything, you decide what "anything" is. A bid package is a bundle of scope you put out to a defined set of bidders and buy as a single agreement. You can cut the job four ways.

Northgate uses all four. Trade packaging for most of the building; a system package for the curtain wall (a delegated-design performance specification, not a set of shop tickets); a phase release for earthwork and foundations so we could move dirt in March against 90% documents; and area splits inside the drywall package so Solano could run two crews without tripping over themselves.

The trade-off, stated honestly

More packages give you more price competition, more granular cost visibility, and more ability to replace a failing subcontractor without blowing up the whole job. They also give you more seams. Every additional package is another boundary at which two proposals can each say "by others."

Fewer packages give you fewer seams and a single throat to choke. They also give you fewer bidders, less price transparency, and a subcontractor who has just been handed a great deal of leverage over your schedule.

Here is how I decide, and it is not a formula:

                      Is the scope a single, testable
                      performance system (envelope,
                      medical gas, fire alarm)?
                                 |
                +----------------+----------------+
               YES                                NO
                |                                  |
      Buy it as ONE package.          Is the trade's work physically
      Performance spec.               interleaved with 3+ other trades
      One warranty. One test.         above the ceiling or inside a wall?
                                                   |
                              +--------------------+-------------------+
                             YES                                       NO
                              |                                        |
                 Buy by TRADE, and spend the              Buy by TRADE, and split by
                 hours you saved on a brutal              AREA only if the manpower
                 scope sheet at the seams.                curve demands two crews.

The rule of thumb that has never failed me: the number of packages should be driven by how many bidders you can attract, not by how the specification is organized. If a package will draw five qualified bidders, it is sized correctly. If it will draw two, it is too big or too strange, and you should cut it differently. If it will draw fifteen tiny bidders who each need supervision, it is too small, and you are about to become the subcontractor.

⚖️ What the contract says. Packaging is not free of contractual consequence. Meridian Health System's CM-at-Risk agreement obligates Kestrel to obtain at least three bona fide bids for every package exceeding $250,000, to open them with the owner's representative present, and to give Pri Sethi the tabulation on request. That clause exists because on a cost-reimbursable GMP the owner ultimately pays the actual subcontract amounts — so the owner has a direct financial interest in how you package and how hard you compete the work. Read your own agreement's competition requirements before you draw the package list. Deciding to buy the whole MEP scope from one design-assist contractor can be an excellent strategy and a contract breach at the same time.

🔄 Check your understanding. You are considering splitting the drywall package into two subcontracts by area — Levels 1–2 and Levels 3–4 — because the manpower curve peaks at 34 workers and no single bidder can staff that. Name the new risk you have just created, and the one sentence you would put in both scope exhibits to control it.

16.3 Prequalification: Deciding Who Is Allowed to Bid

The cheapest way to avoid a subcontractor default is not to invite the bidder in the first place.

Prequalification is the structured evaluation of a subcontractor's capacity to perform this package on this project, done before bid day. It is not a credit check and it is not a vibe. It has gates and it has a score.

Step 1 — Pass/fail gates

Fail any one of these and the score does not matter:

Step 2 — The scored evaluation

Kestrel's thresholds: 75 and above approved; 60 to 74 approved with conditions (a project-specific dollar limit, joint checks to their suppliers, a required bond, or a higher retention); below 60 not approved for this package.

Reading a subcontractor's financial statement at a basic level

You do not need an accounting degree. You need five numbers and one question.

Here is Solano Interior Systems' most recent CPA-reviewed statement (Tier-3 illustrative, like every number in this book):

Working capital = current assets − current liabilities = $8,420,000 − $5,910,000 = $2,510,000.

What it means: this is the cash cushion Solano can burn while waiting to get paid. Against a $2,486,000 Northgate subcontract, their working capital is roughly 101% of the job — comfortable. A common screen is working capital of at least 10% of the largest single job the subcontractor will carry. Solano clears that by a mile.

Current ratio = $8,420,000 ÷ $5,910,000 = 1.42. Above about 1.3 is comfortable; below 1.1 means they are paying this month's bills with next month's draw.

Debt-to-equity = $7,240,000 ÷ $3,180,000 = 2.28 : 1. Above roughly 3:1 is thin for a subcontractor.

Backlog-to-working-capital = $24,700,000 ÷ $2,510,000 = 9.8 : 1. Around 10:1 is the edge of comfortable. Above 15:1, a single bad job takes the company down, and it may take your schedule with it.

Underbillings of $1,340,000 is the number I stare at longest. Underbilling means they have spent money they have not yet billed — unapproved change orders, work performed ahead of the billing cycle, or optimistic revenue recognition. It is the single best early warning of a subcontractor in trouble, and it is the number that will be understated if the statement is merely compiled rather than reviewed or audited.

And the one question: who prepared this, and at what level of assurance? A CPA audit is the highest; a review is middle; a compilation is essentially the contractor's own numbers typed on a CPA's letterhead. Ask. Note it on the scoring form.

Reading the safety numbers, including what EMR really means

The experience modification rate (EMR) is a multiplier applied to a contractor's workers' compensation premium, computed by a rating bureau from their claim history relative to what would be expected for a company of that size, in that class of work, in that state.

An EMR of 1.00 does not mean "good." It means "exactly average for your class and payroll." Below 1.00 means fewer or smaller losses than expected; above 1.00 means more. A 0.72 EMR is genuinely strong. A 1.35 EMR is a real signal.

Four honest cautions, because EMR is the most misused number in prequalification:

1. It is backward-looking, typically built from a three-year window that already lags a year. It tells you about the company two to four years ago.
2. It is volatile for small firms. A twelve-person contractor with one bad claim can carry a 1.6 for three years and be a perfectly safe operation today.
3. It is jurisdictional. Rating methodology and the governing bureau vary by state, and a contractor working in several states may have several modifiers.
4. It measures cost of claims, not exposure to hazard. Pair it with the recordable incident rate, the lost-time rate, and — the number I care about most — whether they can describe a near-miss they investigated and what they changed. Bea Salgado will not approve a subcontractor who cannot answer that question, regardless of the EMR.

The exclusion problem, stated honestly

A prequalification program with a hard EMR cutoff, a bonding requirement, and a three-similar-projects reference test will systematically exclude small firms, newer firms, and firms owned by people who have historically had less access to capital and bonding. That is not a side effect. That is what the program does.

It is also why many public owners run mentor-protégé programs, DBE (Disadvantaged Business Enterprise) participation goals, and bond-assistance or bond-waiver programs on public work, and why some private owners set participation targets of their own. Specific programs, definitions, and goals vary enormously by jurisdiction and funding source — federal-aid projects, state programs, and municipal programs each have their own rules — so verify what applies to your project rather than assuming.

The professional position, and mine, is this: prequalification exists to control real risk, so make each requirement do work. If you require a bond, ask what specific risk the bond is retiring, and whether a joint-check agreement plus a limited subcontract value would retire the same risk while letting a capable eleven-person firm onto the job. Willow Street, your project, is municipal and prevailing-wage, and it will almost certainly have participation goals attached. Design your package sizes so that a smaller firm can actually bid one.

🔄 Check your understanding. A drywall subcontractor scores 68 on your form. Their weak categories are financial capacity (working capital 6% of the proposed subcontract) and backlog versus capacity. Name three conditions you could attach that would let you award without carrying the default risk yourself.

16.4 Scope Sheets: The Seams Are Where the Money Lives

This is the core skill of the chapter. Everything else in buyout is administration. This is the part that makes or loses money.

A scope sheet is a written, package-specific list of exactly what is included in a subcontract, exactly what is excluded, and — critically — where this subcontractor's work stops and the next one's begins. It is developed before bids, issued with the bid package, refined in the scope review meeting, and attached to the executed subcontract as an exhibit.

💡 Aha moment. Estimators price scope. Subcontractors bid specification sections. Those are not the same thing, and the difference between them is where your profit goes.

The specification is organized by product and system, because that is how you file information. It is not organized by responsibility, because responsibility is a commercial question, not a technical one. Nowhere in a 900-page project manual does a section say "and also do the thing between your work and the next guy's." So it lands in nobody's number.

This is a gateway idea, and it is going to come back hard in Chapter 19: scope gaps live between subcontracts, not inside them. Before you learn it, you look for problems inside a package — did the drywall sub price enough board? After you learn it, you stop looking inside packages almost entirely and start walking the boundaries, because inside a package the subcontractor is expert and motivated and will not miss much, while at the boundary both parties are motivated to assume it is the other one's.

The seam checklist: twenty places money hides

Every one of these is scope that a competent bidder will reasonably assume belongs to somebody else. Put every one of them on every scope sheet, for every trade, and force a yes or no.

📊 Diagram (described): where the seam actually is. Picture a section cut through one Northgate exam-room wall, from the deck above to the floor slab. Reading top to bottom, the trades stack like this — and every horizontal line between two of them is a boundary somebody has to buy:

   ===== structural deck / fireproofing ====== (07-3 spray fireproofing)
        ^ head-of-wall joint ................. WHO? (09-1 or 07-4?)
   ||  top track and deflection detail  ||  .. (09-1 framing)
   ||  duct penetration + sleeve        ||  .. sleeve: 23-1 | firestop: 07-4
   ||  conduit penetration + sleeve     ||  .. sleeve: 26-1 | firestop: 07-4
   ||  in-wall backing for headwall     ||  .. WHO? (09-1 or 06-1?)
   ||  in-wall box + mud ring           ||  .. (26-1)
   ||  access panel in gyp wall         ||  .. furnish: 23-1 | install: 09-1
   ||  gypsum board, both faces         ||  .. (09-1)
   ||  acoustic sealant at base         ||  .. WHO? (09-1 or 09-4?)
   ===== floor slab / floor finish ========== (03-1 slab / 09-3 finish)

Four "WHO?" lines in one wall section. Multiply by 18,600 linear feet of partitions.

A complete scope sheet — package 09-1, metal stud framing and gypsum board

This is the artifact. Copy the structure; change the trade.

PACKAGE 09-1 — METAL STUD FRAMING AND GYPSUM BOARD Northgate Outpatient Pavilion | Scope Exhibit A to Subcontract | Rev. 3

Part 1 — Included in the base subcontract amount

Part 2 — Explicitly excluded, and where the work goes instead

Part 3 — Clarifications (the answers from the scope review call)

Part 4 — Schedule and manning obligations

Twenty inclusions, eight exclusions with a named home for each, five clarifications, four schedule obligations. It takes about four hours to write. It is worth roughly a hundred thousand dollars per package.

🧩 Productive struggle. Before you read the next subsection, try this. Here are three exclusion lines, one each from three adjacent Northgate proposals:

• 33-1 Site utilities (Delacroix Underground): "Sanitary, storm, and domestic water to a point five feet (5'-0") outside the building line. Work inside 5'-0" by others."
• 22-1 Plumbing (Cardinal Mechanical): "Under-slab plumbing as shown on Drawings P-101 through P-104. Site utilities by others."
• 21-1 Fire protection (Redline Fire Protection): "Sprinkler system from the riser flange. Fire service main and backflow assembly by others."

Drawing P-101 shows under-slab plumbing terminating at the building line, not at 5'-0" outside.

Give yourself four minutes. Write down every piece of work that is in nobody's number.

The scope-gap hunt, worked

Here is the actual exercise Dani and I ran on Northgate the week after the hook. Take three adjacent trades, put their proposals side by side, and walk every seam.

Then the second pass — the one people skip. Trace every item back into every base bid to see whether somebody already has it.

$389,000 surfaced. $98,000 of it was already bought. $291,000 was genuinely in nobody's number. Both halves of that result are money. Half the value of a scope-gap hunt is finding the scope you were about to buy twice, and if you only hunt for gaps you will pay for the access panels in two subcontracts and never know it.

🔍 Why this works — the mechanism behind scope gaps.

Scope gaps are not evidence that subcontractors are careless. They are the predictable output of the bidding process itself, and the mechanism has three parts.

First, the specification is filed by product, not by responsibility. MasterFormat is a filing system. Nothing in it assigns the work between two systems.

Second, the low bidder wins. At bid time, a bidder facing ambiguous scope has exactly two choices: include it and be higher, or exclude it and be lower. The bidder who includes ambiguous scope loses the job to the bidder who excludes it. Over enough bids, the market selects for exclusion. This is not cynicism; it is arithmetic.

Third, every bidder's exclusion language is written by their estimator, at 11 p.m., for a different project. Boilerplate exclusion pages travel from job to job. "By others" means "not by me," and it is silent about who others is.

Put those together and you get a system that reliably produces holes, and a set of documents in which every hole is technically disclosed. The scope sheet is the only instrument that closes them, because it is the only document written from the boundary's point of view rather than the trade's.

💰 Money check: the same gap, bought at two different times.

Take the blocking gap — $107,000 of in-wall backing. Here is what it costs if you find it on March 20, Year 1, versus month nine, after the walls are framed, roughed in, and boarded on one side.

2.6 times the cost, and the multiplier is not the worst part. At buyout you have five bidders and an unsigned contract. In month nine you have one subcontractor, mobilized, who knows there is no alternative. On a GMP job this comes out of the $1,320,000 construction contingency, and by month nine that contingency has other claimants — and every dollar of it you burn on a scope gap is a dollar not available for a real unknown, and 75 cents of it belonged to Meridian under the savings split.

⚠️ Safety alert: buyout decisions become field hazards.

Three of them, and they all trace back to this chapter.

One: manning you bought that the subcontractor does not have. If you award to a firm that must hire twenty-three people to man your peak, they will hire twenty-three people in one month, and a crew that is 40% new is the crew that gets hurt. Prequalification's backlog-versus-capacity score is a safety metric wearing a financial costume.

Two: an unassigned seam is an unassigned hazard. "Housekeeping" and "temporary protection" and "hoisting" are on the seam checklist for money reasons — and if nobody owns them, nobody does them. Slips, trips, and falls on a debris-covered deck are the daily reality of a job whose cleanup scope lived in nobody's subcontract.

Three, and this is the one that costs people. Recall the Northgate scaffold near-miss in week 34: a frame scaffold was partially modified overnight by a trade that did not erect it, a plank was lifted and not re-secured, and Milo Serrano stepped onto it at 7:20 a.m. Nobody was hurt. Bea Salgado stopped work on the elevation for the day.

That investigation found three failures, and the second one — a scaffold modified by a trade that did not erect it, with no re-inspection — is a scope sheet failure before it is a safety failure. Who owns the scaffold? Who may modify it? Who re-inspects, and under whose competent person? OSHA's scaffold requirements in 29 CFR 1926 Subpart L require a competent person and require re-inspection after alteration — but the regulation does not tell you which of your eleven subcontractors that person works for. Your subcontract does. Clarification C5 in the scope sheet above exists because of week 34. Write it into every scope sheet before you need it.

🔄 Check your understanding. Your painting subcontractor's proposal says "prime and finish all gypsum surfaces per the finish schedule." Your steel subcontractor's proposal says "shop primer only." The gym's exposed structural steel is scheduled to receive a finish coat. Where is the gap, and what does it cost you to find it after the steel is erected and the gym floor is installed?

16.5 Leveling, Negotiating, and Awarding

Bid leveling — the process of adjusting every proposal onto a common scope basis so that you are comparing like with like — you learned in Chapter 13. Buyout is where leveling stops being an estimating exercise and becomes a contract.

The post-bid scope review call

Thirty to sixty minutes per bidder, with the bidder's estimator and their proposed project manager on the phone, and a written agenda that is your scope sheet. You walk it line by line. Dani takes notes; the notes become the Part 3 clarifications above.

The three questions I ask every bidder, in this order:

1. "Walk me through your exclusion page and tell me who you think has each item." Not "do you have blocking." Make them name the other contractor. When they say "the carpenter," you have learned that they believe a carpenter exists on this job, and now you can check.
2. "What is your manpower peak and where are those people right now?" A number and a location. "We'll staff it" is not an answer.
3. "What is in your number that you think nobody else has, and what did you assume that you are not sure about?" This question, asked warmly, has saved me more money than any negotiation tactic I know. Estimators want to tell you what they worried about.

The line: clarifying scope versus bid shopping

We drew this line in Chapter 15 and in Chapter 13. I am going to draw it again, harder, because buyout is where it actually gets crossed.

Bid shopping is taking one subcontractor's price to a competitor after bids are in and using it to get the number beaten. Bid peddling is the same transaction initiated from the other side — the subcontractor who calls back and volunteers to beat a number they should not know. Both are legal in most private contexts and both are a slow way to destroy a company.

Two operating rules make the line easy to hold:

Never disclose another bidder's number, or a range that reveals it. Not to help someone, not to be nice, not to "keep them honest."

When you find a gap, reissue the package to everyone. If the scope changes, the competition resets. This one rule makes shopping structurally impossible, because you cannot use new information on one bidder without giving it to all of them.

🏗️ From the field. Curtis Boone bought the mechanical package on Rivermont Elementary School #12 by taking a $2,140,000 number to the fourth bidder and getting $1,975,000. He was proud of the $165,000. The subcontractor who took it was underfunded, manned the job with five people where the schedule needed eleven, and Curtis supplemented them starting in month five. He spent $287,000 doing it — a net loss of $122,000 on a transaction he counted as a win.

The part he did not put in the report: within a year, his bid coverage on mechanical packages went from five bidders to two. Nobody announced a boycott. Estimators simply stopped returning his calls, because a mechanical estimator spends forty hours on a bid and learns quickly whose jobs are a real competition and whose are an auction with extra steps. Two bidders is not a market. It is a negotiation you have already lost. The most expensive thing Curtis bought that year was the reputation.

Value alternatives from subcontractors

Bidders often propose alternatives — a different product, a different sequence, a different assembly. Take them seriously and evaluate them with the same discipline as any value-engineering item from Chapter 11.

Note the second and third rows. The most valuable subcontractor alternative on Northgate cost nothing and saved four days. Ask every bidder, "If you could change one thing about how this is sequenced, what would it be?" You will get a schedule improvement about a third of the time, and it is free.

The award recommendation memo

Every award above your delegation threshold gets a one-page memo. Not because a form is virtuous — because writing down why you chose someone forces you to have a reason, and because in eleven months, when that subcontractor is failing, somebody will ask what you knew.

AWARD RECOMMENDATION — PACKAGE 09-1 Metal Stud Framing and Gypsum Board | Northgate Outpatient Pavilion | March 19, Year 1

Basis for recommendation. Ferris Wall Systems is $145,000 low and scored 58 on prequalification, below Kestrel's 60-point threshold. Ferris employs 62 field personnel with 51 currently committed to four active projects. Northgate's drywall manpower curve peaks at 34 workers in weeks 52–63. Ferris would need to add 23 people or pull crews from committed work.

Here is the risk, priced. If Ferris mans at 20 instead of 34, the framing-and-board duration on the interior critical path extends from 11 weeks to approximately 19 weeks:

• Delay = 8 weeks = 56 calendar days
• Northgate's total daily exposure to slipping substantial completion = $10,650/CD ($5,150 extended general conditions + $5,500 liquidated damages)
• Exposure = 56 CD × $10,650/CD = $596,400

We would be risking $596,400 to save $145,000. That is a 4.1-to-1 bet against us, and it does not count the disruption to every trade that follows drywall. Solano's prequalification score of 84, its 101%-of-subcontract working capital, its 0.81 EMR, and its named foreman (available from week 48) support award at the higher number.

That is the whole memo. Six inches of paper. It is also the document that, in month eleven, proves the decision was made on analysis rather than on who took you to lunch.

16.6 The Subcontract: The Clauses That Decide Who Pays

The subcontract is where the estimate becomes a promise with teeth. Most standard forms — the AIA subcontract documents, ConsensusDocs subcontract forms, or your company's own — cover the same territory. What varies is the aggressiveness of the terms and, critically, what your state permits.

⚖️ What the contract says: flow-down.

A flow-down clause typically reads, in substance: the Subcontractor assumes toward the Contractor all obligations and responsibilities that the Contractor assumes toward the Owner, with respect to the Subcontractor's work. It is the single most powerful clause in the subcontract, and it is the one most often used carelessly.

What it does well: it makes the subcontractor's obligations match yours automatically, without you re-drafting the prime contract forty times. When Meridian's agreement requires 14-day submittal turnaround, infection-control procedures adjacent to the active clinic, and a 30-day notice for any claim, flow-down puts all of that on Ironbridge, Cardinal, Halcyon, and everyone else.

Three cautions:

1. Attach the prime contract. Physically. As an exhibit, or by making it available and saying in the subcontract that it was made available and where. A subcontractor cannot be bound to terms nobody showed them, and a judge asked to enforce a flowed-down term against a sub who never saw the document is not going to enjoy your argument.
2. Flow-down does not create rights, only obligations. If you want the sub to have your rights against the owner — say, the right to a time extension — you must say so. Many subcontracts flow down every burden and no benefit, which is legally effective and a good way to earn a reputation.
3. Flow-down cannot make the sub responsible for something the prime contract does not require. If your scope sheet demands more than the contract documents do, that is a change to the subcontract, not flow-down.

⚖️ What the contract says: notice.

Every construction contract has notice requirements, and they are unforgiving — a point we made in Chapter 5 and will make again in Chapter 33. In buyout, notice is a sequencing problem.

Meridian's agreement gives Kestrel 21 days from the event to give written notice of a claim for additional time or money. If Kestrel's subcontracts also give the subs 21 days, then a subcontractor who notices you on day 21 has left you zero days to notice the owner. You have just been made liable for their claim with no ability to pass it through.

The notice sandwich: the subcontractor's notice period must be meaningfully shorter than yours. Kestrel's Northgate subcontracts require notice within 7 days, giving 14 days to evaluate, package, and notice Meridian. Some contractors go to 3 or 5 days. Whatever number you pick, pick it deliberately, and check it against your prime contract on every job — the day you copy last job's subcontract onto a project with a 10-day owner notice period is the day you build a trap for yourself.

Pay-if-paid versus pay-when-paid

These two phrases look like synonyms and are not.

Pay-when-paid is a timing clause. It says the contractor will pay the subcontractor within some period after receiving payment from the owner. Courts in most jurisdictions read it as establishing a reasonable time for payment, not as excusing payment altogether. If the owner never pays, the contractor generally must still pay the subcontractor — eventually.

Pay-if-paid is a condition precedent. It says the subcontractor is paid only if the contractor is paid, and that the subcontractor bears the risk of the owner's non-payment. If it is enforceable and clearly written, an owner's insolvency can leave the subcontractor unpaid, permanently.

Enforceability varies significantly by state, and this is not a detail. Some states enforce clearly drafted pay-if-paid clauses as written. Some refuse to enforce them at all as against public policy. Some enforce them only if the condition-precedent language is unmistakably explicit. Some limit them by statute, and some interact with prompt-payment statutes, lien rights, or payment-bond claims in ways that make the clause much less powerful than it reads. Several states have changed their position in the last two decades. Do not carry your knowledge of one state's rule into another state, and do not carry this book's summary into any state. Ask your construction attorney what your project's governing law does with the clause, before you sign — and know that on a bonded public job, a subcontractor may have a payment-bond claim regardless of what your subcontract says.

The professional posture, separate from the legal one: a pay-if-paid clause allocates owner-credit risk to a subcontractor who has no relationship with the owner, no ability to evaluate the owner's finances, and no ability to price the risk. On a job with a solid institutional owner like Meridian, it costs you nothing to soften it, and it is worth real goodwill and often a real price reduction to say so at the scope review call. On a job with a thinly capitalized developer, the clause is doing something — and the subcontractor deserves to know that is why it is there.

Subcontractor bonds versus subcontractor default insurance

Two ways to protect yourself against a subcontractor failing to perform.

A subcontractor performance and payment bond is a three-party surety instrument: the sub is the principal, you are the obligee, and the surety guarantees performance up to a penal sum. Typical cost runs on the order of 1% to 3% of the subcontract value depending on the subcontractor's credit, size, and the surety — verify with the surety, not with a rule of thumb.

Subcontractor default insurance (SDI) is a first-party insurance policy the general contractor buys, covering all enrolled subcontracts, with a self-insured retention (deductible) per loss and usually a co-participation percentage above it.

💰 Money check: the Northgate decision. Kestrel carries a corporate SDI program: 0.75% of enrolled subcontract value, $500,000 self-insured retention per loss, 15% co-participation above the retention. Meridian's contract requires performance bonds on the structural steel and curtain wall packages regardless.

Now the other side of the ledger, because a savings number without its risk is exactly the kind of number this chapter is about. Suppose one subcontractor on a $2,500,000 package defaults and it costs $780,000 to complete their work.

• With a bond: the surety's obligation runs to the penal sum. Kestrel fights about timing, not about money.
• With SDI: Kestrel pays the first $500,000 (the retention), then 15% of the remaining $280,000 = $42,000. Kestrel's out-of-pocket is $542,000; the carrier pays $238,000.

$542,000 divided by $101,541 of annual premium savings is 5.3 years of savings, consumed by one event. That is the actual trade: SDI is cheaper and faster and more inclusive, and it makes you the underwriter. If your prequalification program is rigorous — genuinely rigorous, with financial statements you read and references you call — SDI is usually the better deal and it lets capable small firms onto your jobs. If your prequalification is a form somebody fills out, SDI is a way to convert a subcontractor's failure into your loss.

🔄 Check your understanding. Your subcontract says the subcontractor must give written notice of a claim within 7 days. Your prime contract gives you 21 days. A subcontractor sends you a properly detailed notice on day 6, and you sit on it for 20 days. What have you done to yourself?

16.7 Purchase Orders and Material Procurement

Not every commitment is a subcontract. Some are purchase orders, and the difference is not clerical.

When a PO is right instead of a subcontract

Use a subcontract when you are buying work — labor performed on your site, installed by someone else's people, under your schedule and your safety program. Use a purchase order when you are buying goods — material delivered to your gate, with no installation labor and no site presence.

The practical test: does this vendor's employee set foot on my job site to perform work? If yes, you want a subcontract, because you want the safety obligations, the insurance requirements, the indemnity, the schedule obligations, and the lien-waiver mechanics that a PO does not carry. A vendor who delivers 400 doors is a PO. A vendor who delivers 400 doors and hangs them is a subcontractor, even if their invoice says "material and installation."

Northgate's 26 purchase orders cover things like doors and frames as material, hollow-metal hardware, specialties, casework, appliances, the elevator cab finishes furnished separately, and the anchor bolts and embeds.

Goods versus services, and why the legal difference matters

Contracts for the sale of goods are governed in the United States by Article 2 of the Uniform Commercial Code (UCC), adopted in some form by every state except Louisiana, which has not adopted Article 2. Contracts for services are governed by common law and by construction-specific statutes. When an agreement mixes both, courts generally apply a predominant purpose test — is this mostly a sale of goods with incidental service, or mostly a service with incidental material?

Four practical consequences you will actually meet:

1. The battle of the forms. You send a purchase order with your terms. The supplier sends back an acknowledgment with their terms. Under UCC Article 2, a contract can form even though the forms do not match, and which terms govern is a genuinely messy question. The defense is simple: read the acknowledgment when it arrives, and if it contains terms you cannot live with, object in writing immediately. Most people file it.

2. Implied warranties. Article 2 supplies implied warranties of merchantability and, where applicable, fitness for a particular purpose — unless the supplier disclaims them, which most standard supplier terms do, conspicuously.

3. Limitation of remedies. This is the big one for construction. Nearly every equipment manufacturer's standard terms disclaim consequential damages and limit liability to the purchase price or to repair and replacement. Read that again in the context of a schedule.

If the switchgear arrives ten weeks late, and that delay costs you $74,550 a week in extended general conditions and liquidated damages, your contract remedy against the manufacturer is very likely: nothing. Not "reduced." Nothing. You may get the equipment repaired or replaced. You will not get your delay costs, because you agreed in the acknowledgment you filed that you would not.

This is the most important sentence in this section: you cannot buy schedule protection from a supplier. You can only buy it with float. Every hour you spend negotiating liquidated damages into a material purchase order is an hour you should have spent moving the order three weeks earlier.

4. Statute of limitations and the trigger for it differ between goods and services in many jurisdictions, and the difference can matter years later on a latent defect. Another one for your attorney, not for a rule of thumb.

Price holds, quote expiration, and escalation

A subcontractor's proposal and a supplier's quotation both have a shelf life, and the shelf life is on the document in small type: "Prices firm for 30 days from the date of this quotation." Or worse: "Prices in effect at time of shipment."

Three moves at buyout:

• Log the expiration date of every quote in the buyout log, in its own column. A package that expires before you award is a package you will re-bid at a worse number.
• Negotiate the price hold explicitly. "Firm through October 1, Year 1" is a term you can buy, sometimes for nothing, and it is worth more than a small discount.
• Where you cannot get a hold, use an escalation clause rather than making the vendor guess.

💰 Money check: what an escalation clause is worth. Ironbridge Steel's $3,795,000 subcontract is roughly 55% mill material — $2,087,250. Kestrel and Ironbridge agree to a shared escalation clause: Ironbridge absorbs the first 5% of mill-price movement on that material; increases above 5% are shared 50/50; the clause is symmetric, so decreases are shared the same way.

Suppose mill prices rise 12% before Ironbridge's material release.

• Total increase = 12% × $2,087,250 = $250,470
• Ironbridge's absorbed band = 5% × $2,087,250 = $104,363
• Shared amount = $250,470 − $104,363 = $146,107
• Kestrel's share = 50% × $146,107 = $73,054

Now the alternative. Without an escalation clause, Ironbridge must price the risk into a firm bid. To fully protect a 12% move, they carry $250,470 in their number — and Kestrel pays every dollar of it whether or not prices move.

The clause is worth $177,416 if steel goes up 12% ($250,470 − $73,054), and it costs $0 if prices hold. That is Theme 1 in one table: the party who cannot control a risk should not be forced to price it blindly, because blind pricing is always more expensive than shared pricing.

Two honest cautions. First, this is exactly what the GMP's $575,200 escalation allowance is for; when you draw on it, Pri Sethi will ask what index you used and why, so tie the clause to a published index, name it in the subcontract, and keep the published values with your records. Second, an asymmetric clause — escalation up but no sharing down — is the kind of term that gets you a reputation, and it is also the kind of term a sophisticated owner will strike.

Freight, storage, taxes, and the boring stuff that costs real money

That last row deserves a note, because it connects directly to Chapter 32. Owners will often pay for stored materials, but only against a specific documentation package: a paid invoice or bill of sale transferring title to the owner, evidence of insurance naming the owner for the stored value, segregated and marked storage with the owner's name on it, a right of access for inspection, and on bonded work sometimes a consent of surety. Off-site storage usually requires more of all of that than on-site.

If you buy $840,000 of curtain wall units in October and cannot bill for them until they are installed in February, you have just financed $840,000 of somebody else's product for four months. At an 8% cost of capital that is $22,400 of carrying cost, and more importantly it is $840,000 of working capital that is not available for payroll. Get the stored-materials procedure in writing at buyout, from both the owner and the supplier, or price the float.

🔄 Check your understanding. You issue a purchase order for 186 hollow-metal door frames with your standard terms. Three days later the supplier's order acknowledgment arrives with their terms: implied warranties disclaimed, liability limited to the purchase price, consequential damages excluded, and a broad force majeure clause. You file it. Six weeks later the frames arrive with the wrong hinge preps and the delay costs you eleven days on the door-and-hardware sequence. What is your position, and what should you have done on the day the acknowledgment arrived?

16.8 Long-Lead Items and Back-Scheduling

Here is the discipline that would have saved Northgate twenty-three days and $168,000.

A back-schedule starts at the date material must be physically on site and subtracts every step backward until you arrive at the date a human being must do something today. Not "order early." A date, on a log, with a name next to it.

📊 Diagram (described): the procurement chain, read right to left.

  ISSUE          KESTREL      A/E        [RESUBMIT?]    RELEASE     FABRICATE    SHIP    ON
  SUBMITTAL  ->  REVIEW   ->  REVIEW  -> [+ REVIEW ] -> TO MILL  -> / MANUF.  -> /     -> SITE
                                                          /FACTORY               DELIVER
   <---------------------------- you compute this direction ------------------------------|
                                                                            required-on-site
                                                                            date is the ONLY
                                                                            fixed point

Everything to the left of "on site" is a duration you must obtain from a real person — the fabricator, the manufacturer, the engineer — and verify in writing at buyout. Everything you assume, you will be wrong about.

Northgate long-lead item #1: structural steel — the one that got away

Ironbridge Steel's subcontract was executed March 7, Year 1. Steel erection was scheduled to start August 4, Year 1, and steel erection was on the critical path.

The gating item was not the steel. It was the anchor bolt and embed submittal, because anchor bolts have to be cast into footings months before a column arrives, and because Ironbridge's mill order release was conditioned on approval of that submittal — the column layout and base-plate details that drive the bolt plan also lock the column lengths and the advance bill of material.

Here is the back-schedule that existed. It was in the file. I wrote it.

Sixty-six calendar days from subcontract execution to bolts on the ground, and zero days of float anywhere in the chain.

Here is what actually happened.

Six days. Six days of a document sitting in an office.

And here is the number nobody at Kestrel had written down anywhere: Ironbridge's reserved mill rolling slot required release by April 18. Not April 15 — that was our approval date, which we treated as the milestone. April 18 was Hank Duffy's real constraint, and it lived in Hank's head and on Hank's production schedule and in exactly none of our documents.

Approval came April 21. Three days past the release date. The slot was gone, and the mill's next opening for those shapes was five weeks out — May 23.

Twenty-three calendar days on the critical path. The canonical economics, which you have seen before and will see again in Chapter 29:

Kestrel accelerated — not because the arithmetic demanded it (it saves only $13,050, which is nearly a wash) but because Meridian's leased interim clinic space expires October 1, Year 2, and six days late is survivable while twenty-three is not.

💰 Money check: what six days of desk time cost. Six calendar days of internal review sloth produced a 23-day critical-path delay and $168,000 of acceleration. That is $28,000 per day that the submittal sat on a desk. There is no trade on the job, no crew, no piece of equipment, that burns $28,000 a day. The most expensive thing on your project is paper that is not moving.

And it did not stop at money. The acceleration produced trade stacking, a rework event on deck-edge detailing, and a spike in near-misses in weeks 34 through 36 — including the scaffold event. A procurement failure in April became a safety problem in November. That is Theme 4, and it starts here.

Northgate long-lead item #2: the curtain wall

38,500 square feet of unitized aluminum-and-glass curtain wall, delegated design. Topping out is November 12, Year 1; the building must be dried in by March 28, Year 2. First panels are required on site November 20, Year 1 so installation can start November 25 and run through March 20, Year 2.

📋 Try it: back-schedule the curtain wall.

You are Dani, and it is week two after NTP. Vantage Facade Systems has given you these durations in writing. Compute the date the curtain wall subcontract must be executed so the fabricator can start shop drawings — assuming one review round. Then state what happens if the review takes two rounds.

Required on site: November 20, Year 1.

Work backward. Write your dates before you open the answer.

Northgate long-lead item #3: the electrical switchgear, and the rigging window

This one teaches a different lesson: sometimes the required-on-site date is set by physical access, not by the schedule.

Northgate's electrical service is 3,000 amps at 480/277 volts. The main switchgear is a permitted flatbed load that must be rigged into the level-1 main electrical room through the north loading opening. That opening closes when architectural precast panel P-C4 is set — February 10, Year 2. After that, the room has a wall in front of it.

Now run the two-round scenario, exactly as you did for the curtain wall. A resubmittal cycle adds 22 CD: release moves from August 2 to August 24, delivery moves from January 31 to February 22, Year 2 — twelve days after the rigging window closed.

I have watched a project spend $52,000 cutting a hole in a structural slab to get a piece of equipment into a room it could have walked into six months earlier. Nobody made a bad decision. Everybody made a small late decision, twelve times.

Supply-chain risk: what you actually have to manage

Four risks, and one honest statement.

Single-source components. Some things have one manufacturer, one plant, or one qualified supplier. Elevators, medical gas alarm panels, specialty imaging shielding, some switchgear configurations, and increasingly some electronic components. For every single-source item, ask two questions at buyout: what happens if this plant stops, and is there an approved-equal path, and how long does approving it take? If the answer is "there is no alternate and approval would take ten weeks," that item goes on the risk register from Chapter 6 with a named owner and a monthly check-in.

Allocation. When demand exceeds supply, manufacturers ration. Your order does not get later; it gets partially filled, or it gets filled behind a customer who buys more from them than you do. A purchase order is not a place in line unless the acknowledgment says it is. Ask.

Force majeure in supply contracts. Supplier terms usually contain a broad force majeure clause excusing delay for causes beyond their control — and, as we saw in §16.7, they usually also disclaim consequential damages. Together those two clauses mean the supplier's late delivery is very likely to be your problem, financially and contractually. Which is, again, why float is the only real remedy.

Logistics. Permitted loads, port congestion, rail capacity, driver availability, and the fact that a truck cannot always turn onto your street. Verify the delivery route and the offload plan for oversized items at buyout, not at the gate.

And the honest statement, which I want you to take more seriously than any specific number in this chapter: lead times move, sometimes violently, and they are not a property of the product — they are a property of the market on the day you ask. Items that took four weeks in one year have taken forty in another, and then come back down. Every duration in this chapter is illustrative. Verify every lead time in writing, from the actual supplier, at buyout, for your actual configuration — and re-verify before you release. A lead time you remember from your last job is not data. It is a story about a market that no longer exists.

🔄 Check your understanding. In the steel back-schedule, Caldwell Structural took its full 14-day review, exactly as planned. Was Caldwell the cause of the delay? What was?

16.9 Buyout Variance Reporting and the Bought Budget

Buyout ends with a report, and the report is where the temptation lives.

Ninety days into Northgate, Nadia Haddad wanted the buyout summary for the operations review. Here is the version I gave her. Note what is in it.

The gross number is $382,000. The real number is $27,000. On day 21 that same job looked like $312,000 of savings on nine packages.

Nadia read it and asked the right question, which was not "why is it only twenty-seven thousand." It was: "How confident are you that the two hundred ninety-one thousand is all of it?"

The honest answer was: reasonably, on the interior seams, because we ran the gap hunt on eleven adjacent-trade pairs. Less confident on the site-to-building interface, because two of those packages were still in scope review. So we booked a $95,000 scope-gap reserve inside the construction contingency, named it, and put a date on retiring it. That is what contingency is for — a priced, owned, drawn-down reserve for an identified risk, exactly as Chapter 6 framed it.

Four rules for reporting buyout honestly

1. Report gross variance and net variance separately, always. A single number invites the reader to believe the wrong one. Two numbers force the conversation.

2. A saving is not a saving until the scope sheet is closed. Kestrel's internal rule: a package's favorable variance cannot be reported as a saving until the scope review is complete, the exhibit is attached, and the PM has signed the seam checklist. Before that it is "awarded, scope open," and the estimate value is carried.

3. Never net a saving against an overrun without showing both. If earthwork came in $68,000 under and precast came in $28,000 over, the report says both. The moment you show a net of $40,000, you have destroyed the two pieces of information that were actually useful.

4. Name what a "saving" would have been spent on. On a GMP job this matters enormously. Unused GMP contingency splits 75% owner / 25% Kestrel. A $291,000 scope gap you find at buyout and fund honestly comes out of the buyout variance. The same gap discovered in month nine comes out of contingency at roughly 2.6 times the cost — and 75 cents of every one of those dollars was Meridian's money. Buyout discipline is not just self-protection; on a cost-reimbursable contract it is a fiduciary matter, and Pri Sethi is entitled to see it handled that way.

The bought budget becomes the control budget

When package 32-1 is awarded, the buyout log stops being a procurement document and becomes the foundation of cost control. Each awarded value gets loaded against a cost code. Each subcontract becomes a commitment — money you have promised, whether or not you have spent it. From that day forward the questions change:

That transition is the subject of Chapter 28, and the quality of the handoff is entirely determined by the quality of what you did in these ninety days. A buyout log with clean scope sheets and honest variances becomes a control budget you can forecast from. A buyout log full of unbought scope becomes a cost report that goes wrong slowly for eighteen months and then goes wrong all at once.

Spaced Review

Answer these before you read the responses. Cover the text below each question.

From Chapter 15 — qualifications and exclusions. In Chapter 15 you learned to write a qualifications page for your own proposal: the list of assumptions, clarifications, and exclusions that defines what your price actually covers. Now flip it around. When you are the one reading a subcontractor's qualifications page, what is the single most dangerous kind of statement on it?

The dangerous one is the passive exclusion with no named home: "by others," "by GC," "not included." An exclusion that names a party ("welding of embed plates by the steel erector") tells you where the work went. An exclusion that names nobody tells you only that this bidder does not have it — and the whole point of §16.4 is that the other bidders' pages say the same thing. A bidder's qualifications page is a map of the seams in your project. Read all five of them side by side, not one at a time.

From Chapter 13 — bid leveling and scope sheets. Recall the leveling process: adjust every bid onto a common scope basis before comparing. What changes about that process when you move from estimating to buyout?

The arithmetic is identical; the stakes and the finality change. At bid time, leveling protects the accuracy of your number. At buyout, the leveled scope becomes Exhibit A to an executed contract — a document a subcontractor's lawyer will read in month fourteen. That is why a scope sheet used at buyout must name the trade that carries each excluded item, not merely record that the bidder excluded it. Leveling that ends with "adjusted for parity" is estimating. Leveling that ends with "and item 6 is carried by package 26-1" is buyout.

Deep callback to Chapter 10 — MEP long-lead equipment. Chapter 10 argued that MEP coordination, not structure, sets the interior schedule. Name two pieces of MEP equipment on a project like Northgate whose procurement — not installation — can independently drive the schedule, and say what makes them different from a length of duct.

Air-handling units and the main electrical switchgear are the classic pair, and elevators belong on the list too. What makes them different from duct is that they are engineered-to-order: they do not exist until somebody approves a submittal and a factory builds one for your building. Duct is fabricated from sheet metal in a shop three weeks out. A 3,000-amp switchgear lineup is a manufacturing slot. That is why §16.8 back-schedules the switchgear from a rigging window six months before the equipment is needed electrically — the physical constraint arrives long before the functional one, and MEP procurement is where the interior schedule is actually won.

Project Checkpoint: The Willow Street Buyout Log

In Chapter 15 you assembled the complete bid package for the Willow Street Community Center and wrote a go/no-go recommendation. Assume you went, and assume you won: $6.8 million, 24,000 square feet, 425 calendar days, liquidated damages of $1,200 per calendar day, 5% retention, 100% payment and performance bonds, prevailing wage. Now buy it.

Deliverable — six parts. Add all of it to your Project Notebook.

1. The buyout log. Every package on Willow Street, with columns for CSI division, package number, scope description, estimate value, awarded value, variance, status, quote expiration date, and the named person responsible. Your estimate total from Chapter 13 must reconcile to the sum of the estimate column. Show the reconciliation.

2. Scope sheets for five trades — earthwork, concrete, masonry, mechanical, and electrical. Use the four-part structure from §16.4: Included, Excluded (with the package that carries each excluded item named), Clarifications, and Schedule and manning obligations. Run all twenty seam-checklist items against each trade and force a yes or no on every one. A scope sheet with a blank line is a scope sheet that does not exist.

3. A leveled bid tab for each of the five. Three bidders minimum per package. Show base bids, then the adjustments that put them on a common scope basis, then the leveled totals. Where a bidder's exclusion moves scope to another package, show it moving — the value must come out of one column and appear in another, and your tab must still foot.

4. Award recommendations with justification for all five, using the memo format in §16.5. Make at least one of the five an award to someone other than the low bidder, and price the risk you are buying down, the way the Ferris/Solano analysis did. Show the arithmetic: days at risk × daily exposure, against the dollars saved. Willow Street's exposure is $1,200 per calendar day in liquidated damages plus your own extended general conditions — compute your daily general-conditions rate from your Chapter 13 estimate and use the sum.

5. A long-lead procurement schedule, back-scheduled from required-on-site dates. At minimum: the structural steel and open-web joists, the rooftop units and their curbs, the electrical switchboard, the commercial kitchen equipment package including the grease interceptor, the gymnasium maple flooring system, and the folding partitions. For each, work backward through submittal preparation, your review, A/E review, a resubmittal cycle, fabrication or manufacture, and delivery — and state the date the subcontract or purchase order must be executed. Run at least one of them under both a one-round and a two-round review assumption, and state which one you are planning to.

6. At least six scope gaps between adjacent trades, each with the subcontract that will carry it and a dollar value. Willow Street is generous with these; look hard at the water-main relocation (site utility versus plumbing, and the five-foot rule), the commercial kitchen (who sets the equipment, who makes the final gas, water, and electrical connections, and who ties the hood suppression system into the fire alarm), the wood-framed second floor bearing on the first-floor steel and CMU (bearing plates, hangers, and shear connections — erector or framer?), blocking for gym wall padding and basketball backstops, RTU roof curbs (furnished by whom, set by whom, flashed by whom), and field painting of the exposed gym structure.

Next chapter, in Chapter 17, you will take these commitments and turn them into a mobilization: the permit matrix, the site logistics plan, the procurement log that operationalizes item 5 above, and the schedule for actually getting onto the site. Buyout tells you who is coming. Preconstruction planning tells you when they can start and where they will park.

Chapter Summary

The buyout decision framework — nine questions, in order.

The eight things worth remembering.

1. Buyout savings are only real if the scope is complete. A green variance column caused by unbought scope is a deferred loss.
2. Scope gaps live between subcontracts, not inside them. Walk the boundaries; the interiors take care of themselves.
3. The bidding process structurally produces gaps, because the low bidder wins and excluding ambiguous scope is how you become low.
4. A scope-gap hunt finds double-buys too. On Northgate it surfaced $389,000 and $98,000 of it was already bought.
5. Prequalification is a safety program. Manning capacity you buy is manning capacity that shows up — or twenty-three new hires in one month.
6. Never disclose another bidder's number; when scope changes, reissue to everyone. That one rule makes shopping structurally impossible.
7. You cannot buy schedule protection from a supplier. You can only buy it with float. Suppliers disclaim consequential damages; float does not.
8. Back-schedule from the required-on-site date, and track the fabricator's release date, not your approval date. They are different numbers, and only one of them is real.

Numbers from this chapter worth carrying. Northgate cost of work $40,000,000 across 32 packages, 38 subcontracts, 26 purchase orders. Week-three apparent savings $312,000; genuinely unbought scope found at buyout $291,000; net real buyout savings at day 90 $27,000. Total daily exposure to slipping substantial completion $10,650/CD. The steel chain: 6 days of internal review slip → 3 days past the mill release date → 35-day mill push → 23-day erection delay → $168,000 of acceleration, or $28,000 per day that a submittal sat on a desk.

What's Next

You now have commitments. Thirty-eight subcontracts and twenty-six purchase orders' worth of people who have agreed to show up, and a procurement schedule that says when their material arrives. Chapter 17 turns commitments into a mobilization: the permit matrix and the agencies behind it, the site logistics plan that decides where a crane sits and where a truck turns around, the procurement log that operationalizes every back-schedule you just built, and the mobilization schedule that gets a trailer, a fence, and a temporary power service onto a sloping 6.2-acre site next to a clinic that never closes.

Then, in Chapter 19, you will meet the other half of the sentence this chapter kept repeating. Scope gaps live between subcontracts — and so does everything else. You do not manage the work. You manage the people who contracted to do the work, and the only leverage you have is the document you just spent ninety days writing.