Case Study 37-1 — The Six Weeks Colton Spent Solving the Wrong Problem

Harbor Ridge, Tessa Bright Homes, Coastline Drywall, and everyone in this case study are Tier-3 illustrative composites. The numbers are internally consistent and realistic; they are not a real community. Inspection sequences, contractor licensing, lien rights, and warranty obligations vary substantially by state and locality — nothing here is legal advice.

Setup

The line. Harbor Ridge is thirty-four lots. Colton Reyes runs eleven houses in process on a ninety-two-calendar-day cycle. That pair of numbers is the entire business:

Throughput = WIP ÷ cycle time = 11 ÷ 92 = 0.1196 houses per calendar day → 43.64 houses per year Start interval = 92 ÷ 11 = 8.36 calendar days

Every 8.36 days a finished house has to come off the end of the line and a new one has to go on the front. A blended Harbor Ridge house sells for $438,000, costs $374,500, and earns $63,500 of gross margin.

The trade. Drywall is Coastline Drywall — owner Reggie Prosser, crew lead Ines Duarte. One six-person crew on Colton's line, budgeted at six days per house for hang, tape, finish, and texture at $1.42 per square foot across about 6,800 square feet of board: $9,656 a house. Coastline hangs for other builders too. It is not a subcontractor on a project; it is running its own line, and Tessa Bright is one input to it.

At six days a house, one crew serves 365 ÷ 6 = 60.8 houses a year against the 43.64 Colton needs. Thirty-nine percent headroom. For two years, drywall was the last trade anybody worried about.

What happens

Part 1 — Four mornings that cost Colton his rate

Over nine weeks, four times, Coastline's crew drove to a Tessa Bright house that was not ready to hang.

Not one of the four is a drywall failure. Every one of them is a builder's readiness failure.

Now price them from Ines's side of the truck. Six hangers × 8 hours × $34 per hour loaded = $1,632 of payroll that produced nothing. Four trips: $6,528. Coastline's gross margin on a Tessa Bright house is about $1,352 on $9,656 of revenue, so four wasted mornings consumed the margin on nearly five houses.

Reggie Prosser did not fire Tessa Bright and he did not raise his price. He did what the rational owner of a small business does when a customer becomes unreliable: he stopped betting his crew's day on that customer. Beginning the first week of January he ran the crew across two builders at once, so there was always a second house to fall back on when a Tessa Bright house was not ready.

Elapsed time on a Harbor Ridge house went from six days to nine.

Part 2 — Six weeks of the wrong conversation

Colton is a good superintendent, and this is the part of the story he tells on himself.

He saw nine days and read it as a performance problem. Over six weeks he called Reggie twice and asked him to "get back to six," offered a per-house premium, priced a second drywall subcontractor without pulling the trigger, and told his division manager the drywall sub had gotten soft. None of it worked, because none of it touched the mechanism. Money does not buy capacity: a crew running across two builders takes nine days a house whether you pay $1.42 a foot or $1.60.

Then, on a Tuesday in February, he called me.

Colton: "I've got three houses insulated, inspected, locked, and empty. All at the same stage. Just sitting there."

Me: "How long has that row been there?"

Colton: "It was one house in January. It's three now."

Me: "That's not three problems, Colton. That's one problem drawn to scale. You're looking at a queue, and a queue means you've found your constraint."

A row of houses standing at the same stage is never a coincidence and it is never bad luck. It is your constraint, made visible on the ground, at full size, for free. Learn to see it and you will diagnose the next one in a morning instead of six weeks.

Part 3 — What the ninth day actually costs

Here is the arithmetic Colton had not done, and it is the arithmetic that ends the argument.

At six days a house, drywall had capacity to spare, so a day added to that stage was simply a day added to the cycle — worth about $29,800 a year in gross margin, the standard price of a day on this line. That stays true right up until the crew's capacity crosses below the rate the line requires. At that moment the stage stops being a duration and becomes a ceiling, and the cost of the next day looks nothing like the cost of the last one.

Read the last column twice. The seventh day cost $29,800. The eighth cost $29,200. The ninth cost $137,000 — because it was the day the crew's capacity fell under the line's requirement, and from that point forward annual output is no longer 11 ÷ 92 × 365. It is 365 ÷ 9.

The revenue effect is annual, not one-time. 3.09 houses a year × $438,000 = $1,352,000 of revenue and 3.09 × $63,500 = $196,000 of gross margin — every year, until somebody takes the constraint back out of the line. On Northgate a lost day costs $10,650 once. Here a broken rate bills you annually.

And the cycle time is 99 days, not 95. Little's Law running backward: WIP is still 11, throughput is now 40.56 ÷ 365 = 0.1111 houses a day, so cycle time = 11 ÷ 0.1111 = 99 calendar days. Those extra seven days are the queue — Colton's three locked houses, converted into arithmetic. Guess "6 to 9 days, so 92 to 95," and you are wrong twice: wrong about the cycle, and wrong about the loss by more than a factor of two.

The six weeks Colton spent on the wrong problem carry their own invoice:

42 days at the broken rate = 3.09 houses/yr × (42 ÷ 365) = 0.36 houses 0.36 houses × $63,500 = $22,500 of gross margin, spent on a diagnosis

Roughly a third of one house's entire margin, burned finding out that the problem was in his own truck.

Part 4 — Three responses, priced

Colton's division manager wanted a fourth option: start three more houses and make the volume back up. Reject it. Output is capped at 40.56 by drywall's capacity, so raising WIP from 11 to 14 raises nothing except cycle time:

Cycle time = 14 ÷ 0.1111 = 126 calendar days

Three more houses of unfunded inventory — roughly 3 × $45,000 = $135,000 of additional working capital standing in the mud — and every house in the community takes thirty-four days longer than the plan. Starting houses feels like production. Finishing houses is production.

Part 5 — What they actually did

B first, A as a bridge. The ready-check went in the following Monday — it cost almost nothing and it was the only response aimed at the cause. Option A ran alongside it for about ten houses (10 × $1,088 ≈ $10,900) to hold the rate while readiness earned Coastline's trust back. First-year cost of the pair: roughly $38,800, against $196,000 a year recovered and kept.

The ready-check is one page and it lives in a truck.

Row 9 is the whole document. The ready call is an affirmative commitment with a photograph attached, not the absence of bad news. Anybody can fail to mention a problem; nobody sends a photo of a floor covered in framing scrap.

Then Colton drove to Coastline's shop with the four lot numbers on a page and told Reggie what had happened and what he had changed. That hour appears on no cost report and was the most valuable thing he did all quarter. By June the crew was back on Harbor Ridge full time and drywall was running six days again.

Analysis — the mechanism, step by step

The builder was the constraint the whole time. Colton spent six weeks managing a subcontractor's performance, and the subcontractor's performance was a response to his own. Every link in the chain is individually rational:

Builder's houses are not ready when the crew arrives
        |
        v
Coastline eats $1,632 of payroll per wasted trip
   -- four trips consume the margin on nearly five houses
        |
        v
Reggie protects his crew by always keeping a fallback house at another builder
        |
        v
Elapsed drywall time on a Harbor Ridge house:  6 days  -->  9 days
        |
        v
Drywall capacity 40.6/yr falls BELOW the line's required 43.64/yr
        |
        v
Drywall becomes the ceiling.   Output = 365/9, not 11/92 x 365.
        |
        v
$196,000 a year of gross margin, and a row of locked houses on the ground

Nobody in that chain did anything unreasonable. That is exactly what makes it a systems failure rather than a story about a bad subcontractor, and it is why the fix had to be applied at the top of the chain rather than the bottom.

Why money was never going to work. Capacity was the missing thing, and capacity is the one thing a check does not buy. A premium can move you up the order in which Coastline serves its builders — roughly 1.2 houses a year here — but a crew running across two builders takes nine days a house at any price. Worse, paying for the symptom leaves the cause running, so the wasted trips continue and so does the logic that produced the split. You have bought a more expensive version of the same bottleneck.

Why this is a rate problem and not a delay. A delay is a fixed quantity you can recover with overtime. A rate mismatch compounds and never closes. Drywall at nine days against a line at 8.36 diverges by 0.64 days per house: house 4 waits 1.9 days, house 10 waits 5.7, house 22 waits 13.4, house 34 waits 21.0. No CPM network can draw that, because CPM answers which chain of dependent work is longest and the governing question here is which trade's rate is lowest. That is the whole of §37.3.1, and it is why Chapter 14's technique, correct for Northgate, is close to useless at Harbor Ridge.

The asset Colton actually bought. Option B beat option A by $19,600 a year on this trade alone, and that badly understates it, because the ready-check protects every trade on the line. The next Reggie Prosser is a trim carpenter or a plumber, and he is doing the same arithmetic in his truck this morning. In production housing, being the builder whose houses are ready when the crew shows up is worth more than price and more than any contract you could write — and it is the only line on that table that gets cheaper the longer you run it.

Discussion questions

1. Colton's four failed responses all treat drywall as a performance problem. For each one, name the single piece of evidence that should have ruled it out, and say where in his own records he could have found it in week one.
2. Reggie split the crew rather than raising his price or resigning the account. Argue that splitting was Coastline's best available decision; then argue that a phone call to Colton in November would have been better for Coastline. Which do you believe, and what does that say about how subcontractors escalate problems with a builder they do not want to lose?
3. Option B pays back better than option A and is also slower and less certain. Colton ran both. When would you run A alone? When would running A alone actively make things worse?
4. The marginal-day table shows the eighth day costing $29,200 and the ninth costing $137,000. Explain that difference to a division manager, in under 100 words, without using the phrase "Little's Law."
5. The instinct to start three more houses is committed by competent people roughly once per housing cycle. What makes it so persuasive? Name the single number that, posted weekly, would make it obviously wrong to everyone in the room.

Your turn

Build the ready-check for the trade after drywall, using Harbor Ridge's stage sequence in §37.4.

1. Pick a trade whose start depends on somebody else finishing: interior trim (cycle day 59), cabinets (day 64), paint (day 68), or flooring (day 73).
2. Write eight to ten checks, each one physically verifiable by a person standing in the house. "Coordination complete" is not a check. "Cabinet blocking installed at every upper run and confirmed against the elevations" is.
3. Assign each check an owner and a lead time — 24, 48, or 72 hours.
4. Write the ready call itself: what gets sent, to whom, and with what attached.
5. Finally, compute what one extra day in that stage is worth on Colton's line at 43.64 houses a year, and then what it is worth if that trade's crew capacity falls below 43.64 houses a year. Put both numbers at the top of your one page, where the superintendent will see them every morning.

Carry the page into your Willow Street notebook alongside this chapter's contrast memo. It is the residential form of Chapter 27's make-ready screen, and it fits on a clipboard.