Everything in this book so far has been questions. This chapter is the answer key.
The forty-eight are in the file. The beliefs are on the record, dated, in your own words. The stated numbers are flagged to verify. Your voice is captured. Now we do the thing I promised in Chapter 1: we open your books and we read your year the way a jeweler reads a watch with the back off — job by job, no exceptions, no anecdotes.
I want to be precise about what "open the books" means, because it does not mean what your accountant does. Your accountant reads your books top-down: revenue, cost buckets, a P&L that melts twelve months of jobs into one number per line. That document is true and nearly useless for our purposes, because everything you need to know about your company lives in the differences between jobs, and the P&L's whole job is to average the differences away. A great year and a quietly bleeding year can produce the same P&L. I've owned both.
We read bottom-up. Every job in the year gets pulled out and made to testify for itself: what it sold for, what it really cost, how long it held a crew, what went wrong, what it returned. Then the jobs get scored, ranked, and rolled up into verdicts about the lines. The scoring system has a name — we call it the ECS, the Economic Contribution Score — but the name matters less than the question it forces every single job to answer, which is not the question this industry has ever asked.
The industry asks: did the job close, and what was the margin?
The ECS asks: did the job make you money for the time it consumed?
Those sound like the same question. The entire fortune you've been leaving on the table lives in the difference, so let me pull it apart with the number that runs this whole system.
Gross profit per crew-day. GPPCD. Take a job's gross profit — collected price minus what it really cost to build — and divide by the crew-days it consumed. That's it. That's the king metric, and it's the same arithmetic Dale ran on me in Chapter 1 when the $8,000 window job beat my $20,000 sunroom — the day the sunroom king found out what his scarcest resource actually was. Here's why it dethrones every number you're currently managing by.
Take two jobs. A $30,000 sunroom at 40 percent margin — $12,000 gross, everybody high-fives at the sales meeting. And a boring $6,000 gutter job at 45 percent — $2,700 gross, nobody even mentions it. Now add the dimension the sales meeting never sees. The sunroom held a crew for ten days: $1,200 per crew-day. The gutter job was in and out in one: $2,700 per crew-day. The boring little job out-earned the flagship by more than double — for every day it occupied the one resource you cannot buy more of.
Because that's the thing about crew-days, and it's why they're the denominator: your year is not infinite. You have so many crews and so many working days, and every job you take evicts another job from that calendar. Ticket doesn't capture that. Margin percent doesn't capture it. A big impressive ticket at a decent margin can be — and routinely is — a machine for occupying your best crew at the worst rate in the building. You met this exact math in the sleeper question and the A-team question and the lane chapter. GPPCD is where all of them cash out.
Let me prove margin and GPPCD are two different animals with the job that taught me — because on paper, this one was a triumph.
Referral lead, spring of 2003, and I ran it myself. The customer was a retired New York graduate-school teacher — money in the bank, sharp as a tack, a hardass from the handshake — and she wanted something nobody in our market had ever built: an endless pool inside a glass sunroom. We were the only game in town for it. We'd been on a good run, I had a couple of crews I figured were capable, so I did my homework — called the Endless Pools people, talked through the kit, the skill set, the timeline — built a plan, and priced it at $60,000. She said yes. Drawings, engineering, approvals. We broke ground about the first of March.
We dug out for the pool vault. It rained the next day. So we bought pumps — remember, we had never built one of these — and we pumped it out. Then the vault caved in. The first time. That March it rained every single day for thirty straight, and we didn't have a roof up yet. Her yard looked like a bomb went off, and there was nothing to do but stand in it.
Now here's the part that earns this story its place in this chapter. She paid her $60,000 — exactly what we agreed, never a dollar's fight about it. And by the time it was done, the build had cost $85,000. Her sixty, plus twenty-five of mine. I paid $25,000 out of my own pocket for the privilege of the worst three months my production calendar ever had.
But here's the thing I need you to understand, because it's the whole lesson: the twenty-five wasn't the expensive part. If a $25,000 cash loss were the whole story, this would just be a bad-job story, and every contractor has a drawer full of those. What made it the disaster it was is the denominator. That job held my crews — my good crews — for the better part of three months. Every one of those crew-days had a market value: the sunrooms and window jobs they didn't build while they were pumping out a crater in the rain. The price wasn't the mistake; no price I could have quoted her would have fixed it. The mistake was opportunity cost — carrying the overhead, the schedule, and the best labor in my company while we fought a hole in the ground, and evicting a whole quarter's worth of profitable, boring, buildable work off the calendar to do it. The books eventually showed me the $25,000. Nothing I owned showed me the quarter. Margin measured the job. GPPCD measured the job against everything the job cost me the chance to build. Two very different things, and only one of them was on my reports.
And I'll tell you how close it came, because "opportunity cost" sounds like an accounting abstraction until it's got its hands around your throat: that job almost put me under. There was a month in that stretch where we could not install a single sunroom — the marquee product, the thing my company was known for, frozen solid while my best people fought the crater. What kept the doors open was windows. The little eight-thousand-dollar jobs, in and out in a day or two, the product Dale had already tried to tell me about — they quietly made payroll while the flagship nearly sank the boat. If it weren't for windows, we'd have been out of business. The line I'd have bet the truck on almost cost me the truck, the company, and everything else; the line I overlooked saved all three.
I did get some nice pictures.
One more thing about that year, and then we'll score some jobs. 2003 wasn't just the endless-pool year. It was the year we nearly doubled — the new showroom, the awards, the magazine cover, the Post-it note from Chapter 3. The year the whole company paid four million to install three. The endless pool wasn't a freak accident that happened to a healthy company. It was the flagship symptom of a company running full speed with its eye off the ball — the same disease, one job small enough to see whole. That's why it's the story I reach for when an owner tells me his margins are fine. Margins were fine on paper that year, too, right up until they weren't. The books don't have feelings. But they were the only thing in the building telling the truth.
Now, the scoring. Every job in the year lands in one of three bands, and I need you to hold two dollar figures in your head to read them right, because the bands are where owners jump to conclusions.
The first figure is your break-even floor — what a crew-day has to earn just to cover its share of the nut you stated back in Block One. The overhead, the trucks, the insurance, the lights, divided across your real crew capacity. Say it comes to $2,000 a crew-day; yours will be yours. Below that line, a job didn't underperform. It cost you money to build.
The second figure is your profit target — the floor plus an actual return for the risk, the callbacks you'll eat, the capital you floated, the gray hair. Call it $3,500 in this example.
Three bands:
PRINTS — clears the target. These jobs are the engine. The diagnostic question they answer is what do they have in common? Line, lead source, salesman, customer profile, crew. That common thread is your real lane — the one the books vote for, which may or may not be the one you stated in Block Six.
THIN — between the floor and the target. And here's where precision matters, because owners see the middle band and hear "bad." No. A THIN job covered its overhead and contributed something. It kept crews busy and paid the lights. It is not a loss — it's just not why you're in business, and a company can absolutely starve to death eating nothing but THIN. The question the THIN band answers is what would move these up? Usually it's not price. Usually it's speed. Hold that thought.
LOSES — below the floor. You paid for the privilege of building it. And when the year gets scored, almost every owner finds jobs down here that the sales meeting celebrated — big tickets, marquee installs, the showcase work from Chapter 12. This band is where beliefs come to die, and it's why the exam captured your beliefs first. If I'd shown you this table cold, you'd have argued with the table. But it's not my table. It's your jobs, your books, your stated nut. The books don't have feelings, and they also don't have an agenda. They were just never asked.
Two honesty rules before I show you the lever, because a scoring system is only as good as what it refuses to do.
Rule one: no anecdotes. Every job, the whole year, no exceptions. Not the jobs you remember — memory is the belief system this book exists to audit. The nightmare job you'd never take again and the sleeper you're sheepish about both go through the same math as everything else. Sometimes the nightmare scores fine and the memory was billing the whole category for one bad draw. Sometimes the sleeper is even better than you suspected. The scan doesn't skip organs.
Rule two: no invented numbers. The scoring runs on price, cost, and crew-days — and I told you in Chapter 10 that crew-days are the number almost nobody tracks. So here's how an honest system handles the hole: it estimates from trade norms, and it flags the estimate. Every score built on an estimated denominator says so, out loud, and the verdict carries a caveat until you confirm the real days. What it never does is quietly guess and hand you false precision — because one polished-looking wrong number would poison your trust in every right one. And the medicine that follows is non-negotiable: start writing crew-days down this week. One field on a job record. It is the single highest-value habit in this entire book, and it costs nothing but the deciding.
Now the lever. This is the sentence owners frame.
When an owner sees his THIN band, his reflex is forty years old and it's always the same: we need to raise prices. Sometimes he's right. But run the math before you touch the price tag, because GPPCD has two inputs, and the industry only ever pulls the numerator.
That $30,000 sunroom at $1,200 a crew-day. To hit a $3,500 target by price, you'd have to find $23,000 of new money from a customer who already shopped you against two other bids. Good luck. But take the build from ten days to eight — tighter scheduling, materials staged, the electrician locked to a date, punch list killed on the last day instead of a callback — and the same job, at the same price, jumps to $1,500. Take it to six days and you're at $2,000 — a 67 percent improvement in what the job pays you per day, and the customer's price never moved a dollar. She's happier, in fact, because her house was a job site for six days instead of ten.
Speed is the lever. Not hurry — speed. Fewer crew-days for the same build, which is a production discipline, not a sales discipline. It's why your production manager sat in the exam. It's why callbacks and permit purgatory and the flaky sub from Block Four aren't service annoyances — every one of them is denominator, quietly repricing jobs after the ink dried. Price raises start fights with the market. Speed raises happen entirely inside your own walls, and nobody can shop you against them.
Here's what exists now that didn't exist when you started this book: a scored year. Every job in a band. Every line ranked by what it actually returns per crew-day. Your real lane, your real stars, your real dogs — the reality column, complete at last.
And sitting right next to it, captured before the books could contaminate them: thirty-one beliefs, in your own words, with the date on the page.
Part III is what happens when the two columns finally look at each other. I've run that meeting more times than I can count, and I'll tell you what I've never seen once: an owner who was wrong about everything. That's not how it goes. You'll be right about most of it — you didn't survive this industry being a fool. But nobody's right about all of it, and the handful of misses, sorted by the size of the gap, will be worth more than everything a marketer ever sold you.
Time to line them up.
From The 48 Questions No Marketer Ever Had the Guts to Ask You. Read Chapter One free →