How a Business
Actually Works

Read this first

Most people who own a business don't actually know what a business is. Not in a way they could write on a napkin. They know their trade: how to cut hair, pour concrete, write code, bake bread. But the trade is not the business. The trade is the thing the business sells.

This is the whole problem. A person becomes excellent at the work, assumes that makes them good at business, then is confused when they're the best plumber in town and still broke.

So let's fix the root. By the end of this you should be able to look at any business (yours, your friend's, a stranger's) and see the machine underneath. The same machine is under all of them. A landscaping company, a law firm, a SaaS startup, and a taco truck are the same machine wearing different clothes.

Three promises about how this is written:

• Every term gets defined the first time it appears. If a word could mean ten things, I'll tell you the one thing I mean. If you catch me being vague, I've failed.
• Everything connects by cause and effect. Business is linear: if this number moves, then that number moves, in a direction you can predict. Change a price, and you'll know exactly which downstream numbers move and which don't.
• One real business gets traced through everything. His name is Ray, he runs a mobile rock-crushing service, and in Part 15 his entire business runs through every idea here, with real arithmetic. The scattered examples along the way are warm-ups. Ray is the proof.

One scope note: this covers how the machine works. It does not cover legal structure, taxes in detail, licenses, or insurance. Those vary by location; ask a local accountant. What's here is true everywhere.

Part 1

What a business is, in one sentence

Read it again. Every word is load-bearing.

"Solves a problem." Nobody pays you for effort. They pay you to make a problem go away. The dentist doesn't sell drilling; he sells "the pain stops" and "I'm not embarrassed to smile." Even lawyers billing by the hour aren't selling hours. The client pays for the contract reviewed or the lawsuit survived, and tolerates hourly billing as the meter. The moment the hours stop producing the outcome, they stop being paid for. Forget this and you'll spend your life talking about your process to people who only care about their outcome.

"For a group of people." Not one person. A group big enough and reachable enough that you can keep finding more of them. A problem only one weird person has is a hobby. Note the precise claim: the group must be big. Your slice can be small. Part 10 shows the number of customers you personally need is almost always smaller than you fear. Big pond, small catch, good living.

"At a cost lower than what they'll pay you." This gap is the entire point. It's called margin, and Part 4 measures it precisely. Costs you $40 to solve, they pay $100: you made $60 of value that didn't exist before. Not theft, not luck. It's the reward for solving the problem cheaper than they could solve it themselves. Business is the act of manufacturing that gap and keeping it.

"Over and over again." A single sale is a transaction. A business is a repeatable transaction. Anyone can sell their couch once. And "repeatable" has a concrete test: you can name, in advance, where the next ten customers will come from, and be roughly right. If every sale is a surprise, you have a streak, not a business.

Missing any one of the four? You have a job, a hobby, a one-time hustle, or a charity. All fine things. Just not businesses, and they won't behave like one.

Part 2

The machine, start to finish

The whole machine in one line. Memorize it.

That loop is the business. Everything else is detail.

Get attention. Before anyone can buy, they have to know you exist. Attention comes three ways. You can rent it (ads, which stop working the moment you stop paying). You can earn it (referrals, reviews, reputation: slow to build, cheap to keep). Or you can own it: a list of people who already know you, in the form of phone numbers, emails, and followers. The owned list is the most valuable asset most small businesses never bother to build. No attention, no business. A perfect product nobody hears about makes exactly zero dollars.

Make an offer. An offer is a specific promise: "Give me this much money, and I'll solve this specific problem, by this time, like this." "We do marketing" is not an offer. "We'll get your dental practice 20 new patients in 60 days or you don't pay" is an offer. The clearer the promise, the easier everything downstream.

Deliver the result. Make the problem go away. The result, not the work. Activity is not delivery. The customer doesn't care that you were busy. They care that the problem is gone. (Skill 2 in Part 7 is the tool for telling the difference.)

Get paid more than it cost you. All of it has to clear: materials, labor, ads, rent, and your own wage. That last one is the detail most owners get wrong, and Part 4 fixes it. If the loop costs more than it collects, you have an expensive way to lose money with extra steps.

Reinvest in attention. Once Part 4's math proves each customer is profitable, reinvest a fixed percentage of revenue (healthy small businesses often run 5–10%) and treat it like rent: paid every month, not "when we can afford it." A business that profitably turns $1 of attention into $3 of customer doesn't have a growth problem. It has a "how fast can I pour in dollars" problem. That is the only good problem in business.

When a business is broken, it's broken at one of those five steps. Your job is to find which one. Not all of them. One. (Part 6 is the procedure.)

Part 3

The four jobs every business has

Zoom in and every business, no exceptions, is four jobs running at once. "Running a business is overwhelming" means the four jobs were never separated, so they feel like one giant blob. Separate them and the blob becomes four manageable things, each with its own number.

Job 1: Get leads. A lead is a person who raised their hand: a call, a form, a "how much do you charge?" Not a sale. A hand in the air. The output is leads per week. If it's zero, nothing else matters.

Job 2: Turn leads into customers. Sales. The outputs are conversion rate (the fraction of leads who buy; 10 leads, 3 buy, 30%) and average sale size. A business can double revenue without a single new lead just by getting better at this job. Most owners never realize that.

Job 3: Deliver the result. The output is a customer whose problem is actually gone. Done well, it feeds the other three jobs free, through referrals and repeats. Done badly, it poisons everything upstream. You cannot outmarket a product that doesn't deliver.

Job 4: Keep them and sell them again. The first sale to any person is the hardest and most expensive you'll ever make to them. The second is cheap. The output is lifetime value: what a customer is worth across the whole relationship, defined in Part 4. This is the most profitable job in business, and the one small businesses neglect most.

The map to Part 2: Job 1 is "get attention," Job 2 runs "make an offer" through "get paid," Job 3 is "deliver," Job 4 makes "over and over again" true. Same machine, two zoom levels.

Write the four down. Which is your business worst at right now? That's your constraint, and the constraint is Part 6's subject.

Part 4

The money, said plainly

People avoid this part because it has numbers. Don't. The numbers are simple, and not knowing them is why most businesses die.

Revenue is what you collect. Revenue is not profit, and confusing the two has bankrupted more owners than any competitor. "We did a million in sales" says nothing about whether anyone made money.

Cost of delivering is what it costs to fulfill one sale: the materials, the labor for that job, the fuel to get there. The test: if the sale doesn't happen, does the cost disappear? Yes means cost of delivering. No means fixed cost.

Gross margin is what's left. Sell a $100 haircut, pay the barber $40: gross margin is $60, or 60%. Both forms matter. Margin dollars ($60) say how much fuel each sale adds. Margin percent (60%) says how efficiently revenue becomes fuel, and survives price changes for comparison. This is why two businesses with identical revenue live in different universes: one keeps $60 per sale, the other $10.

Fixed costs get paid whether you sell anything or not: rent, software, insurance, salaries. They show up every month with their hand out.

Your own wage is a fixed cost. This sentence is missing from most business advice, and its absence ruins people. Decide what you'd pay a stranger to do your job in the business (the market wage for the operator role) and count it as a fixed cost before calculating profit. A business that only "makes money" because the owner works free isn't profitable; it's a badly paying job with extra risk. Can't afford to pay yourself yet? Fine. Early businesses often can't. But write the unpaid wage down as the real number it is, and aim the machine at closing the gap. Don't let "we're profitable!" mean "I paid everyone but me."

Profit is gross margin from all sales minus all fixed costs, wage included. $20,000 of margin this month, $15,000 of fixed costs: $5,000 profit. (One more honesty check: that's pre-tax. Set aside a percentage every month, at the rate your accountant gives you, so April isn't a heart attack.)

Now the two ratios that decide whether you live or die.

CAC, the cost to acquire a customer. Everything spent getting attention and converting it over a period (ads, the salesperson's time, the software) divided by new customers that period. Spent $1,000, got 10: CAC is $100. Customers have a price. Know it the way you know the price of gas.

One precision most guides skip: there are two CACs. Paid CAC counts only customers from paid channels, against what you paid. Blended CAC divides total spend by all new customers, referrals and walk-ins included. Mostly word-of-mouth customers make blended CAC look beautifully low, and that low number says nothing about whether your ads work. Paid CAC judges a channel. Blended CAC judges the business.

LTV, lifetime value. What one customer is worth in gross margin (not revenue) across the whole relationship. Gym member pays $50/month, stays 20 months, 80% margin: LTV = 50 × 20 × 0.8 = $800. Margin, not revenue, because you'll compare it to CAC, and acquisition costs are paid out of margin. Notice the ingredients: price × frequency × how long they stay. That last term is churn, the rate customers leave, and it's the silent killer inside every LTV. Halve how long customers stay and you halve LTV without touching price or cost.

Pay $100 for a customer worth $800: a machine. Pour money in. Pay $100 for a customer worth $90: a shredder, and more sales kill you faster. Most owners who "can't figure out why growth isn't helping" are running a shredder and don't know it.

Two adjustments so the rule fits every business:

Customers buy once (roofing, wedding photography)? There's no repeat "lifetime," so the rule collapses to: margin on the single sale ≥ 3× CAC. Referrals become your whole Job 4. The repeat purchase you're owed is the customer's brother-in-law.

Money comes in slowly ($800 over 20 months)? Also track payback period: months until a customer's accumulated margin covers their CAC. CAC $100, margin $40/month: payback 2.5 months. Shorter is safer, because of what's next.

Cash. Profit is a number on a page. Cash is what's in the bank today. You can be profitable on paper and die anyway: you paid for materials in March, the customer paid you in June, and payroll was due in April. A business doesn't die when it stops being profitable. It dies when it runs out of cash. Those are not the same day.

Measure cash as runway: bank balance ÷ monthly fixed costs = months you'd survive if sales stopped. Three months is a floor; six is comfortable. And know the three levers that move cash without touching profit. Collect sooner: deposits, invoice the day the job finishes, card on file, shorter terms. Pay later: ask suppliers for terms; many say yes if asked. Hold less stock: inventory is cash sleeping on a shelf in a costume. Watch the bank balance like a heart monitor, because that's what it is.

Part 5

The driver tree: business as one multiplication problem

This is the shortest part and maybe the most useful page in the guide. Parts 1–4 compress into two equations. Write them for your own business and every problem becomes traceable: a symptom in one number is caused by a number upstream, and you can follow the chain backward like a plumber following a pipe.

The entire business in two lines. Five numbers to grow (leads, conversion, average sale, frequency, margin percent), one to shrink (fixed costs). Every tactic ever invented, from ads and scripts to upsells, loyalty programs, and price changes, is an attempt to move one of those six. When someone pitches you anything, ask: which number does this move? If they can't answer, neither can the tactic.

Three consequences fall out, each an if-then you can take to the bank:

If the numbers multiply, then small improvements stack violently. Improve leads 10%, conversion 10%, average sale 10% (three modest wins) and revenue grows 1.1 × 1.1 × 1.1 = 33%, not 30%. The reverse is how businesses quietly rot: three numbers each slip 10%, revenue drops 27%, and every individual slip looked too small to panic about.

If a number is multiplied by zero, then everything is zero. Perfect conversion times zero leads is zero customers. The world's best product times zero attention is zero revenue. Find your zero before polishing anything else.

If profit is margin minus fixed costs, then growth fixes nothing when margin per sale is negative. More units of a money-losing sale dig the hole faster. Check margin before pouring in leads. (This is Part 12's whole argument, previewed.)

Trace one chain to see the linearity. Ray raises his price per ton 10%. Average sale rises 10% → revenue rises ~10% (if customers stay; Part 9 gives the exact breakeven) → cost of delivering doesn't move, so margin dollars rise more than 10% → profit rises more still, because fixed costs didn't move either. One input, four predictable downstream effects. That's what "knowing how business works" means. Not vibes. Arithmetic.

Write your two equations today, with real numbers. Any of the six you can't fill in is itself a finding. Go measure it this week.

Part 6

The bottleneck: fix one thing, and here's how to find it

The most common owner mistake: improving everything at once and therefore improving nothing.

At any moment, a business has one constraint: the single narrowest point holding back the whole machine. Picture a pipe with sections of different widths. Money flows only as fast as the narrowest section allows. Widening any other section does nothing. Find the narrow part, widen it, and a different section becomes narrowest. Move there. One at a time. Forever. That's what "running a business" actually is, day to day.

The original sin of most business advice is saying "find your bottleneck" and walking away. Here's the actual procedure. Under an hour.

Step 1: Fill in your six driver-tree numbers (Part 5) for the last 90 days.

Step 2: Look where work piles up or runs dry. Constraints are visible:

• Empty calendar, phone silent → leads (Job 1).
• Plenty of inquiries, few buyers. Quotes unanswered, "let me think about it" → conversion (Job 2).
• Calendar full, working flat out, little profit → price or margin (busy because underpriced; see Part 9) or waste (paying for minutes the customer isn't; see Skill 2).
• Selling fine, but refunds, callbacks, bad reviews → delivery (Job 3).
• Happy customers who buy once and vanish → retention (Job 4).
• Every number fine, bank account still scared → cash timing (Part 4's levers) or fixed costs.

Step 3: Put every spare hour and dollar on that one constraint until it's no longer the narrowest point. Then re-run the procedure. The constraint moves when you fix it. That's success, not failure.

The discipline: improving anything other than the constraint is wasted effort, even when the improvement is real. A better logo while leads go unanswered is a nicer paint job on a car with no engine.

Part 7

The seven skills: how to actually see a business

Everything so far is knowledge. This part is skill: trained ways of seeing that let you walk into any business, including yours, and find what's broken. None requires a degree. Each has an exercise you can do this week.

Skill 1 · Draw the machine (process mapping, or "SIPOC")

You can't fix a machine you can't draw. SIPOC is a borrowed manufacturing tool; the name is the five columns you write: Suppliers, Inputs, Process, Outputs, Customers.

• Suppliers: who provides what you need (vendors, ad platform, subcontractors, landlord).
• Inputs: what they provide (materials, leads, labor, the space).
• Process: the 5–7 steps that turn a stranger into a paying customer with a solved problem. Not 30 steps. The height at which you'd explain it to a new employee on day one.
• Outputs: what comes out (solved problem, invoice, review, referral).
• Customers: who receives each output.

Exercise (15 minutes, one sheet): Write your process row: stranger hears of us → contacts us → gets a quote → says yes → we deliver → we get paid → we follow up. Then for each step, write who does it and what number proves it's working (step one's number is leads per week; the quote step's is conversion rate; and so on).

Why it matters: every step is a place the machine can leak, and now each leak has a name, an owner, and a gauge. You stop saying "business is slow," which is a feeling, and start saying "step 3 converts at 15%, down from 30%," which is a fact with an address.

Skill 2 · Tell value from waste

Value is any minute of work or dollar of cost that moves the customer's problem closer to solved: something they'd willingly pay for if they saw it itemized. Waste is every other minute and dollar. The skill is seeing the difference in your own day, which is hard, because waste feels like work. Being busy is the most virtuous-feeling thing in business.

The test: imagine your invoice listed everything you actually did. "Crushing your rock to spec: 9 hours. Sitting idle waiting for blasted rock because the pit plan changed: 2 hours. Re-screening a stockpile we crushed to the wrong spec: 3 hours. Second trip to the site because the spare jaw dies were left at the yard: half a day." The customer gladly pays for line one. Every other line is waste: cost you carry that creates nothing they'd pay for.

The classic wastes, in small-business English. Waiting: the lead that sat in your inbox two days and went cold; the most expensive waste in small business and the least visible. Rework: anything done twice because it wasn't right once; callbacks, refunds, redos. Extra motion: bad routing, three trips to the supplier. Over-doing: polish the customer never asked for and won't notice, while underdelivering what they did ask for. Idle stock: cash sleeping on shelves, or iron sleeping in the yard. Wrong person on the task: the owner, the most expensive hour in the company, doing $15/hour data entry. (Part 11 is about this one.)

Why it matters: cutting cost by "working faster" makes people frantic and quality worse. Deleting waste minutes makes the same people produce more value in the same hours. When Part 6 says your constraint is margin, this is usually the fix. Don't hurry. Stop paying for minutes nobody would buy.

Skill 3 · Napkin math

The ability to compute by hand, in minutes: margin per sale, monthly profit, CAC, LTV, runway, and the two driver-tree equations. Sixth-grade arithmetic. The skill isn't difficulty. It's the habit of reaching for the napkin before reaching for an opinion. "Should I run ads?" isn't a debate; it's CAC versus LTV. "Can I afford a hire?" isn't a feeling; it's margin × added sales versus wage. Exercise: compute all eight scoreboard numbers (Part 14) this week. Any you can't compute, go measure.

Skill 4 · Find the constraint

Part 6 in one line: locate the narrowest point, widen only that, repeat. Listed here because it belongs in the set you drill. Exercise: run Part 6's procedure quarterly, on the calendar, like a dentist appointment for the business.

Skill 5 · Ask "why" five times

Symptoms are loud; causes are quiet. Refuse to fix the symptom until you've followed the chain to the cause. Ask why, then ask why about the answer, about five times, until you hit something you can permanently change.

Customers aren't coming back. Why? Jobs keep running past schedule. Why? The crusher breaks down mid-job. Why? Wear parts get changed when they fail, never before. Why? There's no maintenance schedule; everything is reactive. There. Fix "no maintenance schedule" and the whole chain above it heals. Fix the symptom instead (a discount to win the unhappy contractor's next job) and you'll pay forever for a problem that's still there.

Skill 6 · Change one thing, then measure

The scientific method, kitchen version: pick the number you're moving → change exactly one thing → wait for real signal → compare → keep it or put it back. Two changes at once teach you nothing. You raised the price and redesigned the website, revenue moved, and now you're guessing again, just with more confidence. One at a time feels slow and is the fastest route there is, because every change produces a fact. A year of this is 20–30 facts about what moves your numbers. Your competitors will have a year of vibes.

Skill 7 · Write it down (what a "system" literally is)

Every business guide says "build systems." Almost none says what a system physically is, which lets owners nod along for years without building one. The literal definition:

That's all. Not software, not a binder with a logo. "How we mobilize the crushing spread" as nine written steps, owned by the crew lead, proven by crushing within 24 hours of hitting the site. A task that lives only in someone's head isn't a system. It's a memory, and it leaves in the parking lot every night at 5pm. Exercise: take the one task you'd least like to re-explain to a new hire and write the checklist while doing it. During, not after. After is how steps get forgotten. Your business now contains one system. Part 11 shows why this unglamorous skill is the exit door from owning a job.

These seven are the toolbox. The next two parts are the most powerful things to point them at: what customers value, and what you charge.

Part 8

Why people actually buy (the value equation)

You can't make a good offer without knowing what "value" means to a buyer. People think value means cheap. It doesn't. Value is what someone gets versus what it costs them to get it, and "cost" is much more than money. Four levers: two to push up, two to push down.

1. The dream outcome (up). Not the product. The result. People don't want a mattress; they want to wake up rested. Sell the rested. The bigger and more vivid the result you can credibly promise, the more value.

2. Perceived likelihood of success (up). Do they believe it will work for them? Promise the moon, and if they don't believe you, value collapses to zero. Guarantees, reviews, before-and-afters, case studies: all exist to push this lever. "Will this actually work for me?" is the question every buyer silently asks and most sellers never answer.

3. Time delay (down). Sooner is worth more. Much more. The same outcome in a week is worth multiples of the same outcome in a year. Even a quick early win ("we'll fix the worst room today") raises the value of the whole engagement.

4. Effort and sacrifice (down). How much work, hassle, and risk does the customer carry? "Done for you" beats "done with you" beats "do it yourself," and commands a higher price at each step, because you took effort off their side of the scale. People pay enormous sums to not have to do something.

A big outcome they believe is likely, delivered fast, with little effort on their part: maximize the top two, minimize the bottom two, and you can charge a lot, because you're worth a lot. Most cheap businesses never tried to move these levers. Price was the only dial they knew existed. Move the levers and price stops being the conversation.

(These levers ARE the if-then chain: each one raises what customers will pay or how many say yes, moving average sale or conversion rate in Part 5. Nothing in this guide floats free of the driver tree.)

Part 9

Charge more (yes, more; here's the math and the method)

Most small business owners undercharge. Not a little. A lot. They price off fear ("nobody will pay more") instead of value. Raising price is usually the fastest fix for a struggling business, and unlike every other improvement it needs no new skill, staff, or spend.

The asymmetry. Sell at $100 with $40 delivery cost: keep $60. Raise 30% to $130: cost is still $40, keep $90. Price up 30%; profit per sale up 50%. Every dollar of a price increase falls straight to the bottom line, because cost doesn't move. No other lever in the driver tree does this.

The breakeven formula, so you can stop fearing lost customers. A raise may cost you some. Here's exactly how many you can afford to lose:

Here: $60 ÷ $90 = 67%. Keep two-thirds of your customers and the raise pays for itself. Keep more and you're ahead, with fewer customers and less work. A 30% raise that loses 15% of customers: 100 × $60 = $6,000 of margin before; 85 × $90 = $7,650 after. More money, fewer headaches. The customers who leave over a fair increase are disproportionately your most demanding, lowest-margin ones. (Not always. Occasionally a price-savvy good one walks, which is why you test instead of leap.)

How to raise prices without betting the company (Skill 6: one change, measured):

1. Set the floor and ceiling. Floor: cost of delivering plus your share of fixed costs. Below it every sale loses money, and no volume fixes that. Ceiling: what the value equation says the result is worth to the customer. Your price lives between them. For most undercharging owners, it lives far closer to the ceiling than they think.
2. Quote the new price to new customers only. Existing customers notice nothing; new quotes are the experiment.
3. Watch conversion for 4–6 weeks. If it barely drops, your old price was a donation. (A useful shock: if nobody ever balks at your price, you are definitively underpriced. A healthy price gets some no's.)
4. Then migrate existing customers with notice, a reason, and grandfathered timing. Or keep your best ones at the old price on purpose, as a loyalty decision rather than a fear you obeyed.

Why higher prices compound. Price is a signal (people partly infer quality from it), and higher margin funds better delivery: better materials, more time per job, faster response. That makes the higher price honest, which makes referrals likelier, which lowers CAC. Cheap traps you in the opposite loop: thin margin → can't deliver well → unimpressed customers → no referrals → clawing for every cold lead → small and exhausted.

The caveat, said clearly: raising price without raising value is just greed, and it won't hold. Do both. Make it worth more (Part 8), then charge what it's worth.

The honest boundary: a minority of businesses are true price-takers, selling a commodity where the market sets the price and customers genuinely choose on price alone (raw materials, some retail, gas stations). If that's you, "charge more" is spelled differently: change what you sell until it's not a commodity. Bundle service, add speed or convenience or a guarantee. Move a Part 8 lever, so there's something to charge for. Nobody escapes commodity pricing by being a slightly nicer commodity.

Part 10

You need fewer customers than you think

A reframe that calms people down. Most owners imagine they need a flood of customers. Usually they need a trickle. Do the arithmetic on your actual life.

Want $120,000 a year in profit, with $1,000 of profit per customer? You need 120 customers a year. Ten a month. One every three days. The goal isn't "go viral" or "beat the whole market." It's "find one customer every three days," a completely different, completely doable problem.

Connect it to Part 9: raise price and margin so each customer is worth $2,000, and you need 60 a year. About one a week. The business got easier and you make the same money.

"Charge more" and "fewer customers" are joined at the hip, and they run as a loop: higher value per customer → fewer needed → more attention for each → better delivery → more referrals → even fewer cold leads to chase. The high-volume, low-price game grinds owners into dust. This loop is the refusal to play it.

(Reconciling with Part 1's "group big enough": the market must be large enough that one customer every three days is easy to find. Your share stays tiny. You don't need to win the ocean. You need a reliable fishing spot.)

Your number is target profit ÷ profit per customer ÷ 12. It's almost always smaller and more human than the panic in your head.

Part 11

The owner's real job

The hardest shift, and the one that separates owning a job from owning a business:

At the start, you are the machine: leads, sales, delivery, books. Normal, correct, unavoidable. The trap isn't being the machine at the start; it's never beginning the climb out. Never climb out and you own a job you can't quit, can't sell, and that collapses the day you get sick, because the business is you.

The distinction worth tattooing somewhere: working in the business (doing the trade) versus working on it (building systems, hiring, improving the offer, fixing the constraint). An hour worked in earns once and is gone. An hour worked on compounds: the system keeps producing after you've gone home. This is leverage, meaning more output from the same input. The richest, freest owners aren't the hardest workers. They built machines that produce whether or not they touched anything today.

The hand-off procedure. "Just delegate" is the "just find your bottleneck" of hiring advice: true and useless. The actual method, one task at a time:

1. Pick the task by the wrong-person test (Skill 2): the thing you do regularly that someone could be paid least to do well. Shed $15/hour tasks before $50/hour tasks.
2. Write the checklist while doing the task yourself (Skill 7). During, not after.
3. Show them once. They watch you run it.
4. Watch them once. Wherever they stumble, fix the checklist, not just the person. A stumble means a step was missing or unclear.
5. Hand it off with its number: the measure that proves it's done right (from your SIPOC map).
6. Check the number, not the person. Spot-check weekly, then monthly. Number's healthy? Stay out of the kitchen.

Each pass permanently buys back hours of your week. Your hours are the scarcest, most expensive resource in the company. Spend the recovered ones only on things that compound: the constraint, the offer, the systems, the key hires.

When can you afford to hire? Napkin math, not vibes. A hire pays for themselves if margin per sale × additional sales they enable exceeds their monthly cost, or if they free enough of your hours to generate the same. A $4,000/month employee who frees you to close $8,000/month of new margin isn't a cost. Staying solo to "save money" is often the most expensive decision an owner makes. The bill arrives quietly, as growth that never happened.

The test: one month of vacation, no phone. Does the business survive? Every system written and task handed off moves the answer toward yes. "Yes" is what makes a business sellable, restable, survivable. (And per Part 12, this is the fifth step of building a business, not the first. Pre-profit, you're supposed to still be the machine. No guilt. Just don't retire there.)

Part 12

The order of operations

When everything feels broken, people freeze. Here's the order to fix things in. Each step assumes the one before it. Skipping ahead doesn't make you fast; it multiplies a broken machine.

First: prove people will pay. Can you get strangers (not friends, not "great idea!", strangers) to hand you actual money? If not, no marketing or system will save you. You don't have a problem worth solving yet, or you're offering it to the wrong group. Fix the offer or the audience.

And give this step a deadline, because "still validating" can hide a corpse for years. The kill criterion: if 90 days of real attempts (real offers, real strangers, real price) can't produce a handful of paying customers, stop. Change the offer, the audience, or the idea. Not the logo. Killing a dead idea in 90 days isn't failure. It's the cheapest education you'll ever buy.

Second: make one sale profitable. A single sale that makes money after all its real costs, including your time at a real wage. Lose money on one and you'll lose more on a hundred. This is the step everyone skips, and why "we grew and went broke" is such a common story.

Third: make it repeatable. Can the profitable sale happen again on purpose? One channel that reliably produces leads, one approach that reliably converts. Predictability, not volume. The Part 1 test: name where the next ten customers come from, be roughly right.

Fourth: now pour fuel. Only now: more ad spend, more people, more capacity. Scaling before this point multiplies a broken machine; after it, a working one. Same action, opposite result. The only difference is the order.

Fifth: build the machine that runs without you (Part 11). The work of years, not weeks.

Most stuck businesses are attempting step four without finishing step two. Find which step you're actually on. Be honest. Do that step.

Part 13

The lies small business owners believe

Comforting ideas that quietly keep people broke. Check your own head.

"If I just work harder, it'll work." Effort anywhere but the constraint produces nothing but exhaustion. Hard work at the bottleneck is gold. Everywhere else, it's a hamster wheel with better intentions.

"I can't raise prices; my customers are price sensitive." Some are. Your best ones usually aren't, and you've been letting your cheapest, most demanding customers set the price for everyone. Run the breakeven formula before you run the fear.

"I don't need marketing; my work speaks for itself." Your work cannot speak. It has no mouth. The best product nobody hears about loses to the mediocre product everybody hears about, every time. Attention is a job, not a reward good work earns automatically.

"More revenue will fix it." If each sale loses money, more revenue digs faster. Margin first, then volume. Revenue hides profit problems right up until it can't.

"I'll hire when I can afford it." Often backwards. The right hire, one who frees your hours or directly adds sales, is how you get to where you can afford more. Staying solo to save money can be the most expensive choice you make.

"Being busy means it's working." Busy is not profitable; motion is not progress. Run the invoice test: how many of your busy minutes would a customer pay for? You can be slammed every day and going broke. That's the most common way to go broke.

"We're profitable" (while paying yourself nothing). If profit only exists because the owner works free, the business is losing money and the owner is donating the difference. Put your wage in the fixed costs and look at the real number. It might hurt. It's still the real number.

"Great month. The business is working." One month is weather. The driver tree run quarterly is climate. Judge the machine, not the month.

Part 14

The scoreboard

You can't run what you don't measure, and you'll drown measuring everything. Eight numbers. Weekly. That's the whole discipline. Where each comes from is listed, so "I don't have that data" lasts exactly one week:

1. Leads per week. Count every hand-raise: calls, forms, DMs, walk-ins. A paper tally works. (Job 1)
2. Conversion rate. Leads that became customers ÷ total leads, same tally. (Job 2)
3. Average sale size. Revenue ÷ number of sales. (Job 2)
4. Gross margin per sale. Price minus cost of delivering, dollars and percent. (Job 3's economics)
5. CAC. Attention-and-sales spend ÷ new customers, paid and blended. (Jobs 1–2)
6. LTV. Margin per purchase × lifetime purchases. And the ratio: LTV ≥ 3× CAC? (Job 4)
7. Monthly profit. Total margin minus all fixed costs, your wage included.
8. Cash runway. Bank balance ÷ monthly fixed costs, in months. The heart monitor.

When one goes ugly, it points at exactly one place in the machine. The full if-this-then-that:

Eight numbers weekly, plus this table: you always know which job is the bottleneck and what to try first. Not bureaucracy. The instrument panel. You wouldn't fly a plane without one.

Part 15

The whole machine, once, with real numbers

Meet Ray. He runs a mobile rock-crushing service: he hauls a crushing spread (jaw crusher, screen plant, excavator, loader) onto a contractor's site and turns blasted rock and demolition concrete into road base, at $6.00 a ton. Heavy civil. Heavy iron. Heavy fixed costs. Let's run his whole business through this guide.

The driver tree (Part 5, Skill 3). Cost of delivering one ton: fuel $1.20, wear parts and maintenance $0.90, operator labor $0.90, mobilization amortized $0.60 = $3.60. Gross margin: $2.40 a ton (40%). The spread crushes about 1,250 tons a day and last quarter averaged 16 crushing days a month → 20,000 tons → revenue $120,000 a month, gross margin $48,000. Fixed costs: equipment finance $33,000 (the iron gets paid whether it crushes or sleeps), insurance, shop, and office $5,000, and Ray's operator wage $10,000 (in fixed costs, where it belongs; see Part 4). Total: $48,000. Profit: $48,000 − $48,000 = zero. A $1.4-million-a-year company earning nothing. And Ray feels profitable, because his wage masks it. Part 4's honest arithmetic just earned its keep.

Acquisition. About $3,000 a month of business development (association dues, site visits, nights spent estimating) brings 10 bid invitations; he wins 20%: 2 jobs a month, roughly one from a new contractor. CAC ≈ $3,000. A contractor historically gives him 3 jobs (~12,500 tons each) before drifting away, so LTV = 37,500 tons × $2.40 = $90,000 of margin. LTV:CAC is 30-to-1, ten times the rule. By the ratio, he should buy more attention. Right?

Wrong. Run the procedure (Part 6). Step 2: where does the machine leak? The calendar. The spread sits idle 6 days a month between jobs, while fixed costs tick at $48,000 ÷ 22 available days ≈ $2,200 a day, crushing or rusting. More bid invitations wouldn't help; the leak is downstream of leads. Five whys (Skill 5): the spread sits idle. Why? The next job isn't lined up when the current one ends. Why? Quotes go out a week late. Why? Ray prices every bid himself, at night, after running a loader all day. Why? There's no rate sheet; every quote is built from scratch. Bottom found: the constraint is quote turnaround, wearing a utilization costume.

The fix is a system (Skill 7). A standard rate sheet (tonnage bands, material types, mobilization zones) owned by his office manager, proven by a number: every bid out within 24 hours. Cost: one weekend. Result, measured over a quarter (Skill 6): win rate climbs from 20% to 30% (contractors often take the first credible number), the gaps fill, idle days drop from 6 to 2. Four reclaimed crushing days × 1,250 tons × $2.40 = $12,000 a month of new margin at zero new fixed cost. Profit goes from $0 to $12,000 a month. Same iron, same crew. Arithmetic.

The constraint moves. Re-run the procedure. Next leak: contractors do 3 jobs and drift. Why? Nobody calls them after demobilization, so their next project gets crushed by whoever shows up asking. Fix: another system. A post-job report (tonnage, gradation specs, photos) plus a tracked call when each contractor's next project breaks ground. Average relationship grows from 3 jobs to 5: LTV rises from $90,000 to $150,000. A kept contractor costs a phone call; a replaced one costs $3,000 and a bid cycle.

Then price (Part 9). He tests $6.50 a ton on new bids only. Margin per ton: $2.40 → $2.90, up 21% on an 8% price move (the asymmetry). Breakeven retention: $2.40 ÷ $2.90 = 83%, so he can lose 17% of his wins before the raise hurts. Win rate barely moves, because he now wins on 24-hour quotes and reliability, not price. Each job is now worth 12,500 × $2.90 = $36,250 of margin.

Then capacity (Part 11). Backlog stretches to 8 weeks; the constraint moves to capacity. Napkin math: a second spread adds ~20,000 tons a month × $2.90 ≈ $58,000 of margin against roughly $38,000 of new fixed costs: about $20,000 a month. And it pencils only because the rate sheet, the follow-up system, and the maintenance schedule already exist. Note the order: buying more iron before fixing idle days would have doubled the idle-day problem at twice the finance payment. Step four before step two, the classic grave (Part 12). He staffs it with the six-step hand-off: checklist, show once, watch once, hand off with its number, check the number. He's now working on the machine, and the vacation test gets closer to "yes" every quarter.

Same iron, same rock, same town. The difference between the break-even version and the compounding one was never effort. Ray was working nights either way. It was seeing the machine: drawing it, reading its eight numbers, finding the one narrowest point, fixing that, repeating. Every move was arithmetic before it was action.

The whole thing on a napkin

If you forget everything else, keep this

A business solves a problem for a group of people, at a profit, repeatably. It runs as a loop (get attention, make an offer, deliver the result, get paid more than it cost, reinvest in attention), and inside the loop are four jobs: get leads, convert them, deliver, keep them. The whole machine is two equations: Revenue = Leads × Conversion × Average sale × Purchases per year, and Profit = Revenue × Margin% − Fixed costs (your wage included). At any moment exactly one point in that machine is the bottleneck. Your job is to find it with the numbers, widen it, and move to the next, using seven learnable skills: draw the machine, tell value from waste, do napkin math, find the constraint, ask why five times, change one thing at a time, and write systems down as checklist + owner + number. Keep LTV at 3× CAC, keep months of cash in the bank, charge more than you're comfortable with by being worth more than you're comfortable with, and remember you need fewer customers than you fear. The end goal is a machine that runs without you, because that's the difference between owning a business and owning a job.

That's it. That's the whole thing. It's not complicated. It's just not easy. Now go find your bottleneck.