Introduction — The Spreadsheet With No Column for Miles

Three hundred and forty rows, eleven months, and no column for the thing that mattered.

The spreadsheet with no column for miles

Kit was not careless. That is the first thing to establish, because the story only means something if you accept it.

Kit tracked everything. Three hundred and forty rows, eleven months, every shift recorded on the evening it happened. Date. Hours online. Gross earnings. Base pay separated from tips, because Kit had read somewhere that mixing them hides what is really going on. Which app. Which hours. A note about weather, added in March, after a wet Tuesday paid unexpectedly well.

⚠ Kit could tell you, from data rather than impression, that Thursday evenings beat Friday evenings by about eleven per cent in that market. Almost nobody in this trade can tell you anything like that about their own work.

The headline number was twenty-two dollars and change per hour, averaged across eleven months including the bad weeks.

Then a transmission failed at a hundred and forty-three thousand miles.

The car had been bought at eighty-one thousand miles, nine years old, for a figure Kit could still recite. It had been fine. It had needed the ordinary things — tyres, a battery, brakes twice — and Kit had paid for all of them out of the same account the earnings went into, which is how almost everybody does it and which is exactly the problem.

⚠ The repair quote was more than the car was worth.

So Kit sat down with the spreadsheet and did, for the first time, a piece of arithmetic that eleven months of careful record-keeping had never prompted.

⚠ Sixty-two thousand miles on the odometer since purchase. Thirty-one thousand of them in eleven months.

Half the car's remaining life, consumed in under a year, by work that had reported itself as twenty-two dollars an hour.

⚠ Kit had not been earning twenty-two dollars an hour. Kit had been earning twenty-two dollars an hour and spending a car — and the car did not send an invoice until the day it stopped.

What a car costs when nobody bills you for it

Here is why this is a structural problem rather than a personal failing.

Every cost in an ordinary business announces itself. Rent arrives monthly. Stock has to be bought before it can be sold. Wages are paid on a date. You cannot fail to notice them because somebody sends you a bill.

⚠ Vehicle depreciation is the only major cost in this trade that never announces itself. It accrues on every mile, silently, and then presents itself once, years later, in a single number that has nothing obviously to do with the Tuesday shifts that caused it.

And the way the work reports itself makes this worse rather than better. Open any platform at the end of a shift and it will tell you what you earned. Some will tell you how long you were online. None of them will tell you how far you drove, because your mileage is not their business — it is yours, sitting on your driveway, quietly worth less than it was this morning.

⚠ So the driver has one half of an arithmetic problem presented to them constantly, in large friendly numbers, and the other half not presented at all.

What that does in practice.

⚠ You accept offers you would refuse if you could see the whole calculation ☐ ⚠ You work zones that pay well per hour and badly per mile ☐ ⚠ You compare yourself to other drivers on the wrong number ☐ ⚠ You feel like you are doing well right up until a repair bill, and then you feel like you were a fool, and you were not — you were working on incomplete instrumentation ☐

The fix is not complicated.

⚠ Photograph the odometer at the start of a shift and at the end. Two numbers, four seconds, and you now have the denominator that every meaningful figure in this book depends on. Chapter 7.

Why the offer decision is the whole business

The second thing this book is built on is smaller and more immediate.

⚠ You will make between twenty and fifty accept-or-decline decisions in a working day, most of them in under a minute, many of them while the car is moving.

Every one of those decisions is an economic decision. Take this and you have committed the next thirty minutes and the next eleven miles. Decline it and you have committed nothing and earned nothing. That is the whole business, repeated forty times, and it is the only real decision the job contains.

⚠ And it is presented to you on a screen that shows you some of what you need and not all of it.

You are typically shown a payment figure and some indication of distance. You are frequently not shown where you will end up relative to where the work is, what the wait at the restaurant will be, whether the address is a tower block with no parking, or what the whole thing will cost you in miles once you have driven back to somewhere with orders in it.

⚠ Some of that is unknowable. Some of it is knowable and you have to supply it yourself, from your own records, in about fifteen seconds. Chapter 21.

And then there is the part almost nobody does.

⚠ You have no data at all about your own decisions unless you record them. A driver who logs only what they accepted has a record of their own compliance. A driver who logs what they declined, and why, and what it would have paid per mile, can tell within a fortnight whether their instincts are actually costing them money. Chapter 5.5.

The four rules

Four sentences. Short deliberately, because you have to be able to hold them in your head at a kerbside with a phone buzzing.

⚠ Rule 1: You never handle the phone while the car is moving.

⚠ First, because it is the one that kills people — you, or somebody else, or somebody's child on a crossing.

This is the rule the trade is structurally arranged against. Offers expire. Navigation reroutes. Customers message. Support pings. The entire working rhythm of app-based delivery is built on interacting with a screen, and the majority of drivers do a substantial amount of that interaction at thirty miles an hour.

⚠ It does not have to be that way, and Chapter 4 is a detailed account of exactly how to build a working setup that does not require it — including the honest arithmetic on what it costs you in expired offers, which is a real number and much smaller than people assume.

⚠ Rule 2: You decide before you drive.

⚠ The accept-or-decline is made stationary, on numbers you can actually calculate, and once the car is moving the decision has already been made.

This rule is the practical enforcement mechanism for rule one and an economic discipline in its own right. Chapters 5 and 21.

⚠ Rule 3: The food is yours from the moment you take it.

⚠ Temperature, time, seal, and where it sits. Nobody is checking, the customer cannot tell what happened between the counter and the door, and the consequences of getting it wrong are somebody else's illness. Chapter 6.

⚠ Rule 4: You count the miles, not the deliveries.

⚠ The odometer is the only honest instrument in this business. Everything else is one half of a calculation. Chapter 7.

How to read this book

Read the introduction and Chapters 4 to 7 first, in order. Those are the four rules and everything after assumes them.

⚠ Then read Chapter 29 before you take another shift. It is the cost-per-mile build, it takes about forty minutes with a folder of receipts, and it produces the single number that makes every offer decision in this book possible. Without it, Chapter 21 is theory.

⚠ Then Chapters 21 and 5 together, which are the offer decision and its discipline.

The rest can be read in order or used as reference. Part Two is the vehicle, compliance and records. Part Three is the work itself. Part Four is money and systems. Part Five is independent clients and growth.

The resources. Sixty-eight of them, at the back, in six parts. Working documents rather than illustrations. Every financial figure is blank and in U.S. dollars.

The prompts. One hundred and forty-eight, four per chapter, collected in the sixth resource file. ⚠ Not one of them asks a tool to make a driving decision, and not one contains a customer's address.

⚠ The warning markers throughout are load-bearing. They flag genuine risk — to you on the road, to somebody eating what you carried, to your vehicle, or to your ability to answer a question later. A reader skimming only the marked passages would get a fair summary of what actually matters here.

⚠ One thing this book will not do is tell you whether this work is worth doing.

That is a question with a real answer and the answer is different for different people in different markets with different cars, and it depends almost entirely on a number that most drivers have never calculated.

⚠ What this book will do is give you the arithmetic. After Chapter 29 you will know what a mile costs you. After Chapter 21 you will know what an offer is worth. And after a month of Chapter 7's odometer photographs you will know, for the first time, what you are actually earning.

⚠ Kit's spreadsheet was excellent. It was missing one column, and that column was the difference between a business and a slow sale of a car.

©2026 James Henderson / https://localhandyman.work

Chapter 1 — Understanding the Industry and What You Are Actually Selling

You are selling the depreciation of a vehicle, and the pay screen shows none of it.

⚠ You are not selling deliveries. You are selling the use of a vehicle, and the vehicle is the thing being consumed.

1.1 What you are actually selling

Four things, and only one of them is visible on the pay screen.

⚠ One: the depreciation of a vehicle you own. ⚠ The largest input and the one nobody invoices. Every mile you drive moves your car measurably closer to being worth nothing. Chapter 29.2.

Two: your time. ⚠ All of it — driving, waiting at counters, walking into buildings, sitting between offers. Not just the minutes an app counts as "active."

Three: fuel and consumables. The only cost that announces itself, which is why it is the only one most drivers can quote.

⚠ Four: risk. ⚠ You are on the road for hours at a time, at peak hours, in the dark, in weather, frequently while tired. That is a real cost even when nothing happens, and it is priced into nothing.

⚠ The pay screen shows you a payment for a delivery. It shows you none of the four things you actually put in.

What this reframing changes.

⚠ It makes cost per mile the central figure rather than dollars per hour. Chapter 36.3 ☐ ⚠ It makes the decline a legitimate business action rather than a lost opportunity. Chapter 5.5 ☐ ⚠ It makes the dead leg — the miles home with nothing in the car — a cost you can actually manage. Chapter 22.4 ☐

1.2 Who pays you, and who does not

⚠ Worth being precise about, because it explains most of the friction in this trade.

⚠ The structural fact: the party whose delays cost you most — the restaurant — is the only party in the transaction with no financial relationship to you at all. That is why the wait is the single most damaging unpriced cost in food delivery, and why Chapter 23 treats unassigning as a skill rather than a failure.

1.3 The four costs nobody shows you

⚠ 1. Depreciation. ⚠ Silent, per-mile, and it arrives once, years later, as a repair quote larger than the car's value. Chapter 29.2.

⚠ 2. Dead miles. ⚠ Every mile driven with nothing in the car. To the pickup, back from a delivery that ended somewhere with no demand, and repositioning. In many markets this is thirty to fifty per cent of all miles driven, and it is invisible because no app reports it. Chapter 22.4.

⚠ 3. Wait time. ⚠ Standing in a restaurant for eighteen minutes is unpaid on most platforms, and it is time in which your car is off, your earnings are zero, and your evening is passing. Chapter 23.2.

⚠ 4. Tax and reserves. ⚠ Every dollar that arrives is gross. A proportion of it belongs to a tax authority and a proportion of it belongs to your next car, and both need to leave the account before you feel wealthy. Chapters 19.5 and 36.6.

⚠ All four are real, all four are yours, and none of them appear anywhere on a platform's screen. The purpose of this book is to put them somewhere you can see.

1.4 What the trade competes on, and what it should

What drivers talk about. Which app pays best ☐ Which zone is busiest ☐ Hourly averages ☐ Promotions and bonuses ☐ Acceptance rate ☐ ⚠ Whose night was better ☐

⚠ Almost every one of those is a dollars-per-hour conversation, and dollars per hour is the number that hides the car.

What actually determines whether this works.

⚠ Your cost per mile, calculated from real figures. Chapter 29.1 ☐ ⚠ Your dollars per mile on accepted offers. Chapter 21.2 ☐ ⚠ Your dead mile proportion. Chapter 22.4 ☐ ⚠ Your unpaid wait time. Chapter 23.2 ☐ ⚠ Whether you drive tired. Chapter 26.2 ☐ ⚠ Whether you touch the phone while moving. Chapter 4 ☐

⚠ Two drivers in the same city on the same platform earning the same gross can be having completely different years, and the entire difference is in that list. None of it is visible from outside and almost none of it gets discussed.

1.5 Where the money actually is

⚠ And the largest single lever nobody uses: reducing dead miles. A driver who cuts dead miles from forty per cent to twenty-five per cent has increased earnings per mile by a fifth without earning a cent more. Chapter 22.4.

1.6 How AI Helps

Use it for. ⚠ Cost-per-mile arithmetic from your own figures ☐ ⚠ Analysing your own offer log for patterns ☐ Building the record templates ☐ Drafting customer messages ☐ Drafting the questions in Resource 2 ☐

Never use it for. ⚠ Deciding whether to accept an offer while driving ☐ ⚠ Anything containing a customer's address ☐ ⚠ Telling you what platforms pay in your market — it does not know ☐

1.7 Common Mistakes

⚠ Tracking dollars per hour and nothing else. Section 1.4.

⚠ Not counting dead miles. Section 1.3.

⚠ Treating wait time as part of the job rather than as a cost.

Spending gross earnings. Chapter 19.5.

⚠ Chasing a promotion whose miles do not work.

⚠ Comparing yourself to other drivers on a number that hides the car.

Believing fuel is the main vehicle cost. It is not. Chapter 29.2.

1.8 Safety, Legal, Ethical, and Professional Considerations

⚠ Establish, before your first shift, whether your insurance covers delivery work. Many ordinary policies exclude it entirely, and driving uninsured is not a risk to manage but a reason not to start. Chapter 16.2.

⚠ Establish whether any registration, licence, or permit applies to delivery work where you live, including for alcohol. Resource 3.

⚠ Establish whether any food handling requirement applies to you. Chapter 16.4.

⚠ Never accept work you cannot do safely because the arithmetic looks good. Chapter 8.2.

1.9 Do This Now

☐ ⚠ Photograph your odometer today. That is the start of your denominator. Chapter 7.1. ☐ ⚠ Write down the four costs from Section 1.3 and put the list where you plan your week. ☐ ⚠ Ask your insurer today whether delivery work is covered. Resource 19. ☐ Open Resource 3 with the three questions from Section 1.8. ☐ ⚠ Estimate, honestly, what proportion of your miles last week were dead. Chapter 22.4. ☐ Start Resource 15, the mileage log, this shift.

1.10 Chapter Summary

You are selling the depreciation of a vehicle, your whole time including the unpaid parts, fuel, and risk — and the pay screen shows none of them. Four costs are invisible by design: depreciation, dead miles, wait time, and the tax and reserves that leave every gross dollar smaller than it looked. The trade talks in dollars per hour, which is precisely the number that hides the car, while the things that actually determine whether this works are cost per mile, dollars per mile, dead mile proportion, unpaid waiting, fatigue, and whether you touch the phone while moving. And the largest unused lever in the business is cutting dead miles, which raises what you earn per mile without earning another cent.

Prompt 1 — My four costs, estimated

"Help me build a rough picture of what a week of food delivery driving actually costs me. My figures: [miles driven, hours online, hours actually waiting at pickups, fuel spent, gross earnings]. Calculate: fuel cost per mile; gross earnings per mile; gross earnings per hour online; and gross earnings per hour if I include waiting time as unpaid. Show the arithmetic. Do not estimate depreciation yet — just flag that it is missing and that it is likely the largest cost."

Prompt 2 — Dead mile estimate

"Here is a log of one shift: [for each delivery, the miles from where I was to the pickup, the miles from pickup to drop-off, and the miles from drop-off to my next pickup or home]. Calculate total miles, miles with an order in the car, and miles with nothing in the car. Give me the dead mile proportion as a percentage. Do not suggest how to reduce it yet."

Prompt 3 — Questions for my insurer

"Draft the questions I must get answered in writing from a vehicle insurer before starting food and grocery delivery work, covering: whether my current policy excludes delivery or commercial use; what cover I would need instead; whether cover differs between carrying food and carrying groceries; whether cover applies while I am logged into an app but have no order; whether it applies while an order is in the car; what happens to any claim if I was using a phone; and what evidence they would want after an incident. Phrase as questions only. Do not guess at any answers."

Prompt 4 — What I should be tracking

"I currently track these things about my delivery work: [list]. Compare that against a list of what actually determines profitability in this trade: cost per mile, dollars per mile on accepted offers, dead mile proportion, unpaid wait time, and hours driven while tired. Tell me which of those I am not capturing, and for each one, the simplest thing I could start recording tomorrow to capture it."

©2026 James Henderson / https://localhandyman.work