Introduction

Ines's clinic landing page: seven rounds instead of three, a verdict drawn from a hundred and eighty visitors, a price that never fitted the audience, and a headline claim nobody could source. The four rules follow from it.

Ines took the landing page job because it looked small.

One page. A founder with a piece of software that helped small clinics handle their appointment reminders. The existing page was four years old, written by the founder, and converted at a number nobody had ever measured because nobody had set up the measurement. The brief was three sentences long and the last one said make it convert better.

Ines quoted for three rounds of revisions. It went to seven.

Round four is where it went wrong, and it went wrong for a reason that had nothing to do with the writing. The founder's business partner — who had not been in the kickoff call, had not seen the brief, and had a different idea about who the software was for — joined the thread and said the page was aimed at the wrong people. Not wrong in a way anybody could resolve by editing a sentence. Wrong at the level of who the reader was. Rounds five, six and seven were two people arguing about their own business through Ines's document, and Ines was paid for three.

Then the page went live and, two weeks later, the founder wrote to say it had not worked.

One hundred and eighty visitors. Four sign-ups. The old page, they said, had done better — though when Ines asked over what period and against what number, no answer came back, because there had never been a number. One hundred and eighty visitors is not a result. It is not even the beginning of a result. Two more sign-ups either way would have moved the rate by more than a percentage point, and two more sign-ups is noise. Ines knew that. The founder did not, and by then it was too late to say so, because saying it after a bad number sounds like an excuse and saying it before sounds like professional advice.

And underneath all of that was the thing that actually mattered, which nobody had discussed at any point.

The software was priced at a level that made sense for a practice with several clinicians and no sense at all for the single-room practices the page was written to attract. The founder had picked the audience because those were the people who emailed them. The founder had picked the price because that was what the last competitor charged. Nobody had put those two decisions next to each other. Ines had spent thirty-one hours writing the most persuasive possible case for a purchase that, for most of the people reading it, was the wrong purchase at that price.

Copy does not fix that. Copy has never fixed that.

There was one more thing, and it is the one that could have caused real trouble. The headline Ines eventually wrote — after the partner's fourth rewrite, in a version that was not entirely Ines's — claimed the software cut admin time in half. That number had come from a testimonial on the old page. The testimonial had come from a customer who said it three years earlier in an email. Nobody had measured anything. The claim went live above the fold on a page selling to healthcare practices, and when Ines asked where the number came from, the founder said they thought it was from a case study, and then could not find the case study.

That is the trade. Not the writing. The writing is the part you can do at eleven at night with a cup of tea.

The four rules

Everything in this book comes out of four facts. They are not tips. They are the conditions the work happens under, and every checkpoint in the following thirty-seven chapters traces back to one of them.

One — you are paid to change behaviour, so the ethics are in the sentence rather than beside it.

An article informs. A social post entertains or reminds. Copy is written to move somebody from not-buying to buying, and it is judged on whether it did. That is a different job, and it carries a different weight, because every technique that raises response is one notch away from a technique that deceives. Urgency becomes a fake deadline. Scarcity becomes an invented limit. Social proof becomes a testimonial nobody said. Specificity becomes a number nobody measured. There is no bright line in the craft itself; the line is a question you have to ask deliberately, about every claim, every time: would this survive the reader finding out everything? Nobody else in the transaction is going to ask it. The client wants sales. The agency wants the page shipped. The reader is not in the room.

Two — the offer decides the outcome, and the offer is not yours.

The price, the product, the audience and the alternative the reader is actually considering account for most of what happens on any page. Copy accounts for a smaller share than anyone in this industry likes to admit. When the offer is right, ordinary copy works. When the offer is wrong, excellent copy produces a slightly better-attended failure. You will be hired to fix conversion and handed a problem that is not a conversion problem, and you will be judged on a variable you were never allowed to touch. Most of the skill in this trade is recognising that before you quote, saying it in a way that does not lose you the job, and writing it down when the client declines to change anything.

Three — everything looks testable and almost nothing is tested.

This is the trap that separates copy from every other writing trade. An article's result is obviously murky, so nobody pretends otherwise. A landing page has a conversion rate, updated hourly, on a dashboard the client can see. It looks like evidence. At the traffic volumes most clients have, it is not — it is a small number divided by another small number, moving around for reasons that have nothing to do with the words. Pages get killed on a hundred and eighty visitors. Winners get rolled out on a coincidence. Someone runs a "test" while also changing the price, the ad, the audience and the button colour. You will be handed conclusions drawn from noise, and the only defence is to say what a number can and cannot support before it exists rather than after.

Four — everybody believes they can write, so the copy is decided by seniority.

Nobody rewrites a developer's function in a meeting. Everybody rewrites a headline. Copy is the one deliverable where every person in the room has an opinion, holds it confidently, and has no way to tell taste from evidence — including, sometimes, you. The predictable result is that the version that ships is the one preferred by whoever is most senior, usually rewritten to be more about the company and less about the reader, and it is that version whose performance is attributed to you. The only referee that works is a document agreed before the writing started: a voice guide, a message map, and a research file full of the customer's own words. Not because it wins every argument, but because it changes the argument from I prefer this to this is what we agreed and here is why.

What this book is built on

The economic unit is revenue per approved asset — what a landing page, an email sequence, a voice guide or a set of ads actually returned against every hour it consumed, research and revisions included. Not per word. Per-word pricing prices the research at zero, and the research is the job.

The leading indicator is hours per approved asset, tracked by asset type and by client. It falls steeply within a client as their voice guide and research bank mature, and does not fall at all across new clients. A flat trend over a year means you are starting from scratch every time.

The master variable is the brief-completion rate — the share of projects where the reader, the offer, the price, the permitted claims, the approvers and the measure of success were all established in writing before a word of copy was written. It is the single number that predicts a profitable month, because the unbilled hours in this trade almost all come from something that was never agreed.

The risk indicator is the count of claims you have published that the client cannot substantiate, tracked cumulatively. It should be zero. It is the one that ends up in front of a regulator, or in front of somebody who bought on the strength of it.

How to use this book

Read Part One in order. It is the part that decides whether you should do this at all, and Chapters 5 through 8 are the four rules in full.

After that, use it as a manual. The resources are the working documents — the brief question set, the claims and proof register, the copy audit, the rate card, the profitability sheet. The prompts are at the end of every chapter, four each, and the ⚠ marks the clause that stops the specific failure that prompt is exposed to. When you adapt one, that is the clause to keep.

Every figure is blank because every figure depends on things this book cannot know. Fill them in from your own log.

There is somebody at the end of every piece of copy you will ever write, deciding whether to spend money they may not have much of. They are the reason for the claim check, the proof register and Chapter 5. They are also, in the end, the only reliable route to work that keeps coming back.

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

Chapter 1 — What Freelance Copywriting Actually Is

The job is paid decision-easing for one specific reader; you sell judgement, research and sentences in the order clients value them least to most.

1.1 The job is not writing sentences

Ask ten people what a copywriter does and nine will describe typing.

The typing is real, and it is somewhere between a tenth and a fifth of the hours on a healthy project. The rest is finding out what the reader already believes, what the offer actually is, what may honestly be said about it, what the client has been telling themselves that is not true, and what the piece is competing against in the reader's head. Then it is cutting — because the first draft of anything persuasive is always too long, too polite and too much about the company.

A useful way to hold it: a copywriter is paid to make a decision easier for one specific person who has not asked to be sold to. Everything else follows.

That framing does two things immediately. It tells you the reader is the subject rather than the client, which is the whole of Chapter 5. And it tells you that if the decision is genuinely wrong for that reader — wrong price, wrong product, wrong person — no amount of writing makes it easier honestly, which is the whole of Chapter 6.

It also explains why this trade sits so awkwardly next to the neighbouring ones. Article writing asserts facts to inform. Social content maintains a presence. Email marketing operates a permission asset. Copywriting produces a piece of writing whose entire purpose is that somebody does something they were not going to do. Same keyboard. Different obligation.

1.2 What you are actually selling

You are selling three things, in this order of value, and clients rank them in the opposite order.

Judgement about what to say. Which of eleven possible things about a product is the one that moves this reader. Which objection to answer first. Which proof point carries weight and which sounds like everybody else. This is where the money is and it is almost invisible on the page.

Research that turned into language. The customer's own words, taken from reviews, support tickets, sales calls and interviews, used instead of the words the company uses internally. Almost every piece of copy that outperforms does it by sounding like the reader rather than the boardroom.

Sentences. The craft. Rhythm, specificity, the discipline of cutting. It matters — but it is the layer clients think they are buying and the layer they most confidently rewrite.

⚠ A client who believes they are buying sentences will price accordingly and revise accordingly. Chapter 4.7 covers the conversation that resets this, and it happens before the quote or it does not happen.

1.3 What a project actually contains

Take a mid-sized landing page. Here is where the hours actually go on a project run properly.

⚠ Research and structure together are more than a third of the project and appear nowhere in the deliverable. They are the first thing a client asks you to skip and the reason projects go over when you agree.

1.4 The four rules

The whole book runs on four, set out in full in Chapters 5 to 8. ⚠ Resource 3.

Rule one — you are paid to change behaviour, so the ethics are in the sentence rather than beside it. Every technique that raises response is one notch from a technique that deceives, and nobody else in the transaction is going to draw the line.

Rule two — the offer decides the outcome, and the offer is not yours. Price, product, audience and alternative account for most of the result. You will be judged on a variable you do not control.

Rule three — everything looks testable and almost nothing is tested. Small-business traffic produces numbers that feel like evidence and are noise, and conclusions get drawn from them anyway.

Rule four — everybody believes they can write, so the copy is decided by seniority. The version that ships is often the one preferred by whoever outranks the room, and its performance is attributed to you.

Every checkpoint in this book traces to one of the four. When something in a later chapter looks fussy, it is usually rule two or rule four in disguise.

1.5 The five things a client cannot do themselves

This trade exists because of five specific incapacities, and knowing them tells you what to sell.

They cannot see their own offer from outside. The founder who has explained the product four hundred times cannot hear how it lands on somebody hearing it once. This is not stupidity — it is the unavoidable effect of knowing your own thing too well.

They cannot write in their customer's language. They write in their industry's language, because that is the language they think in all day. The gap between the two is where most conversion lives.

They cannot cut. Everything on the page is there because somebody internally cared about it. Removing it feels like betraying a colleague.

They cannot decide what to leave out at the level of argument. A page that says nine things says nothing. Choosing the one thing requires being willing to be wrong in front of the client, which an employee usually is not.

They cannot say the offer is the problem. Chapter 6. An internal person who says the price is wrong is making a career move. You are making a professional observation, and that is the single most valuable thing an outsider brings.

1.6 Where the hours actually go

Across a working month, in a business that has been running long enough to have repeat clients:

⚠ Roughly two hours in five are spent on things a client does not think they are paying for. Price per asset rather than per word or per hour and this becomes your problem to manage rather than your problem to justify. Chapter 17.2.

1.7 Who is already doing this

Four groups, and you compete with each differently.

Agencies. They win on capacity, process and being a safe choice for a nervous marketing director. They lose on price, on speed for small jobs, and on the fact that the person writing is often three levels below the person who sold it.

Career freelancers with a decade of controls behind them. They win on evidence and specialism. They lose on availability. Many of them will not take a project under a fee level that leaves a large market open.

Cheap volume services. They win on price and turnaround. They do not do research, they do not interrogate offers, and their output reads like it was written by somebody who has never met the reader — because it was.

The client themselves, or their marketing coordinator. By far your largest competitor. Most small businesses write their own copy and always will. You are not competing to be better than them at typing; you are competing to be worth the fee against doing nothing, which is a different argument.

⚠ The client's real alternative is usually "leave the page as it is." Chapter 21.6 covers reading the reader's alternative; the same discipline applies to your own sales conversation.

1.8 What makes somebody good at it

Not what most people expect.

Curiosity about strangers. The people who are good at this genuinely want to know why somebody nearly bought and did not. If customer interviews sound like a chore rather than the interesting part, this trade will be uphill.

Comfort with being disliked briefly. Telling a founder their price is wrong for their audience is uncomfortable, necessary, and the thing they will thank you for later — if they ever do.

Willingness to cut your own best sentence. Most drafts are improved by deleting the line the writer is proudest of.

Tolerance for ambiguous verdicts. Rule three means you will rarely know for certain that something worked. If you need scoreboard confirmation to feel competent, this will grind.

Ordinary reading stamina. Two hundred reviews, forty support tickets and three sales-call recordings is a normal week's input on one project.

Speed of typing appears nowhere on that list.

1.9 What the first year actually looks like

Roughly, and with the obvious caveat that yours will differ.

Months one to three. No paid work for most of it. You build a research method, write two or three assets end to end for businesses you invent or for a real business you do for free with permission, and you assemble a voice guide template and a claims register. Your hours per asset are terrible and that is expected.

Months four to six. First paid projects, usually small, usually from somebody who already knows you. Your first committee arrives around here and it will go badly. Your first "it didn't work" arrives too, on a sample far too small to mean anything, and Chapter 7 is what you wish you had read first.

Months seven to nine. You start refusing things. The brief becomes a document you send rather than one you wait for. Hours per asset drop noticeably on any client you have worked with twice, which is your first evidence that the research bank is an asset.

Months ten to twelve. Two or three clients returning, one of them possibly on a retainer. Enough log entries to compute revenue per approved asset by type, which tells you what to stop selling.

The month-two figure and the month-eleven figure for the same deliverable often differ by a factor of three. That is the whole business, and it comes from research reuse and from briefs, not from typing faster.

1.10 Chapter summary

Copywriting is paid decision-easing for one specific reader, and the writing is a minority of the work. You sell judgement first, research second and sentences third, in the order clients value them least to most. A properly run project spends more than a third of its hours on research and structure that never appear in the deliverable. Four rules govern everything: the ethics live in the sentence, the offer is not yours, almost nothing is genuinely tested, and seniority decides what ships. Your real competitor is the client doing nothing. The first year is a slow collapse in hours per asset, driven entirely by research reuse and written briefs.

AI prompts for this chapter

Prompt 1 — The offer interrogation before a quote

I am a freelance copywriter about to quote for a project. Here is what the client has told me: [PASTE]. Generate a list of questions I should get answered in writing before I quote, grouped into: the reader, the offer, the proof, the approvers, and the measure of success. ⚠ Do not write any copy, headlines or taglines. ⚠ Do not invent details about the product, the price or the audience — where information is missing, produce a question rather than an assumption. ⚠ Flag any question whose answer would change my price.

Prompt 2 — Separating a copy problem from an offer problem

Here is a client's description of their situation: [PASTE]. Sort the possible causes of their underperformance into three lists: things copy can change, things only the offer can change, and things that are neither. ⚠ Base every item strictly on what is in the pasted text — do not add assumed causes. ⚠ Do not tell me the copy is the cause unless the pasted text contains specific evidence for it. ⚠ Mark anything I would need to ask the client to resolve.

Prompt 3 — Estimating where the hours will go

For a copywriting project of this description — [PASTE] — produce an hour-block estimate across: brief and offer interrogation, audience research, message map, drafting, cutting, claim check, revisions, and delivery note. ⚠ Give ranges rather than single figures. ⚠ Do not supply a fee, a rate or a currency figure of any kind. ⚠ Name the specific factors in the brief that would push each block to the top of its range.

Prompt 4 — What this project is actually competing against

A client sells [PASTE PRODUCT AND PRICE] to [PASTE AUDIENCE]. List the alternatives that reader is realistically weighing, including doing nothing and continuing with their current arrangement. ⚠ Do not include alternatives you cannot justify from the pasted description. ⚠ Do not name specific competitor companies or make claims about them. ⚠ Mark which alternatives I would need to confirm with actual customers rather than assume.

⚠ AI checkpoint for this chapter

One — did the model write copy when I asked it to ask questions? Then it has skipped the part of the job that is worth money. Rerun with the constraint restated.

Two — did it fill a gap with an assumption instead of a question? Every invented detail about a reader or an offer becomes a claim later. Convert each one back into a question for the client.

Three — did it tell me the copy is the problem? It will default to that, because that is what the request implies. Rule two says check the offer first, and no model can see the offer.

Four — did it produce a rate, a fee or a benchmark figure? Delete it. It is drawn from published numbers that do not account for research depth or approval structure, and quoting from it will cost you a project.

Do This Now

1. Write down, in one sentence each, the three things you are selling — judgement, research, sentences — as they apply to the kind of client you expect to work with.

2. Take any landing page you can find for a small business and list the three questions you would need answered before you could improve it. Notice that none of them are about writing.

3. Start a blank project log with these columns: client, asset type, total hours, approvers, whether a written brief existed, and outcome. You will need a year of it.

4. Read one product's reviews for twenty minutes and write down ten phrases customers used that the company does not use anywhere on its own site.

5. Complete Resource 1 honestly, particularly the questions about being disliked briefly and tolerating ambiguous verdicts.

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