Introduction — The Statistic That Was Never True
Eleven hours, a rounded figure with its qualifier gone, and a verdict three months late.
Marek wrote articles for three agencies and two direct clients, at a per-word rate that had seemed reasonable when it was agreed.
The 2,400-word piece took eleven hours. Not because it was difficult, but because there was no outline. Marek researched for four hours, wrote for five, delivered it, and then received a note saying the client had expected a different angle — not a different topic, a different angle — which meant most of the research had been for the wrong article. The rewrite took two more hours and was not billable, because nothing in the arrangement said what a rewrite was. At the agreed rate the piece paid a little under what a competent evening's work should, and Marek had spent two evenings and part of a weekend on it.
That was the visible problem. Three others were quieter.
The first was a number. A piece published under a doctor's byline contained a statistic about how long a particular condition takes to resolve. Marek had taken it from a well-written article on a reputable-looking health site. That article had taken it from a press release. The press release had summarised a study — and had rounded, and had dropped the qualifying clause about which patient group it applied to. By the time it reached Marek it was a clean, quotable sentence with no hedging in it at all, and it was not true of the people who would read it. Nobody checked. It ran for a year under a real doctor's name, and the doctor had signed it off because doctors are busy and the writer is supposed to have checked.
The second was a handover. A piece delivered in March was published in July, retitled to something Marek would not have chosen, with two paragraphs added by somebody at the client who wanted a product mentioned. One of those paragraphs made a claim that the sourcing in the rest of the article did not support, and it sat directly above a properly cited section, which made the whole piece look carelessly researched. Marek found out by seeing it in a search result.
The third was a verdict. At the end of the quarter one agency reported that the articles "weren't performing". The measurement was ranking position at ninety days, on a site with four hundred other pages, half of which had been rewritten by somebody else during the same period, tracked by a tool whose traffic figures disagreed with the client's own analytics by roughly a third. There was no baseline. There was no record of what had been published when. And the piece that had actually earned the client an enquiry — Marek later learned, in passing — was one that ranked nowhere at all and had been forwarded by hand between two people who worked in the industry.
By month five Marek was working most evenings, being paid by the word for a job whose cost is research, and being judged on a number arriving three months late about a page that had been changed after delivery.
Nobody behaved badly. The health site was reputable. The press release was written by a person doing their job. The client who added a paragraph was proud of the product. The agency was reporting what its tool reported.
Every part of it was ordinary, and all of it was expensive.
What this book is about
This book is about the four things underneath that story, because every one of them is a property of the trade rather than a mistake anybody made.
Every sentence is a claim in somebody else's name. You are asserting facts about the world, under a byline that belongs to a person or a business who probably has not checked them. The chain from a study to a press release to an article to your draft loses a qualifier at every step, and the version that reaches you is the cleanest and the least true. This is why fact-checking is a separate pass in this book rather than a step in editing, and why Chapter 24 exists at all.
Research is unbounded and the fee is not. There is always another source, another paper, another person who would know. Nothing in the work itself tells you when to stop — the topic does not end, the reading does not end, and at eleven at night the next source always looks worth opening. The only thing that ends research is a decision made in advance, which is why the outline is the most important document in this trade.
You lose control at handover. The piece will be retitled. It will be edited by somebody who does not know which sentence the citation was attached to. A paragraph will be added. It will be published four months late, or split into two, or not published at all. What appears under that byline is not what you delivered, and you will be judged on the difference.
The verdict arrives late and belongs to somebody else. Rankings settle over months. Traffic tools disagree with each other and with the client's own analytics. Everything else on the site moved during the same period. And the piece that actually did something for the business may be the one that ranked nowhere and was forwarded by hand.
What this book will give you
An outline that lists the claims before they are written and identifies the source for each, signed off before drafting begins — which bounds the research, prevents the structural rewrite, and turns "a different angle" from a free rewrite into a scope conversation. A research budget stated in hours, agreed in the brief. An eleven-point fact-check run as a separate pass, on every number, date, name, quotation and link. A source library that makes the fourth piece in a subject cheaper than the first. A style guide that makes a voice checkable rather than a matter of taste. A delivery note that survives the handover and says what was checked, what was sourced, and what must not be changed without re-checking. And a report that states what you can honestly be measured on, agreed before the first piece rather than after a bad quarter.
The economic unit is revenue per finished piece — what a commission returns against every hour it consumed, research included.
The leading indicator is research hours per thousand words, which falls sharply once a source library exists and does not fall at all without one.
The master variable is the outline sign-off rate — the share of pieces drafted only after a written outline was approved. It is the single number that predicts whether a month is profitable.
And the risk indicator is the count of published claims you could not source, tracked cumulatively, because that is the one that ends up under somebody else's name.
Who this is for
Somebody paid by the word for a job whose cost is research. Somebody who has rewritten a piece because the angle was wrong and not charged for it. Somebody who has cited a statistic they found in an article rather than in the study. And anybody who has been told their work "isn't performing" by a number measured three months late on a page that was edited after delivery.
There is one more person in this, and they never asked to be here: the reader. They will act on what you wrote — take the advice, believe the number, follow the link — and they will do it on the strength of a name that is not yours. Chapter 5, Chapter 9 and Chapter 24 exist for them.
©2026 James Henderson / https://localhandyman.work