Introduction — The Deck That Took Thirty-Eight Hours
A deck quoted at twelve hours took thirty-eight, and none of the extra time was design.
Rowan quoted a sixty-slide investor deck at about twelve hours of work. It took thirty-eight.
Not one of the extra twenty-six hours was spent designing.
⚠ The content arrived in five batches over three weeks. The first batch was eleven slides of rough text and a promise that the financials were "nearly done." The second batch changed the order. The third arrived with a new market-size figure and no source. The fourth removed two sections that had already been designed. The fifth was the financials, four days before the meeting, in a spreadsheet with three tabs and no indication of which one was current.
Rowan had started designing after the first batch, because the client was in a hurry and it seemed helpful. By the time the deck was finished, thirty-one slides had been built and rebuilt at least twice, a chart had been redrawn four times against three different versions of the same numbers, and the whole thing had gone through seven review rounds.
⚠ The client was not difficult. They were not unreasonable, they paid on time, and they were pleased with the result. They simply did not have their content ready when they commissioned the design, and neither of them had said so out loud.
That is the shape of this trade. It is not really a design problem.
The three failures that follow from the same thing
Rowan had three more expensive lessons in the same year, and each of them came from the same root: a deck is not an artefact you make and hand over. It is a thing that gets performed, edited, changed and reused by people who are not you.
⚠ Month seven. A client supplied a chart showing a revenue trend. The y-axis started at a number Rowan did not choose, and a rise of about three per cent looked like a near-doubling. Rowan drew it as supplied, because it came from the client's own finance team and it was not Rowan's data. The chart went into a funding conversation. It was queried, awkwardly, in a room Rowan was not in.
⚠ Nobody had lied. The client had exported a default chart and passed it on. Rowan had redrawn it beautifully. And the version that went in front of investors was a picture Rowan had made.
⚠ Month nine. A finished deck, delivered on a Friday, looked perfect on Rowan's machine. The client opened it on Monday on a different computer that did not have the two typefaces Rowan had used. Everything reflowed. Text overflowed its boxes, headlines wrapped onto three lines, and a board saw a deck that looked broken. The design was fine. The file was not.
⚠ Month eleven. A client who had bought a fifty-slide deck later added forty slides themselves. None of them matched: different margins, different type sizes, five slightly different blues. The deck had been built as fifty individual slides rather than on a set of layouts, so there was nothing for the client to build on. They hired somebody else to rebuild all ninety.
What changed
⚠ Rowan's fix was not design skill, better software or a faster machine. It was a gate.
⚠ Nothing gets designed until the storyline is approved in writing: one sentence per slide, in order, saying what each slide claims. It takes an hour or two, it is cheap to change, and it is where the argument about structure belongs — because that argument is going to happen either way, and having it over sentences costs a fraction of having it over finished slides.
The other three changes were equally unglamorous. Every chart now gets a source and a date, and a supplied chart with an unusual axis gets a question before it gets drawn. Every deck is built on layouts rather than as loose slides. And every file is opened on a second machine before delivery.
By month twenty-six:
⚠ Rounds per deck fell from about 4.6 to about 1.4.
⚠ The proportion of jobs where the final content existed before design began rose from about a quarter to roughly three-quarters — because Rowan started asking, and started pricing the answer.
⚠ Revenue per build hour roughly doubled, on the same skills, in the same software.
⚠ And the revenue mix changed shape entirely: deck builds fell to about a third of income, with template systems, retained deck support, chart work and structure consulting making up the rest — most of which does not scale with the number of slides drawn.
The four rules
This book is built on four facts about the work. Each gets a chapter, and everything afterwards refers back.
Rule 1 — You are designing a talk somebody else has to give. Chapter 5.
⚠ The deck is performed, live, by a person who did not make it, often nervous, in a room you will never see. A slide is not a page — it is something an audience reads while somebody talks over it. Beautiful slides regularly make bad talks.
Rule 2 — A chart is a claim about somebody else's numbers. Chapter 6.
⚠ You did not gather the data. You frequently cannot verify it. And when you redraw it, the claim becomes something you made — which is why the source, the date and the axis are your business even though the numbers are not.
Rule 3 — The file will be edited by somebody who is not a designer. Chapter 7.
⚠ Whatever you deliver gets opened on a machine that does not have your fonts, edited by somebody who has never used a layout master, and extended by forty slides you did not build. A design that only works in your hands is a design that failed.
Rule 4 — You are paid for a deck and the content is not ready. Chapter 8.
⚠ This is the economics. The client commissions design before the argument is settled, the numbers are final or the story is agreed — and every hour of that unsettledness is paid for in rounds, by you, unless the structure is gated first.
The unit and the indicator
⚠ The economic unit is revenue per build hour — net revenue divided by every hour from first enquiry to handover, including the intake exchange, content chasing, the storyline, the rounds and the delivery check. Chapter 31.
⚠ Not per slide, because slides are not units of work. A dense chart slide and a section divider take wildly different amounts of time, and pricing by the slide means the hard ones subsidise nothing and the easy ones earn nothing.
⚠ The leading indicator is rounds per deck — how many review cycles a job went through before sign-off. Chapter 27.
⚠ It costs a tally mark. It responds immediately to the storyline gate. It moves before revenue does, and before a client goes quiet. And almost nobody in this trade counts it, which is why almost everybody in this trade quotes as though every job were their easiest one.
Who this book is for
Anybody designing presentations for businesses, agencies, founders, trainers or speakers — as a side hustle or on the way to something larger.
⚠ And particularly three people. Somebody whose jobs consistently take two or three times what they quoted. Somebody who has never counted the review rounds. And anybody who has redrawn a chart exactly as supplied without asking where the numbers came from.
How to read it
Read Chapters 5 to 8 first, in order. They are the four rules.
⚠ Then read Chapter 8 again before you quote anything, because it is the rule that decides whether the job was profitable — and it is decided before you open the software.
⚠ If you want the fastest return today: write the storyline for your next deck as one sentence per slide, send it, and do not design anything until it comes back approved. That single gate is the highest-return action in this book and it costs nothing but an hour.
©2026 James Henderson / https://localhandyman.work