Introduction — Twenty-Nine Children and One Wrong Answer

One character in an answer key, found mid-lesson, and a listing that never recovered in two years.

⚠ Month fourteen. A teacher Marin has never met printed a fractions worksheet on Monday night, handed it out on Tuesday morning, and worked through it with the class. Item eleven on the answer key was wrong. Twenty-nine children found it at the same moment, and the person standing at the front had to explain it.

---

What that error actually cost

⚠ It was one character. It took four minutes to fix and about ninety seconds to have caused.

⚠ Consequence	⚠ Detail
⚠ Time to correct the file	⚠ ⚠ 4 minutes
⚠ Reviews within a week	⚠ ⚠ 3, all one star
⚠ Sales of that resource, next quarter	⚠ ⚠ Down about 70%
⚠ Sales of that resource, two years later	⚠ ⚠ Never fully recovered
⚠ Teachers who told Marin	⚠ ⚠ 2 of the 3, in detail

⚠ "They were not angry with me. One of them apologised for mentioning it. She had stood in front of her class and been wrong because of something I made, and she apologised." Section 7.2.

⚠ That is the whole difference between this business and every other digital product in this series: the buyer is not the user, and the failure happens in public, to somebody else. Rule One.

---

The number that had never been calculated

⚠ Marin's first year: sixty worksheets, about four hours each, roughly two hundred and forty hours. Section 30.2.

⚠ Line	⚠ Year one
⚠ Resources published	⚠ ⚠ 60
⚠ Hours each	⚠ ⚠ ~4
⚠ Class time each covered	⚠ ⚠ ~20 minutes
⚠ Coverage per production hour	⚠ ⚠ ~5 minutes
⚠ Answer keys checked by solving them	⚠ ⚠ None

⚠ Five minutes of classroom time for every hour Marin worked. Section 30.1.

⚠ "I measured my year in how many things I had made. Sixty sounds like a great deal until you notice it is twenty hours of lessons." Section 1.3.

---

The unit that changed the arithmetic

⚠ Month nineteen. Marin built one six-lesson unit instead of six worksheets. Section 3.2.

⚠ Measure	⚠ 60 worksheets	⚠ One 6-lesson unit
⚠ Hours	⚠ ⚠ ~240	⚠ ⚠ ~22
⚠ Class time covered	⚠ ⚠ ~1,200 minutes	⚠ ⚠ ~270 minutes
⚠ Coverage per production hour	⚠ ⚠ ~5 min	⚠ ⚠ ~12 min
⚠ Price	⚠ ⚠ Low, each	⚠ ⚠ Several times one worksheet
⚠ Revenue, year two	⚠ ⚠ Moderate, spread over 60	⚠ ⚠ More than all sixty combined

⚠ "The unit is what a teacher actually wants, because what they are short of is not worksheets. It is planned lessons." Section 2.7.

⚠ A teacher buying a worksheet has still got to plan the lesson around it. A teacher buying a unit has bought back a fortnight of Sunday evenings. Section 17.4.

---

What the error-report rate showed

⚠ After month fourteen, Marin started counting reported errors per hundred sales. Section 7.4.

⚠ Period	⚠ Error reports per 100 sales	⚠ What changed
⚠ Year one	⚠ ⚠ ~4.1	⚠ ⚠ Nothing checked
⚠ Year two, first half	⚠ ⚠ ~1.8	⚠ ⚠ Solve-it-yourself pass added
⚠ Year two, second half	⚠ ⚠ ~0.9	⚠ ⚠ Print test added
⚠ Year three	⚠ ⚠ ~0.3	⚠ ⚠ Second solver on assessments

⚠ The rate fell for two terms before review scores rose, and review scores rose two terms before revenue did. Section 24.4.

⚠ "It is the only number in this business that moves before anything else does, and it costs nothing to count." Section 35.5.

---

The alignment that quietly expired

⚠ Month thirty-one. Marin discovered that twenty-three resources claimed alignment to codes from a superseded version of a framework. Section 8.4.

⚠ Line	⚠ Figure
⚠ Resources claiming alignment	⚠ ⚠ 41
⚠ Aligned to a superseded version	⚠ ⚠ 23
⚠ Version recorded at the time of making	⚠ ⚠ None
⚠ Hours to re-check all 41	⚠ ⚠ ~19
⚠ Hours it would have taken with a register	⚠ ⚠ ~3

⚠ "I had made a claim about a document I do not control, and I had not written down which version of it I was claiming." Section 8.3.

⚠ Nobody had complained. A claim that has quietly stopped being true produces no signal at all until somebody who checks these things notices. Section 8.9.

---

The seasonality nobody plans for

⚠ Period	⚠ Share of annual revenue
⚠ The six weeks around term start	⚠ ⚠ ~41%
⚠ The rest of the autumn term	⚠ ⚠ ~22%
⚠ Spring term	⚠ ⚠ ~24%
⚠ The long summer break, excluding term-start	⚠ ⚠ ~13%

⚠ Nearly half a year's income arrives in six weeks, and the work that produces it was done four months earlier. Section 31.1.

⚠ "I built resources in the season I was selling them and sold nothing in the season I should have been building." Section 32.2.

---

What Marin changed

⚠ Change	⚠ Month	⚠ Effect
⚠ Solve every key as a student would	⚠ 15	⚠ ⚠ Error rate 4.1 → 1.8
⚠ Print test on ordinary equipment	⚠ 17	⚠ ⚠ Error rate 1.8 → 0.9
⚠ Units instead of single worksheets	⚠ 19	⚠ ⚠ Coverage 5 → 12 min/hour
⚠ Alignment register with versions	⚠ 31	⚠ ⚠ Re-checks: 19 h → 3 h
⚠ Build in spring, publish before term	⚠ 33	⚠ ⚠ Term-start share used, not wasted
⚠ A list, so corrections reach somebody	⚠ 34	⚠ ⚠ Repeat purchases, and reachable buyers

⚠ Four of those six cost nothing at all. Two of them cost under an hour each. Section 32.4.

---

Where the money was, at month forty-four

⚠ Layer	⚠ Share of revenue	⚠ Started
⚠ Units and multi-lesson plans	⚠ ⚠ 44%	⚠ Month 19
⚠ Bundles of existing resources	⚠ ⚠ 21%	⚠ Month 26
⚠ Single worksheets	⚠ ⚠ 18%	⚠ Month 1
⚠ Assessments	⚠ ⚠ 11%	⚠ Month 29
⚠ Editable versions	⚠ ⚠ 6%	⚠ Month 37

⚠ The sixty worksheets that took a year are under a fifth of the revenue. The bundles took about fourteen hours to assemble. Section 3.9.

---

What this book is

⚠ Thirty-seven chapters, sixty-eight fill-in resources, and a hundred and forty-eight AI prompts — four per chapter, none of which will write your answer key.

⚠ It is built on four rules and on one number: coverage minutes per production hour, which is what a teacher is actually buying and what no marketplace reports. Section 30.2.

⚠ The leading indicator is the error-report rate per hundred sales, which falls two terms before review scores rise and four before revenue does. Section 7.4.

⚠ And the argument running through it is one sentence: you are not selling a teacher a worksheet, you are selling them the ability to stand in front of thirty children on Tuesday morning and have it work first time, unsupervised, with no chance to fix it. Section 6.1.

---

How to read it

⚠ Chapters 5 to 8 first, in order. They are the four rules.

⚠ Then Chapter 21, before you publish anything with an answer key. Section 21.1.

⚠ Then Chapter 23, before you publish anything a teacher will print. Section 23.2.

⚠ Then Chapter 15 before you use a single piece of clip art commercially, because that is the licence people discover they did not have. Section 15.3.

⚠ Everything else is procedure, and the procedures are only worth having because they are cheaper than the Tuesday mornings that produced them. Section 34.8.

---

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

Chapter 1 — Why Teacher Worksheets and Lesson Plans Work as a Side Hustle

You sell prepared classroom time, not paper. Sixty worksheets is twenty hours of lessons.

⚠ This business has a real advantage that almost nobody states plainly: your customer is short of exactly one thing, and it is the thing you can make. It also has a risk almost nobody states plainly: your product is used, in public, by somebody who cannot fix it.

1.1 What you are actually selling

⚠ Not a worksheet. Prepared classroom time. Section 17.4.

⚠ A teacher's scarcest resource is Sunday evening, not paper. Section 2.7.

⚠ A single worksheet still leaves them planning the lesson around it.

⚠ A unit hands back a fortnight of planning, which is why it is worth several times as much. Section 3.2.

⚠ Which means the unit of value here is minutes of class time covered, and everything in Chapter 30 follows from that. Section 30.1.

1.2 The four rules, in short

⚠ Each is a structural fact rather than advice. Resource 3.

⚠ The buyer is not the user. Chapter 5.

⚠ It has to work first time, in a room, with no chance to fix it. Chapter 6.

⚠ An error is found by thirty people at once. Chapter 7.

⚠ Alignment is a claim about somebody else's framework. Chapter 8.

⚠ Every difficult decision in this book resolves to one of those four. Section 1.9.

1.3 Why classroom time is the unit

⚠ Because that is what the buyer is short of and what they are actually purchasing. Section 2.7.

⚠ Twenty minutes covered for four hours of your work is five coverage-minutes per hour. Section 30.2.

⚠ Two hundred and seventy minutes covered for twenty-two hours is about twelve.

⚠ The second is not twice as good; it is a different business. Section 3.2.

⚠ And counting resources published rather than minutes covered is how somebody spends a year producing twenty hours of lessons. Section 1.10's scenario.

1.4 Who buys, and when they buy

⚠ Teachers, mostly at their own expense, in bursts. Section 31.1.

⚠ Roughly two fifths of a year's revenue arrives in the weeks around term start. Section 31.1.

⚠ They buy on a Sunday for a Tuesday, which is why the preview matters. Section 5.8.

⚠ They buy under time pressure, which is why the listing must answer their question in ninety seconds. Section 2.1.

⚠ And they buy again from somebody whose last resource worked, which is the whole of Chapter 25. Section 25.4.

1.5 What it costs to start

⚠ Very little money, and a great deal of care. Section 14.2.

⚠ A word processor or a design tool you may already have.

⚠ A printer, which is not optional in this business. Section 23.2.

⚠ Commercially licensed clip art and fonts, which is the one real early cost. Section 14.4.

⚠ And a marketplace, which takes a percentage rather than a fee. Section 16.5.

1.6 What it costs to keep

⚠ The line that decides whether this is a business. Section 17.3.

⚠ Answer keys stay correct, forever. Rule Three.

⚠ Alignment claims stay true through framework revisions. Rule Four.

⚠ Every reported error is corrected quickly, in every version. Section 7.6.

⚠ And the whole catalogue is reviewed each term, which is an hour or two you will not want to spend. Section 26.3.

1.7 Who does well at this

⚠ Four traits, and none of them is design skill. Resource 1.

⚠ Somebody who has taught the grade and the lesson. Section 4.5.

⚠ Somebody who will solve every item on their own key. Section 21.2.

⚠ Somebody who will print it in black and white on both sides before publishing. Section 23.3.

⚠ Somebody who thinks in units rather than in worksheets. Section 3.2.

1.8 Who does badly

⚠ Also four, and they are more common. Section 4.9.

⚠ Somebody making things they would have enjoyed teaching. Section 4.3.

⚠ Somebody who counts resources published. Section 1.3.

⚠ Somebody who treats the answer key as the last five minutes. Section 21.1.

⚠ And somebody who claims alignment without recording which version. Section 8.3.

1.9 What this book measures

⚠ One economic unit and one leading indicator, both counted by hand. Section 35.1.

⚠ Coverage minutes per production hour — what a teacher is actually buying. Section 30.2.

⚠ The error-report rate per hundred sales — which moves before anything else. Section 7.4.

⚠ Neither appears on any marketplace dashboard.

⚠ Marin's coverage went from five to twelve, and the error rate from about four per hundred to under one, and neither change cost money. Section 30.10's scenario.

1.10 Chapter summary

⚠ You are selling prepared classroom time, not paper.

⚠ Coverage minutes per production hour is the unit. Counting resources is not.

⚠ Two fifths of the year arrives in six weeks, from work done months earlier.

⚠ Answer keys and alignment claims must stay true forever.

⚠ Neither number this book uses is on any dashboard.

---

A realistic scenario

Marin's first year, counted two ways.

⚠ Measure	⚠ Figure
⚠ Resources published	⚠ ⚠ 60
⚠ Hours	⚠ ⚠ ~240
⚠ Class time covered, in total	⚠ ⚠ ~20 hours
⚠ Coverage per production hour	⚠ ⚠ ~5 minutes
⚠ Answer keys solved before publishing	⚠ 0
⚠ Resources print-tested	⚠ 0

⚠ "Sixty resources sounds like a productive year. Twenty hours of lessons sounds like a fortnight, and they are the same thing." Section 1.3.

What the year actually produced:

⚠ Outcome	⚠ Detail
⚠ Revenue	⚠ ⚠ Modest, spread thinly
⚠ Error reports	⚠ ⚠ ~4.1 per hundred sales
⚠ Reviews mentioning a mistake	⚠ ⚠ Several
⚠ Resources needing correction later	⚠ ⚠ 11 of 60
⚠ Alignment versions recorded	⚠ ⚠ None

⚠ "Eleven of my sixty needed fixing, and I found that out one angry Tuesday at a time." Section 7.4.

The traits, honestly assessed:

⚠ Trait	⚠ Did Marin have it in month one?
⚠ Had taught the grade and the lesson	⚠ ⚠ Yes
⚠ Would solve every item on the key	⚠ ⚠ No
⚠ Would print-test before publishing	⚠ ⚠ No
⚠ Thought in units rather than worksheets	⚠ ⚠ No

⚠ "One of four, and it is the one that cannot be learned quickly. The other three are habits and they took me eighteen months to acquire." Section 34.8.

And what the second year showed:

⚠ Measure	⚠ Year one	⚠ Year two
⚠ Resources published	⚠ ⚠ 60	⚠ ⚠ 22
⚠ Coverage per production hour	⚠ ⚠ ~5 min	⚠ ⚠ ~11 min
⚠ Error reports per 100 sales	⚠ ⚠ ~4.1	⚠ ⚠ ~1.3
⚠ Revenue	⚠ ⚠ Modest	⚠ ⚠ Substantially higher

⚠ Fewer than half as many resources, and more of everything that mattered. Section 30.7.

⚠ Marin's note: "I spent a year proud of a number that measured how many evenings I had used up. Nobody had ever suggested measuring how many lessons I had covered, and that is the only number a teacher is actually buying."

---

AI prompts for this chapter

Prompt 1 — Is This Business a Fit for Me?

"Assess honestly whether this business suits me, and do not be encouraging. ⚠ Ask me: have I actually taught the grade and lesson I want to make resources for; will I solve every item on my own answer key as a student would; will I print it in black and white on both sides before publishing; and do I think in units or in single worksheets? ⚠ Then ask about my hours in my worst term. ⚠ Tell me plainly which of these I fail, and be clear that the first is the one that cannot be learned quickly."

Prompt 2 — What Am I Actually Selling?

"Help me understand what a teacher is actually buying from me. ⚠ Not paper — prepared classroom time. A single worksheet still leaves them planning the lesson around it; a unit hands back a fortnight of Sunday evenings. ⚠ I will describe what I plan to make. ⚠ Tell me how many minutes of class time it covers and how many hours it will take me to build. ⚠ Then calculate coverage minutes per production hour and tell me plainly whether I am building worksheets or lessons."

Prompt 3 — Count the Real Cost of Starting

"Help me count what this actually costs before I begin. ⚠ Money first: a design tool, a printer — which is not optional here — and commercially licensed clip art and fonts, which is the one real early cost. ⚠ Then HOURS: objective and alignment, building student pages, building the answer key, teacher instructions, print testing, and the listing. ⚠ I will describe my first resource. ⚠ Give me an honest total and tell me how many evenings that represents."

Prompt 4 — What Will This Resource Cost Me to Keep?

"Estimate the ongoing obligation I am taking on. ⚠ Ask me what my resource depends on: a standards framework, a curriculum version, named tools, prices, or anything that could be revised. ⚠ Then remind me what must stay true forever: the answer key, and any alignment claim. ⚠ Propose a review interval and estimate the annual maintenance hours. ⚠ Be honest that this work is unpaid and produces no signal at all until a teacher is standing in front of a class with something wrong."

---

⚠ AI checkpoint for this chapter

One — did it count resources published? Count minutes of class time covered.

Two — did it call this passive? Keys and alignment claims must stay true forever.

Three — did it treat the answer key as an afterthought? It is the highest-risk object here.

Four — did it ignore who uses it? A teacher buys. Thirty children use it.

---

Do This Now

1. Work out the coverage minutes per production hour of your last resource.

2. Count how many of your published answer keys you have solved yourself.

3. Print one of your worksheets in black and white, on both sides.

4. Write down which framework version your alignment claims refer to.

5. Ask yourself whether you are building worksheets or lessons.

---

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