Introduction — Nine of Seventy-Four

Seventy-four founding members, seventy-one at month fourteen, and nine of them the same people.

⚠ Rae opened with 74 founding members. Fourteen months later there were 71.

⚠ Nine of them were the same people.

---

What the number was doing

⚠ Rae ran a paid community for people who make things and sell them at markets — pricing, stock, pitches, the weekend logistics of it. $18 a month.

⚠ The launch went well by every measure anybody uses. 74 members in the first fortnight, mostly from a mailing list Rae had built over two years.

And then, month by month, the total held:

Month	⚠ Members
1	74
4	69
7	73
10	70
⚠ 14	⚠ 71

⚠ Rae described the community, to anybody who asked, as "stable". The word appeared in Rae's own notes four times.

---

What was actually happening

⚠ In month fifteen Rae built a table nobody had suggested: members grouped by the month they joined, tracked across every month since.

Joined	⚠ Started	⚠ M3	⚠ M6	⚠ M12	⚠ M14
⚠ Month 1	⚠ 74	61	38	14	⚠ 9
Month 4	22	15	9	—	⚠ 6
Month 7	31	19	12	—	⚠ 11
Month 10	28	17	—	—	⚠ 14
Month 13	26	—	—	—	⚠ 21

⚠ Rae had not been running a community of about seventy people. Rae had been running a community that lost roughly nine per cent of its members every month and replaced them, continuously, for fourteen months.

⚠ The headline number never moved because the recruitment exactly matched the leaving — and the recruitment was costing Rae most of a working week every month, unpaid, invisible, and recorded nowhere.

---

The nine who stayed

⚠ Rae then looked at what the nine survivors from the founding cohort had in common. There was exactly one thing.

⚠ All nine had posted something in the community within their first fourteen days.

⚠ Of the sixty-five who had left, fifty-one had never posted at all. Not once, in the entire time they paid.

⚠ They had joined, looked, paid for between two and nine months, and left — and Rae had never spoken to a single one of them individually, because there had never seemed to be a reason to.

---

What nobody said

⚠ Not one of the sixty-five told Rae they were leaving.

⚠ Not one gave a reason. Two replied to the automated cancellation email with a version of "no problem, just a bit busy" — which Rae now regards as the least informative sentence in this trade.

⚠ The cancellations arrived as notifications from a payment processor, individually, spread across fourteen months, each one small enough to absorb.

⚠ And every one of them was a decision made weeks earlier, silently, at a moment Rae could have seen and did not know to look for.

---

The sentence this book is built on

⚠ Rae had not lost members. Rae had replaced them, continuously, for fourteen months, at a cost that never appeared anywhere — and had called the number that concealed it stability.

---

What this book argues

⚠ You did not launch a product. You took on a promise that renews every month — and each month you are not selling a membership, you are being re-bought by everybody who has one.

Which produces four rules:

⚠ Rule One — you are selling a room made of other people. ⚠ The product is substantially the other members, which means it is partly outside your control, it is at its worst on the first day, and one wrong member damages it for everybody. Chapter 5.

⚠ Rule Two — the promise renews every month, and so does the work. ⚠ There is no finished state and very little of it can be front-loaded. Recurring revenue is recurring obligation, and the month you least want to deliver is the month it matters most. Chapter 6.

⚠ Rule Three — nobody tells you they are leaving. ⚠ The decision is made weeks before the cancellation, the revenue graph conceals it, and the only view that shows the truth is a cohort table. Chapter 7.

⚠ Rule Four — the first fourteen days decide the next fourteen months. ⚠ Nine of seventy-four, and the only thing they had in common was posting once in the first fortnight. Chapter 8.

---

What this book is not

⚠ It is not about launching. ⚠ ⚠ Rae's launch was a success. Chapter 18 covers it in one chapter because it is one chapter of the work.

⚠ It is not about growth. ⚠ A community that retains does not need much growth. One that does not retain cannot be grown out of.

⚠ It states no consumer-rights rule, refund requirement, payment regulation or tax obligation. ⚠ ⚠ Chapter 15 tells you what to establish and who to ask — and recurring payments are precisely the category those rules exist to govern.

⚠ And it makes no promise about member numbers or retention.

---

Where to start

⚠ Chapters 5 to 8, in order. ⚠ The four rules.

⚠ Then Chapter 19 — onboarding — before you take a single payment. ⚠ ⚠ It is the highest-leverage chapter in the book, it costs a weekend to build, and it is the difference between nine of seventy-four and something considerably better.

⚠ Then Chapter 11, and write the promise down before you charge anybody for it.

⚠ And build the cohort table in month one, even with one cohort in it. Resource 7 — it cannot be reconstructed later, and it is the only view that tells you the truth.

⚠ Rae's note, written in month sixteen and kept at the top of the community's operations document: "Seventy-four, seventy-one, and nine. The number I reported for over a year was true and it described nothing. I was running a treadmill and calling it a business, and the whole of it was visible in a table I could have built in twenty minutes at any point."

---

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

Chapter 1 — Why a Paid Membership Community Works as a Side Hustle

The revenue renews monthly and so does the obligation. Only one of those is planned for.

⚠ Recurring revenue is the most attractive sentence in small business and the least examined. This chapter is about what it actually obliges — because the money arrives monthly and so does the work, and only one of those is usually planned for.

1.1 The proposition: recurring revenue, and what it costs

What is genuinely true:

⚠ The revenue is predictable. ⚠ Which is rare, and is the real prize. ⚠ Resource 2.

⚠ It compounds slowly and reliably ⚠ if retention holds — and everything in this book turns on that condition. Chapter 30.

⚠ A small number of members is a real income. ⚠ ⚠ 120 members at $18 is $2,160 a month, which is a different proposition from most side hustles.

⚠ It can be run in evenings.

⚠ And you are paid before you deliver, ⚠ which is unusual and good for cash flow.

What it costs, and this is the part that is underestimated:

⚠ The obligation renews with the payment. ⚠ Rule Two — you cannot have a quiet month, and there is no version where the work stops while the revenue continues.

⚠ The product is other people, ⚠ which you do not fully control. Rule One.

⚠ And the losses are silent. ⚠ Rule Three.

1.2 What a member is actually paying for

⚠ Rarely the content, which is the thing most founders build.

In rough order of what actually retains people:

⚠ Other members. ⚠ Section 5.3 — being among people with the same problem is most of the value, and it is the part you cannot make yourself.

⚠ Being known. ⚠ ⚠ Somebody using their name. Somebody replying. Section 8.5.

⚠ Access to you, ⚠ in a bounded form. Section 4.5.

⚠ A rhythm. ⚠ ⚠ Something happening on a schedule, which structures their week. Section 23.7.

⚠ Answers to specific questions.

⚠ And, last, the content library. ⚠ ⚠ Which everybody builds first and which almost nobody stays for. Section 20.5.

1.3 Why it is easy to start and hard to keep

Easy to start:

⚠ A platform, a price, a promise, and people who already know you. It can genuinely be live in a fortnight.

Hard to keep, in the order it gets hard:

⚠ The room is empty and you are charging for it. ⚠ Section 5.5 — month one.

⚠ Launch energy runs out around month three. ⚠ Section 6.4.

⚠ The quiet sets in. ⚠ ⚠ Members stop posting, you post into silence, and it compounds. Section 23.2.

⚠ The first cancellations arrive with no explanation. ⚠ Section 7.1.

⚠ And month eight, ⚠ ⚠ which is where most paid communities end. Section 2.8.

1.4 The four rules, and where they come from

⚠ All four came out of one cohort table built fourteen months too late. ⚠ Resource 3.

⚠ Rule One — you are selling a room made of other people. ⚠ Fifty-one of sixty-five leavers never posted once. They had paid for a room they never entered. Chapter 5.

⚠ Rule Two — the promise renews every month, and so does the work. ⚠ Chapter 6.

⚠ Rule Three — nobody tells you they are leaving. ⚠ Not one of sixty-five. Chapter 7.

⚠ Rule Four — the first fourteen days decide the next fourteen months. ⚠ Nine survivors, and all nine had posted in their first fortnight. Chapter 8.

1.5 What this work rewards that other content work does not

⚠ Consistency over brilliance. ⚠ A predictable adequate thing beats an unpredictable excellent one.

⚠ Warmth. ⚠ ⚠ Which is genuinely a commercial skill here — using names, replying, noticing who has gone quiet. Section 8.5.

⚠ Narrowness. ⚠ Section 10.2 — members need to recognise each other.

⚠ Restraint. ⚠ ⚠ The smallest promise you can keep beats the largest one you can imagine. Section 11.5.

⚠ And patience. ⚠ ⚠ A community is worth most in year three, which is longer than most people plan for.

1.6 What it punishes that others forgive

⚠ Over-promising. ⚠ Section 11.4 — a promise made in launch enthusiasm is charged for monthly forever.

⚠ Disappearing. ⚠ ⚠ A week's silence from the host in a paid room is noticed by everybody and is the fastest route to cancellations.

⚠ Ignoring the quiet member. ⚠ Section 8.9.

⚠ Letting one member set the tone. ⚠ Section 5.4.

⚠ And growth without retention — ⚠ ⚠ which produces the treadmill in the introduction and can run for years while looking like a business. Section 30.5.

1.7 The central inversion of this book

⚠ You did not launch a product. You took on a promise that renews every month.

⚠ Each month you are not selling a membership. You are being re-bought — by every member, silently, without a conversation.

Which changes the questions:

⚠ Not "how do I get more members" but ⚠ "how long does a member stay".

⚠ **Not "what should I make this month" but ⚠ **"what did I promise, and did I deliver it". Section 11.1.

⚠ **Not "is the community growing" but ⚠ **"what does the cohort table say". Section 7.3.

⚠ And not "how is it going" but "how many members joined this month, and how many of the ones who joined a year ago are still here".

1.8 Who does well at this, and who struggles

Does well:

⚠ Somebody who genuinely likes the people. This is not sentimental — it is a job requirement.

⚠ Somebody who can deliver something adequate on a bad week.

⚠ Somebody comfortable being a host rather than a performer. ⚠ Section 5.6.

⚠ And somebody who will do the unglamorous half — ⚠ replying, noticing, welcoming.

Struggles:

⚠ Somebody who wants to make the content and not run the room. ⚠ ⚠ Which is most people who start one, and it is the commonest reason they fail.

⚠ Somebody who needs the community to be lively before they enjoy it. ⚠ Section 5.5. ⚠ Resource 1.

⚠ Somebody who cannot take a cancellation personally without it affecting the next month's work.

⚠ And anybody who wants to build it once. ⚠ Rule Two.

1.9 What the first year actually looks like

Period	⚠ What happens
Months 1–2	⚠ Launch energy. The room is quiet and you are working hardest.
⚠ Month 3	⚠ First cancellations, unexplained. Normal.
Months 4–6	⚠ Rhythm forms; participation is the whole job
⚠ Months 7–9	⚠ The decision point. Most communities end here.
⚠ Months 10–12	⚠ If retention holds, it starts compounding

⚠ Realistic first-year revenue: modest, and highly sensitive to churn rather than to signups. Two communities with identical acquisition and a four-point churn difference are a completely different business by month eighteen. Section 30.6.

⚠ What should be true at twelve months: a written promise you have kept every month; an onboarding sequence that runs whether or not you are available; a cohort table with twelve rows in it; a churn figure you can state; and a member list exported and held off the platform. Chapter 36.

1.10 Chapter summary

⚠ The revenue renews monthly and so does the obligation. Only one of those is usually planned for.

⚠ Members stay for other members, for being known, and for the rhythm. Almost nobody stays for the content library.

⚠ Most communities end at month eight, and it is a retention failure rather than a content one.

⚠ Each month you are being re-bought, silently, by everybody who has a membership.

⚠ Churn matters more than acquisition. Four points of it is a different business by month eighteen.

---

A realistic scenario

The library Rae spent four months building.

⚠ Months two to five. Rae had decided the community needed depth, and set about building a proper resource library: eleven long guides, a template pack, and a searchable archive of everything Rae had ever written on the subject.

⚠ Roughly ninety hours of work across four months, on top of the weekly delivery.

Rae's reasoning at the time, which was not stupid:

⚠ New members would see immediate value.

⚠ It would justify the price.

⚠ And it was work that could be done once, ⚠ which appealed enormously.

What the platform's own statistics showed at month six:

⚠ Of 69 members, 23 had opened any library item at all.

⚠ Eleven had opened more than one.

⚠ The most-opened guide had been read by nine people.

⚠ And the template pack, which Rae had spent about eighteen hours on, had been downloaded four times.

What members were actually using:

⚠ The weekly question thread. Opened by roughly 80% of members every week.

⚠ The monthly call, attended by 12 to 20 — ⚠ and, more importantly, the recording was watched by another 15 to 25.

⚠ And each other's posts.

What Rae did with the finding:

⚠ Stopped adding to the library entirely. ⚠ ⚠ It stayed, it was fine, and it was never mentioned in the promise again. Section 20.5.

⚠ Moved the ninety hours into the weekly thread and into replying — ⚠ specifically, into replying to every single post within a few hours. Section 23.5.

⚠ And started the practice that changed the business: personally welcoming every new member by name, in public, with a question they could answer in one line. Section 8.4.

What happened over the following two quarters:

⚠ Participation in the first fourteen days rose from about 20% of new members to about 55%.

⚠ Monthly churn fell from roughly 9% to about 5.5%.

⚠ Which, at 70 members, is the difference between replacing 6.3 people a month and replacing 3.9 — ⚠ and the recruitment to replace them was the largest uncosted block in Rae's month. Section 30.5.

⚠ Rae's note: "Ninety hours on a library that nine people read. The thing that actually retained anybody was replying to posts and using people's names, which costs about forty minutes a day and which I'd been treating as the bit I did when I had time left over. I had it exactly backwards, and the platform's own statistics had been saying so for four months."

---

AI prompts for this chapter

Prompt 1 — Is a Membership Community Actually Right for Me?

"Interrogate whether a paid membership community suits me, sceptically rather than encouragingly. ⚠ Ask me: do I actually like the people I would be serving, which is a job requirement rather than a sentiment; can I deliver something adequate on a bad week, every week, indefinitely; am I comfortable being a HOST rather than a performer; will I do the unglamorous half — replying, welcoming, noticing who has gone quiet; and do I want to make content or run a room? ⚠ Be direct if my answers suggest I want to make content, since that is the commonest reason these fail. ⚠ Then explain what recurring revenue actually obliges: the work renews with the payment, and there is no version where one continues without the other."

Prompt 2 — What Would Members Actually Stay For?

"Help me work out what my members would actually pay to keep, as opposed to what I would enjoy building. ⚠ The order I believe is real: other members; being known, meaning somebody uses their name and replies; bounded access to me; a rhythm that structures their week; answers to specific questions; and — last — a content library, which almost everybody builds first and almost nobody stays for. ⚠ Challenge that order against my specific audience after asking about them. ⚠ Then tell me honestly what proportion of my planned effort is going into the last item, and what it would look like to move that effort into the first three."

Prompt 3 — Am I Building the Wrong Thing?

"Help me audit where my delivery hours actually go against what members actually use. ⚠ Ask me: what I spend the most hours on; what proportion of members open, attend or use each thing; and what my platform's own statistics say rather than what I assume. ⚠ Be sceptical about anything that took many hours and is used by a small fraction — I once spent ninety hours on a library that nine people read while the weekly question thread was opened by eighty per cent of members. ⚠ Then help me plan the reallocation: what to stop making, what to stop promising, and what to move those hours into instead."

Prompt 4 — What Should Be True at Twelve Months

"Help me define what a membership community should have produced after a year, in terms that are not member count. ⚠ My candidates: a written promise I have kept every month; an onboarding sequence that runs whether or not I am available; a cohort table with twelve rows in it; a monthly churn figure I can state; and a member list exported and held off the platform. ⚠ Challenge or improve that list. ⚠ Then set realistic month-by-month expectations for year one, including that launch energy runs out around month three, the first unexplained cancellations arrive then, and most communities end at month eight — and tell me what specifically decides which side of month eight I land on."

---

⚠ AI checkpoint for this chapter

One — did it plan the content and not the room? Members stay for other members and for being known.

Two — did it treat recurring revenue as passive? The obligation renews with the payment.

Three — did it prioritise growth over retention? Four points of churn is a different business by month eighteen.

Four — did it assume a library retains people? Almost nobody stays for it.

---

Do This Now

1. Write down what you think members will stay for. Then read Section 1.2 again.

2. Look at your platform statistics: what do members actually open?

3. Count the hours you spend on content against the hours you spend replying.

4. Build the cohort table this week, even with one cohort in it.

5. Read Chapters 5 to 8 before you take a payment.

---

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