YURIOS // LAB
The Codex → Part VII · Business Chapter 42

Operating as a One-Person Studio

Solo, part-time, without burning out

The operational chapter. How to run this thing solo, part-time, without burning out — and without it quietly turning into a second job that pays nothing.

Everything before this chapter was what to build. This is how to keep building it for years, on 5–10 hours a week, as one person. The constraint is the whole design problem. Most of the advice you'll read about solo business assumes you quit your job and went all-in. You didn't. That changes the math, and most of this chapter is about what changes.

The one rule

Treat the project like a business and yourself like an employee.

Two halves. Both required, and each fails in a specific way without the other.

  • Business without employee is the workaholic trap: you optimise the venture and grind the human running it into dust. You ship for six months, hit a flat week, and quit — not because the project failed but because you did, and there was no slack in the system to absorb it.
  • Employee without business is the hobby trap: you make things you enjoy, never ship, never charge, never measure, and a year later you have a folder of half-finished cards and no audience, no revenue, no evidence anything is working.

The rule forces the tension that keeps both alive. As a business, you owe the project a strategy, a cadence, and honest accounting. As that business's only employee, the project owes you humane hours, a day off, and a paycheck eventually. You are on both sides of the table. Negotiate fairly.

Have others written about this? The prior art

Yes — extensively. "One person, deliberately small, profitable, sane" is a well-developed genre now, and you should stand on it rather than rediscover it. The useful sources cluster around a handful of people who actually did it:

  • Pieter Levels (@levelsio) — the patron saint of indie hacking. Built Nomad List, Remote OK, PhotoAI and a dozen others to a multi-million-dollar solo income with no funding, no employees, no office, and a deliberately boring stack (vanilla PHP, jQuery, one server). His doctrine: ship in 24–72 hours, embrace the ugly MVP, improve live, automate relentlessly, keep margins absurd because you have no investors and no staff. His "12 startups in 12 months" challenge is the canonical "execution beats analysis" stunt. (levels.io)
  • Paul Jarvis, Company of One — the philosophical anchor. The thesis: question growth as the default. Define "enough," then optimise for profit, resilience, autonomy, and a life you'd actually want, not for headcount or revenue records. Stay small on purpose; treat scaling as a cost, not a trophy. (summary)
  • Sahil Lavingia, The Minimalist Entrepreneur — the cautionary arc. Raised ~$8M for Gumroad chasing unicorn scale, missed the Series B, laid off ~75% of the team, then discovered the "failed" company was profitable, sustainable, and genuinely helping creators. His reframe — community before product, charge before you've built, own the business so it doesn't own you — comes from someone who tried it the expensive way first. (Mixergy summary)
  • Jason Fried & DHH (37signals), It Doesn't Have to Be Crazy at Work — the anti-hustle correction. "Stop celebrating Crazy, start celebrating Calm." Forty hours is plenty; protect attention; the company itself is your best product. Their "calm company" framing is the direct antidote to indie-hacker grind culture. (37signals)
  • Justin Welsh — the systems-and-content operator. Walked away from a nine-figure-scaling executive career after a personal crisis and built a one-person business to several million in revenue at ~92% margin on two products, a newsletter, and ruthless content systems — no ads, no staff, a couple of tools. The lesson for us: audience + digital products + repeatable content systems is the highest-leverage solo shape. (Growth in Reverse)
  • Brian Moran, The 12 Week Year — the execution cadence. Annual plans fail because twelve months is long enough to procrastinate. Compress the "year" to twelve weeks: short enough to feel urgent, long enough to finish something real. We adopt this directly below.
  • Cal Newport, Deep Work / Slow Productivity — the focus discipline. Fixed-schedule productivity (decide your hours in advance, force the work to fit), time-blocking, and the shutdown ritual ("schedule, shutdown, complete") that lets a part-timer actually stop and recover. Essential when your build time is scarce and bordered by a day job. (Cal Newport)

Where they agree (the consensus you can just adopt): profitability over growth; smallest viable version first; charge early; automate the boring; keep the stack and the tool list lean; protect your attention and your hours; measure leading activities, not just lagging revenue.

Where companion-building diverges from all of them. None of these authors built an attachment-based product as an anonymous solo creator. That introduces two problems they don't cover and this chapter has to:

  1. You are part of the product. The creator persona (ch. 36) is itself an artefact people form attachments to. That is leverage and it is a liability — see Health below.
  2. You're doing this part-time on 5–10 hours/week, not full-time. Levels, Welsh, and Lavingia all went all-in. Their absolute throughput numbers don't transfer; their ratios and disciplines do. Read them for principles, not for pace.

The week structure

A sustainable solo cadence at 5–10 hours/week. Decide the hours in advance (Newport's fixed-schedule productivity) and defend them like appointments, because they are.

Block Hours Output
Make 4–6 Build artefacts. The thing itself. Code, cards, lore, art.
Ship 1 Public release: post, drop, push. Something leaves the building.
Talk 1–2 Reply, community, DMs (on your terms — see below).
Plan 30 min Friday review + next-week task list.

Two rules govern the table:

  • The make/ship ratio is the whole game. The temptation is to make more and ship less — polishing in private feels productive and risks nothing. Resist it. Shipping is what compounds the audience; making in private compounds nothing. If a week is tight, cut make, never ship. A small thing shipped beats a big thing hidden, every single week.
  • Batch the Make block, don't sprinkle it. Ten minutes here and there produces nothing on creative work. One protected 2–3 hour Strategic Block (Moran's term) beats six fragmented half-hours. Most people get their real building done in one weekend session plus one weeknight. Find your two slots and make them sacred.

Close every working session with a shutdown ritual: log what you did, write the single next action, close the laptop. The part-timer's enemy is the project leaking into the other 160 hours of the week as low-grade background anxiety. The ritual is what gives you permission to stop.

The 12-week year for a part-time studio

Annual planning is a trap at this scale — twelve months is so long you can drift for two of them and not notice. Run 12-week cycles instead (Moran). Each cycle:

  1. Pick 1–2 outcomes, not five. At 5–10 hrs/week a cycle is ~80–120 working hours total. That's realistically one reference implementation, or two character cards plus an audience push — not both. Choosing is the work.
  2. Write the lead measures. Lag measures are results you don't fully control (followers, revenue). Lead measures are the activities that produce them and that you do control (cards shipped, posts published, build commits, outreach DMs sent). Track the lead measures weekly; they're the steering wheel.
  3. Score the week, aim for 85%. Moran's finding: hit ~85% of your planned weekly activities and the outcomes follow. Don't chase 100% — that's a sign you under-planned. Below ~60% for two weeks running, something's wrong; diagnose it (see maintenance mode).
  4. Stop at week 12, review, then go again. Twelve weeks is "your year." Measure, keep what worked, drop what didn't, set the next one to two outcomes. Two cycles ≈ the six-month gameplan in ch. 40.

The book's month-by-month plan (ch. 40) is the strategy; the 12-week cycle is how you actually run it without an annual plan you'll abandon by February.

Tools the project actually needs

Keep this lean. Solo overhead is a tax you pay forever, and every tool is a small recurring decision. The consensus across every source above: a boring, minimal stack you fully understand beats a clever one you have to maintain.

  • Repo: GitHub. Code, lore, and notes all in markdown where possible — version control is free memory.
  • Docs/notes: the repo itself + Obsidian (or similar) for ephemeral thinking.
  • Newsletter: Buttondown (~$9/mo, indie-friendly), Ghost (self-host, owns everything), or Substack (free, but rents you the audience). Owning the list is owning the audience — prefer that.
  • Payments: Stripe (consulting, courses, SaaS), Ko-fi (donations/one-offs), Patreon (recurring patronage).
  • Hosting: Vercel/Cloudflare/Fly for web; HuggingFace Spaces for demos.
  • LLM: start on hosted APIs (Anthropic/OpenAI/Together). Self-host locally only once cost or sovereignty demands it.
  • Domain: the creator handle + one project domain. Not a portfolio of speculative domains.
  • Discord: only when audience > ~500. Before that it's an empty room you feel obligated to staff.

Total operating cost month 1: < $30. Month 12 with audience: < $200. If your tool spend outruns your revenue, you've bought yourself a job, not a business. (Welsh runs a multi-million business on a literal handful of tools; copy the restraint, not the revenue.)

The "no" list

A solo project dies of yeses. Every yes is a withdrawal from a 5–10 hour weekly account. Pre-decide the standing refusals so you're not re-litigating them under social pressure in a DM:

  • No co-founder for free equity without 3+ months actually working together first.
  • No commissions for personas/content outside the canon you're building. (Bespoke work doesn't compound; it's a treadmill.)
  • No platform-of-the-month chasing. Pick two channels from ch. 37 and hold them. TikTok-this-month/Threads-next is how part-timers burn their entire budget on setup and never post.
  • No agency / dev-shop work that isn't directly companion-adjacent.
  • No drama. Mute, don't reply, don't quote-tweet. It's the cheapest time you'll ever save.
  • No "while I'm at it" scope creep. Reference implementations stay minimal on purpose (ch. 30).
  • No new product before the current one has either shipped or been explicitly killed. One open build at a time.

Jarvis's framing: every "no" to growth-for-its-own-sake is a "yes" to keeping the business something one person can actually run.

The compounding model: a stack of small wins

The whole strategy is small things shipped on a steady cadence, compounding. Lavingia calls it small wins propelling the next; the shape is the same every month — ship a small thing, post about it, talk to the people who showed up.

  • Month 1: 1 character card, 50 followers.
  • Month 3: 2 cards, 1 reference impl, 250 followers.
  • Month 6: 3 cards, 2 reference impls, 1 book pre-sale, 1,000 followers.
  • Month 12: 5 cards, 3 reference impls, 1 book, 1 desktop app, 3,000 followers.
  • Month 24: 10+ cards, 4 reference impls, the SaaS product, 10,000 followers.

These are illustrative floors, not promises. The point isn't the numbers; it's the shape — a flat, repeatable monthly rhythm that survives bad weeks because no single drop carries the whole project.

Money: what staying small actually buys

The reason to stay small isn't modesty — it's leverage. The solo creator-economy shape (audience + digital products + content systems) runs at 85–99% margins because there's no payroll, no office, no investors taking a cut or dictating growth (Levels, Welsh). For a part-timer that margin is the entire point: a project that nets $1,500/month at 90% margin is a raise, while one that grosses $5,000 and nets $400 after tools and contractors is a stressful hobby.

So the financial target isn't revenue, it's profit per hour worked. Optimise the thing that pays the most per scarce hour. Ranked roughly by leverage for this project:

  1. Digital products that sell while you sleep — character cards, lorebook packs, the book, a course. Make once, sell many. Highest leverage.
  2. Recurring patronage — Patreon/newsletter tiers. Predictable, compounds with audience.
  3. Productised services — a fixed-price persona audit or card-+-implementation bundle. Higher rate than hourly, capped time.
  4. Hourly consulting — last resort. Real money, zero compounding, trades your scarcest asset directly for cash. Use it to fund the others, not as the destination.

Pause monetisation entirely if it's corrupting the work (ch. 39) — but when you do charge, charge for the leveraged things first.

Health

This is not a soft section. It's the highest-risk part of running this specific project, and the prior-art authors mostly don't cover it because they didn't build attachment-based products.

Why companion-creators are unusually exposed. Your audience overlaps heavily with people prone to intense one-sided attachment — that's why the product works. But the same dynamic points back at you. Research on "comfort creators" finds that these one-sided bonds generate feelings of closeness that exceed any real relationship, which loads the creator with continuous, uncompensated emotional labour: audiences demand authenticity and availability, and managing thousands of one-sided relationships blurs the line between engagement and emotional drainage. Surveys put full-time-creator burnout well above half the population, with algorithmic pressure and always-on availability as the top stressors. (Grady/UGA, The Creator Economy) You are building the attachment machine and standing in front of it. Hard rules, set in advance, are the only defence:

  • DM hours: structured, not open. Probably no open DMs at all. If yes, fixed windows, batched, never a 24/7 obligation. The audience does not get on-demand access to a person.
  • Anonymous: yes. The persona is public; your civilian identity is not. This is a boundary and a brand decision (ch. 36), and it's far easier to keep from day one than to claw back later.
  • Expectations managed publicly. State your cadence — "I ship monthly" — and let it set the norm. Silent over-delivery trains the audience to expect nightly contact, then resent its absence.
  • One full day per week with zero project work. Non-negotiable. This is the employee half of the one rule enforcing itself.
  • Separate the persona from the self. Criticism of the brand is not criticism of you. Build the wall deliberately; anonymity helps you keep it standing.

Burnout warning signs + the maintenance-mode protocol

Burnout at 5–10 hrs/week rarely looks like collapse; it looks like quiet drift. Watch for:

  • Dread before the Make block where there used to be pull.
  • Shipping stops first. Making in private continues (it's safe); public drops quietly lapse. This is the canary — the make/ship ratio inverting is the earliest measurable sign.
  • Weekly score under ~60% two cycles running with no external cause.
  • Resentment of the audience — replies feel like a tax, not a connection. (Direct symptom of the emotional-labour load above.)
  • "While I'm at it" sprawl — chasing new scope because the core work has gone joyless.

When two or more show up, drop to maintenance mode on purpose — don't quit, downshift:

  1. Cut to the spine: newsletter + one channel only. Pause builds, Discord, video, consulting.
  2. Pre-write and schedule two to four weeks of drops in one sitting, so "shipping" keeps happening with near-zero ongoing load (ch. 40's lore-exhaustion fix).
  3. Announce nothing dramatic. "Heads-down on the next build" is true and sufficient. No farewell posts.
  4. Set a review date (end of the current 12-week cycle). Recover, then decide: resume, re-scope, or honourably wind down.

Maintenance mode is a feature of the system, not a failure of it. A project that can idle is a project that survives the bad quarter. The alternative — grinding through on willpower — is exactly the workaholic-trap failure the one rule exists to prevent.

Templates

Friday review (15–20 min, weekly)

Week of: __________   12-week cycle: __ / 12

DID I SHIP? (the only non-negotiable line)   yes / no — what: ______

LEAD MEASURES (what I controlled)
  Build commits / artefacts:        target __  actual __
  Public posts / drops:             target __  actual __
  Audience touches (replies/DMs):   target __  actual __
  Outreach (warm contacts):         target __  actual __
WEEKLY SCORE: ___% of planned activities done   (aim 85%)

LAG MEASURES (snapshot, don't over-react)
  Followers ___   Subscribers ___   MRR/revenue $___

ENERGY CHECK (the employee half)
  Dread / neutral / pull?  ____    Day off taken?  y / n
  Any burnout warning signs firing?  ____

NEXT WEEK
  The one Make block goes here: __________  (day + hours)
  The one thing I will ship:    __________
  One thing to STOP doing:      __________

12-week cycle plan (1 page, every 12 weeks)

Cycle __ — dates __________ to __________

VISION CHECK: does this cycle move the long arc? (the book's thesis)  y / n

THIS CYCLE'S 1–2 OUTCOMES (be brutal — ~80–120 hrs total)
  1. __________________________
  2. __________________________ (optional)

WEEKLY LEAD-MEASURE TARGETS
  Ship: __ /wk   Build: __ /wk   Post: __ /wk   Outreach: __ /wk

THE "NO" THIS CYCLE (what I'm explicitly declining)
  ____________________________________

RISK: most likely reason this slips → mitigation
  ____________________________________

END-OF-CYCLE REVIEW DATE: __________

The hand-off question: if this becomes a real studio, what do you hire (or automate) first?

Don't hire to grow; hire to remove the lowest-leverage, most-draining hour from your week — and prefer automation to headcount until automation genuinely can't do it (every source above hires reluctantly). Rough order:

  1. Automate before you delegate. Scheduling, cross-posting, payment/email flows, analytics. A few Zapier-style automations replace your first "hire" and never need managing.
  2. First human: a VA / community moderator. Offloads the emotional-labour surface — moderation, triage, routine replies — which is both low-leverage and the biggest burnout vector. Highest-value first hire for this project specifically.
  3. Then: contract art or contract dev, project-by-project, for the bottleneck skill you most lack. Contract, not employee — keep fixed costs near zero.
  4. A co-founder/employee is the last resort, and only against the Jarvis test: does this make the business better for the life you want, or just bigger? If only bigger, don't.

The goal is never "a team." It's a one-person studio that one person can run indefinitely, with help bought surgically where it removes pain.

Not legal or tax advice. This is a checklist of what to set up and who to ask — confirm specifics with a professional in your jurisdiction. See ch. 41 for the fuller legal/compliance picture.

The minimum viable back office for a solo creator:

  • Separate the money on day one. A dedicated business bank account and card, even before forming an entity. Never commingle. This alone makes everything below tractable.
  • Entity: most solo creators start as a sole proprietor and form an LLC when revenue or liability justifies it (in the US; equivalents elsewhere). The LLC's value is liability separation and credibility, not magic tax savings at small scale.
  • Bookkeeping from the first dollar. A simple spreadsheet or Wave/QuickBooks. Categorise income (products, patronage, consulting) and expenses (tools, hosting, contractors, art). Future-you and your accountant will thank you.
  • Set aside tax as you earn, don't reconcile in a panic at year-end. A common rule of thumb is parking ~25–30% of profit; the right number depends entirely on your jurisdiction and bracket — ask.
  • Know your obligations: self-employment/estimated taxes, and sales tax / VAT on digital goods, which can apply based on where your buyers are, not where you are. This catches digital-product sellers constantly — platforms like Gumroad/Patreon may handle some of it; confirm what they cover.
  • Keep the anonymous brand and the legal entity reconciled. Anonymity is a public-facing choice; your bank, tax authority, and any payment processor will still need your real identity. Plan for that gap (ch. 36, ch. 41) rather than being surprised by it.

Spend an hour with an accountant once at setup. It's the highest-ROI "tool" purchase in this chapter.

Where this differs from generic solopreneur advice

To close the loop on the prior art — adopt the consensus, but adjust three things for this project:

  1. Part-time pace, not full-time. Read Levels/Welsh/Lavingia for ratios and discipline; ignore their absolute throughput. Your unit of planning is the 12-week cycle at 80–120 hours, not their full-time year.
  2. Attachment exposure is a first-class risk. The Health section isn't optional polish — for a companion creator it's load-bearing. The generic literature underweights it because they didn't build attachment machines.
  3. Ethics can override the money. Every source optimises for sustainable profit. This project sometimes pauses profit on purpose when the money would pull against the ethics (ch. 5). That's a feature, and it's the one place where being a company of one with no investors is pure advantage: no one can make you grow in a direction you've decided is wrong.

Sources & further reading