Side HustlesBeginner22 min read

The weekly CEO routine: 30 minutes to run a one-person company

A fixed weekly slot, one page of numbers, five blocks (look, decide, protect, plan, tidy), using AI to prepare without handing it decisions, a decision log, handling bad and good weeks, and growing the routine only when needed.

The weekly CEO routine: 30 minutes to run a one-person company

The Weekly CEO Routine: 30 Minutes to Run a One-Person Company

Once you have customers, the work changes. Before, the hard part was starting. Now the hard part is staying in charge: the inbox fills, a customer's question sits for three days, you forget to chase an invoice, you spend a week on something that did not matter and you cannot say whether the business is growing or just busy. A one-person company has no manager, so the founder has to do the manager's job as well as the work. The answer is not a heavy system. It is one short, regular meeting with yourself: thirty minutes, once a week, with the same agenda every time. This tutorial gives you that routine. You will learn the five blocks (look, decide, protect, plan and tidy), one page of numbers worth tracking, how to use AI to prepare the review without handing it your decisions, how to keep a decision log and what to do when a week goes badly. It promises no results. It makes sure that whatever you are doing, you are doing it on purpose.

Difficulty: ★☆☆☆☆ (no technical background; a calendar and a notes app)
Tools needed: A notes app or spreadsheet, a calendar and any chat AI to help prepare
Last updated: November 2026

Overview

A CEO routine is a short, fixed, weekly review of your business, done by you, as the person in charge. It is not a report for anyone else. It is how you look up from the work and answer four questions:

1. Where are we? What happened this week, in plain numbers.
2. What matters most next? The one to three things that move the business forward.
3. What could go wrong? Customers at risk, money problems, deadlines, promises.
4. What will I do, and what will I stop? A short, realistic plan for next week.

Why weekly? Daily is too noisy and monthly is too slow. A week is long enough for patterns to show and short enough to fix a problem before it grows.

Why thirty minutes? Because a routine you can keep beats a perfect one you skip. If it takes two hours, you will do it twice and quit. Thirty minutes is enough to look, decide and plan, if you prepare the numbers in advance and do the same thing each time.

Why does it matter more for a solo founder? Because there is nobody else to notice. A team has meetings, a manager and colleagues who say "didn't we promise this customer a reply?". A solo founder has only their own attention, and attention goes to whatever is loudest, not whatever is most important. The routine moves your attention, once a week, onto what is important.

What it is not.

  • It is not a planning marathon or a goal-setting retreat.
  • It is not a dashboard project. A few numbers in a spreadsheet are enough.
  • It is not a place to criticise yourself. It is a place to decide.
  • It is not a job for AI. AI can help you prepare and spot things you missed. The decisions are yours.
  • How AI helps. It can summarise your notes, turn a messy week's log into a short list, draft replies you have been putting off, spot numbers that moved and play the sceptic about your plan. It cannot see your customers' real situation, hold your priorities for you or take responsibility. Do not let it send, pay, publish or delete without your approval. The final module quest and the next tutorial cover how to set approval rules for any AI helpers.

    This tutorial has seven steps. First, set up a fixed time and a place. Second, the one page of numbers. Third, the five blocks of the 30 minutes. Fourth, using AI to prepare. Fifth, the decision log. Sixth, what to do on a bad week. Seventh, how to grow the routine only when you need to.

    A note on honesty. The numbers and examples are invented. This tutorial gives no benchmarks and no promises of growth. Nothing here is financial, tax or legal advice. Where your records feed tax or legal obligations, check the rules where you live.

    Who This Is Useful For

  • People with their first customers who feel busy but not in control.
  • People with a day job who need a small, reliable habit to keep a side business moving.
  • Anyone who has tried big planning systems and abandoned them.
  • What You Will Learn

  • How to set up a weekly 30-minute review you will actually keep.
  • A one-page set of numbers that tells you if the business is healthy.
  • A five-block agenda: look, decide, protect, plan, tidy.
  • How to prepare the review with AI without outsourcing decisions.
  • How to keep a simple decision log.
  • How to handle a bad week, and a very good one.
  • How to adjust the routine as the business grows.
  • What You Need

    1. At least a few customers or a clear process, so there is something to review. If you have none yet, you can still use this routine for your 14-day customer plan.
    2. A notes app or spreadsheet, and a calendar.
    3. Thirty minutes a week, in a fixed slot, plus ten minutes of preparation.
    4. A chat AI, optional.

    Information as of 14 November 2026. Tools change, and this tutorial names none. Keep customers' private information out of any tool you have not checked.

    Step 1 — A fixed time and a place

    The routine only works if it happens. Make it easy to keep.

    Pick a slot and protect it. Choose a weekly time when you are reasonably fresh and not interrupted. Many people like Friday afternoon (to close the week) or Sunday evening or Monday morning (to open the next). Choose the one that fits your life. A slot after a good night's sleep beats a slot after a long workday. Put it in your calendar as a repeating event, and treat it as a meeting you cannot cancel for a minor reason.

    Keep it to 30 minutes. Use a timer. If you routinely run over, you are either doing work in the review, which belongs in your week, or you have too many numbers. Cut.

    Have one place. A single document or spreadsheet holds your weekly page. Each week, duplicate last week's page and fill in the new one. Over time, you build a record.

    Prepare for ten minutes before. Update the numbers from your payment records and your messages, so the thirty minutes is for thinking, not collecting. If it takes longer, simplify what you track.

    Start with a rest rule. Decide the one day a week the business does not touch you. A review is useless if you are exhausted.

    If you have a day job, the routine is even more important and must be even smaller. A 20-minute version is better than none. Keep the business separate from your employer's time, tools and information.

    If you miss a week, do not make up two reviews. Do a short one, note what slipped and carry on. Do not abandon the habit.

    A tiny setup.

    WEEKLY CEO ROUTINE SETUP
    Day and time: ___ (repeating, 30 minutes)
    Place (document or sheet): ___
    Preparation (10 min) happens: ___
    My rest day: ___
    If I miss a week I will: do a 15-minute version the next day.
    Person I will tell about my numbers each month (optional): ___


    Step 2 — One page of numbers

    Choose a few numbers. If you track twenty, you will track none. Your page should answer one question: is the business healthy this week?

    The core numbers for most one-person companies.

    AreaNumberWhy it matters
    CustomersNew customers this week, and total activeIs demand coming in?
    RevenueMoney received this week, and this month so farAre you being paid?
    CostsMoney spent this week, and this month so farAre costs under control?
    PipelinePeople in conversation, and offers outstandingWhat is coming?
    DeliveryWork promised this week, work done, anything lateAre you keeping promises?
    SupportQuestions or complaints open, and how oldIs anyone waiting?
    TimeHours spent on the business, and on whatAre you within your limits?
    EnergyA 1 to 5 rating of how you feelAre you sustainable?

    Pick the three that matter most for your stage. At the start (first ten customers), pipeline, new customers and delivery matter most. Later, revenue, costs and time may take centre stage. Add others only when you need them.

    Use real, countable numbers. Money received, not "sales expected". Conversations held, not "interest". Hours spent, not "a lot". If you cannot count it, write what you can observe, in a few words.

    Compare, do not just record. A number alone says little. Beside each, write last week's number and the change. A rising cost with flat revenue tells a story.

    Know your thresholds. For each key number, decide in advance what triggers action. For example: "If money received in a month is below my costs for two months in a row, I review pricing." "If any customer question is older than two working days, I clear it first." Thresholds turn numbers into decisions. Link them to your kill criteria from your limits sheet.

    Honest records. Record every sale and cost on the day, or the numbers will be wrong. These records also support tax and legal obligations, so keep them clean. Check the rules where you live.

    Avoid vanity numbers. Followers, page views and likes belong here only if you can link them to customers.

    Your one page.

    WEEK OF ___
    Customers: new ___ (last wk ___), active ___
    Money in: ___ (month to date ___)   Money out: ___ (month to date ___)
    Pipeline: conversations ___, offers outstanding ___
    Delivery: promised ___, done ___, late ___
    Support: open ___, oldest ___ days
    Hours on the business: ___ (limit ___)    Energy (1-5): ___
    Thresholds hit this week: ___

    Step 3 — The five blocks of the 30 minutes

    The same agenda every time keeps the meeting short. Here are five blocks with suggested times.

    Block 1: Look (5 minutes). Read the numbers. Do not interpret yet. Just notice.

  • What went up, what went down, what did not move?
  • Did any threshold trigger?
  • What surprised you?
  • Write two lines: "Good this week" and "Not good this week".

    Block 2: Decide (8 minutes). Choose what matters next.

  • The one to three priorities for next week. Choose from what moves customers, money or delivery. Priorities should fit your real hours from your limits sheet.
  • One thing to stop or reduce. Every week add a "stop" along with a "start". It might be a task, a tool, a meeting or a habit that costs more than it earns. Without stopping things, your list only grows.
  • One decision you have been avoiding. Pricing, a difficult customer, an offer to change. Make it, or set a date to make it.
  • A good priority is specific: "Send the day-5 follow-up to the six people who have not replied" rather than "do outreach".

    Block 3: Protect (7 minutes). Look for what could go wrong.

  • Customers at risk: anyone unhappy, late, waiting for a reply or due to renew.
  • Promises: what did you promise this week or next, and what could break it?
  • Money: invoices unpaid, costs due, a subscription renewing, tax set-asides.
  • Rules and records: anything with a deadline, such as a filing, a renewal or a licence.
  • Data and security: anything shared by mistake, a password to update, a backup to check.
  • Health: your sleep, your stress, your day job, your people. Does anything need protecting?
  • For each risk you find, write one action and a date.

    Block 4: Plan (7 minutes). Turn the priorities into the calendar.

  • Put the priorities in specific time slots in the coming week, in your best hours. A priority without a time slot rarely happens.
  • Put customer-facing deadlines in first, then the rest.
  • Leave slack. Plan about 70 per cent of your hours. Something will come up.
  • Plan the first action for each priority, so Monday starts without thinking.
  • Delegate or automate one small thing, if you have a helper or tool you trust (see the next tutorial and the quest), within approval rules.
  • Block 5: Tidy (3 minutes). Close the loop.

  • Clear tiny tasks that take under two minutes.
  • Archive or delete what is finished.
  • Write the decision log entry (Step 5).
  • Save this week's page, and prepare next week's blank.
  • A quick agenda to keep in front of you.

    CEO ROUTINE (30 min)
    1. LOOK (5): numbers, good / not good
    2. DECIDE (8): 1-3 priorities, 1 thing to stop, 1 decision to make
    3. PROTECT (7): customers at risk, promises, money, rules, data, health
    4. PLAN (7): priorities into calendar slots, first actions, slack, one thing to hand off
    5. TIDY (3): clear tiny tasks, log the decision, save the page

    If you only have 15 minutes, do Look, Decide and Protect in five minutes each, and put priorities in the calendar immediately.

    A word on tone. Be kind. A review where you attack yourself every week will be abandoned. Be honest about what happened, and practical about what to change.

    Step 4 — Using AI to prepare, not to decide

    AI is useful in the ten minutes before the review and in a few places within it. The rule: AI prepares, you decide.

    Where AI helps.

  • Summarising your notes. If you keep a messy daily log, ask AI to condense the week into a short list of events, customers, problems and promises.
  • Spotting numbers that moved. Paste your two weeks of numbers and ask what changed most.
  • Drafting replies. Prepare responses to customer messages that are waiting, which you then read, edit and send yourself.
  • Challenging your plan. Ask what you may be underestimating or missing.
  • Formatting. Turning your notes into the weekly page, a short customer update or a decision log entry.
  • Reminders and checklists. Preparing a checklist for recurring tasks.
  • Where AI does not belong.

  • Deciding priorities. It does not know your goals, limits or customers.
  • Judging customers' real situation. It sees only the words you paste.
  • Sending, paying, publishing or deleting without your explicit approval. If you use any assistant that can act on your accounts, set clear limits first, as the team roster quest covers.
  • Handling private information carelessly. Remove names and identifying details, and check each tool's terms before you paste customer information.
  • A preparation prompt.

    Role: a calm, honest operations assistant for a one-person company. You cannot see my customers or accounts. Do not invent facts, numbers or customer details. Anything you are unsure of, mark [CHECK].
    Task: I will paste my notes and numbers for the week (names removed). (1) Summarise what happened in five lines. (2) List which numbers moved and which crossed my thresholds. (3) List anything that looks like a customer at risk, an unkept promise, an unpaid invoice or a deadline, based only on what I wrote. (4) Suggest two questions I should ask myself in the review. Do not tell me what my priorities should be.
    [paste notes and numbers]

    A challenge prompt for the plan.

    Role: a sceptical advisor. You do not know my business beyond what I paste.
    Task: Here are my priorities and time slots for next week and my available hours. (1) Tell me whether the plan fits the hours. (2) Which priority is most likely to slip, and why? (3) What did I leave out that a one-person company usually forgets (follow-ups, admin, rest)? Keep it short.
    [paste the plan]

    Check the output. Compare what AI says with your own memory and records. If it is wrong, the notes may be unclear, or the tool made it up. You are responsible for what you act on.

    Keep AI's role small at first. One use per week is enough: summarising notes, or challenging the plan. Add more only when the routine is stable.

    Step 5 — The decision log

    A decision log is a simple record of what you decided, why and what you expected to happen. It is one of the most useful things a solo founder can keep, and it takes two minutes a week.

    Why keep it.

  • You forget why you decided things. Three months later, a reason you were sure about may be gone.
  • It improves your judgement. You can compare what you expected with what happened.
  • It stops you from re-arguing settled questions.
  • It protects against hindsight bias. You can see what you actually knew.
  • It is a record if you later work with a partner, adviser or helper.
  • What to record. Only meaningful decisions: pricing, offers, tools, customers you accept or decline, big spends, stopping something and changes to the plan.

    DECISION LOG
    Date | Decision | Why (the facts I had) | What I expect to happen | When I will check | Result (later)

    Examples.

  • "Raised price for new customers from HK$1,100 to HK$1,200. Reason: delivery takes 5 hours, not 4, and I have a waiting list. Expect: fewer yeses, same bookings. Check in 60 days."
  • "Stopped weekly social posts. Reason: three months, no customer traced to them, 2 hours a week. Expect: no change in enquiries. Check at 6 weeks."
  • "Declined a customer who wanted tax advice. Reason: outside my scope, I'm not qualified. Expect: none. Note: referred them to a licensed professional."
  • Review the log monthly. Look at the "result" column. Where were you right? Where wrong? What does that say about your judgement and your information? This is where real learning happens.

    Keep it private, and free of customers' personal information.

    Do not over-record. If you log every small choice, you will stop. Five to ten meaningful entries a month is plenty.

    Step 6 — A bad week, and a very good one

    Weeks vary. The routine matters most when things are not normal.

    A bad week. Something went wrong: a late delivery, a complaint, a lost customer, a sudden cost, an illness or a day job crunch.

  • Do the review anyway, in short form. Skipping it makes the next week worse.
  • Look at facts first. What exactly happened? Write it down in two lines, without blaming.
  • Protect the customer. Contact anyone affected, honestly and early. Apologise sincerely for real problems, follow your refund and revision terms and fix it.
  • Find the cause, not the culprit. Was it capacity, a missing step, unclear scope, a tool failure or a promise you should not have made? One process fix is more valuable than self-criticism.
  • Cut, do not push. If you are overloaded, drop a priority, tell someone the date will move, or pause new outreach. Do not borrow from sleep.
  • Check your kill criteria and limits. Is anything triggered?
  • Look after yourself. Rest, food, a conversation with a person you trust.
  • Write one lesson in the decision log.
  • A very good week. A new customer, a big sale, nice feedback.

  • Do the review anyway. Success can hide problems.
  • Ask why it worked. Was it a message, a channel, a referral, an offer? Write it down so you can repeat it.
  • Check capacity. Can you deliver what you just sold? Open a waiting list if needed.
  • Do not change everything. One good week is one data point.
  • Say thank you to those who helped.
  • Set aside tax and costs from new income, as per your records and local rules.
  • A very quiet week. Nothing happened.

  • Look for what you stopped doing. Outreach, follow-ups, content?
  • Check the pipeline. If it is empty, the next priority is filling it.
  • Do not panic-discount. Review pricing calmly, with evidence.
  • Talk to someone. A past customer, an interviewee or a peer can give you a fresh view.
  • Burnout signs. Dread of opening your inbox, skipping meals, constant tiredness, resentment of customers, dropping things you enjoy. Reduce scope, say no, rest and talk to someone. If low mood or anxiety persists, speak to a health professional. A business is not worth your health.

    Never review when you are in a rage or in despair. Postpone for a day. Decisions made at your worst are rarely good.

    Step 7 — Grow the routine only when you need to

    The routine should be as small as it can be and still work. Add parts when a real problem appears, not before.

    Signs you may need more.

  • You have more than a handful of customers, and tracking them takes more than a line each. Add a simple customer list with status and next step.
  • You have a recurring task that eats time. Document it as a checklist. Consider whether a tool or helper could do part of it, with approval rules.
  • You have money coming and going from several places. Add a short monthly money check: income, costs, tax set-aside and a look at the runway.
  • You are planning a bigger move: a price change, a new offer, a helper or a hire. Add a monthly or quarterly review with a longer horizon.
  • You work with someone, such as a part-time helper or an adviser. Add a short shared note, so they know priorities.
  • A monthly add-on (about 30 minutes). Once a month, in addition to the weekly review:

    MONTHLY REVIEW
    Revenue, costs, effective hourly value, runway: ___
    Customers: gained, lost, why: ___
    Best channel this month: ___
    Decision log: what was I right and wrong about? ___
    Limits sheet: any limit or kill criterion changed or triggered? ___
    Offer and price: change anything? ___
    One thing to learn or fix next month: ___

    A quarterly add-on (about an hour). Look at the 90-day picture: is this the right lane, offer and price? Do I want more, the same or less? What would make the next quarter a good one? Are my limits still true?

    Resist complexity. Every added meeting, number or tool is a cost. If a part of the routine has not changed a decision in two months, remove it.

    Tools. A simple document and a spreadsheet are enough for a long time. Fancy tools can wait.

    When you add helpers (human or AI). Keep control: each helper has one job, a clear limit and a way for you to check its work. The next tutorial, on staying solo or hiring, goes deeper, and the quest has you design a small AI team roster with approval rules.

    Worked Example: Wen's first month of reviews

    Wen has four paying clients for her monthly bookkeeping tidy-up, a day job and about eight free hours a week.

    Setup. Friday at 8:30 pm, 30 minutes, in a spreadsheet tab per week, with ten minutes of preparation on Thursday. Rest day: Sunday.

    Her page. Customers: new 1, active 4. Money in: HK$4,500 this month. Money out: HK$300. Pipeline: 3 conversations, 1 offer outstanding. Delivery: 4 promised, 3 done, 1 late by two days. Support: 2 open, oldest 3 days. Hours: 9 (limit 8). Energy: 3.

    Week 1 review.

  • Look. Good: a new client. Not good: one sheet was late, hours over limit, energy low.
  • Decide. Priorities: (1) finish the late sheet before anything else, (2) answer both open questions tomorrow, (3) one follow-up to the offer outstanding. Stop: checking email at midnight. Decision: whether to cap clients at five until she is faster. She decides yes.
  • Protect. Risk: the late client, who is unhappy. She writes to him tonight, apologises plainly and offers the refund her terms promise if she misses the next date. Another risk: she has not set aside tax. She sets aside 15 per cent of this month's income, as agreed with her accountant friend.
  • Plan. Priorities go into Saturday morning and Tuesday evening slots. She plans 6 hours, not 8. She hands one thing to a tool: a template for client reminders, which she reviews each time.
  • Tidy. She logs: "Cap clients at five for two months. Reason: delivery takes 5 hours and I am over my limit. Expect: slightly slower growth, steadier quality. Check in 8 weeks."
  • With AI. Wen pastes her anonymised notes into a chat AI, and asks for a five-line summary and the numbers that crossed her thresholds. It correctly notes the late sheet and the hours over limit, and incorrectly says she has two late sheets, because her notes were unclear. She fixes her notes. She does not ask the AI for priorities.

    Week 3. A very good week: two enquiries and a referral. Wen notices she is at capacity, so she opens a waiting list instead of accepting everyone. She writes down that the referral came from a client she delivered early for.

    Month review. Revenue is steady, costs are low, her effective hourly value is just below her minimum. She raises the price for new clients, as planned in her pricing review, and records the decision. Her energy score has risen from 3 to 4 since she stopped midnight email.

    The routine took thirty minutes a week, and it changed three real decisions in four weeks.

    Common Mistakes to Avoid

    Three habits kill the routine.

    Mistake #1: Making it too big. Twenty metrics and a two-hour meeting are abandoned in week three. Fix: three to six numbers, five blocks, a timer and a slot you protect.

    Mistake #2: Reviewing without deciding. Reading numbers and feeling informed is not management. Fix: every review ends with one to three specific priorities with time slots, one thing to stop and a logged decision.

    Mistake #3: Letting AI run the meeting. AI can summarise and challenge, but it does not know your customers, limits or goals, and it can get facts wrong. Fix: use it to prepare, check what it says against your records and keep decisions, sending, paying and deleting in your own hands.

    Going Further

    Share your numbers with one person. A friend or peer who asks you about them monthly adds gentle accountability.

    Track one number for a quarter. Pick the one you most want to move, such as follow-ups sent or days to deliver, and watch it.

    Write a one-page "how I run this business". Over time, your routine, checklists and decision log become a manual. It is valuable if you ever want to hand parts to a helper.

    What to learn next. The final tutorial in this track asks a big question: should you stay solo, or bring in help, and when? You will also write an AI team roster in the quest, with one job per helper, a limit and a way to check, and approval for anything that spends money or speaks for you.

    Key Takeaways

    In this tutorial you learned:

  • A weekly 30-minute CEO routine keeps a one-person company under your control. It is a short, fixed meeting with yourself with the same agenda each time.
  • Protect the slot and prepare for ten minutes, so the thirty minutes is for thinking. A 15-minute version beats skipping.
  • Track a few real numbers: customers, money in and out, pipeline, delivery, support, hours and energy. Compare to last week and set thresholds.
  • Use five blocks: look, decide, protect, plan and tidy. Always choose one to three priorities, one thing to stop and one decision.
  • AI prepares, you decide. Use it to summarise, spot changes and challenge your plan, and never let it send, pay, publish or delete without your approval.
  • Keep a decision log with the reason, expectation and result, and review it monthly.
  • Do the review on bad, good and quiet weeks. Protect customers and yourself first, find causes and not culprits.
  • Grow the routine only when a real need appears, with a monthly and quarterly add-on, and remove parts that no longer change decisions.
  • Before you move on, a challenge. Put a 30-minute weekly slot in your calendar, build your one page of numbers with three to six lines and run your first review this week. End it with one to three priorities, one thing to stop and one logged decision.

    Learn AI, after work

    Track your progress, earn XP, and unlock more free tutorials in the AfterWork Academy app.

    Open this tutorial in the app

    More AI tutorials