Skip to content
WhittleOSWhittleOS
← All guidesStarting a business6 min read

How to run a business alone without the week breaking

We run WhittleOS as one person and publish our risk register: 10 risks, 1 open and 7 being tested. The rhythm, the limits and the preparations that keep a solo business running.

By Boris Binyaminov ·

Our own register
10 risks, 7 being tested
Rhythm
Daily, weekly, monthly, quarterly
Plan for
The week you cannot work

The week has to keep fitting

Run it on a rhythm, and write down what could end it

01

Daily

The support inbox

02

Weekly

One thing that brings new buyers

03

Monthly

The books and the risk register

04

Quarterly

Access, domains and backups

Daily support, weekly selling, monthly books and a risk register read on a fixed date keep a one-person business from breaking on the duty nobody noticed.

To run a business alone, keep the jobs that never stop (support, selling and the books) inside a fixed weekly rhythm, and write down what could end the business before it does. We run WhittleOS as one person and publish that list. It holds 10 risks today: 1 open, 7 being tested and 2 closed. The rest of this page is the rhythm, the limits that end solo businesses, and what to prepare for the week you cannot work.

The jobs that do not stop when you are alone

In a team, someone else notices when a job is skipped. Alone, nobody does until the cost arrives. The fix is a cadence: each recurring duty gets a slot, and the slot is kept even in a busy week.

Swipe left to see what happens when a duty is skipped.
WhenDutyWhat happens if you skip it
Every dayAnswer the support inboxSmall questions turn into refunds and public complaints
Every weekDo one thing that brings new buyersRevenue holds for a while, then falls with nobody new arriving
Every weekRead the week's numbers: signups, payments, cancellationsA problem is noticed a month after it started
Every monthClose the books and pay the billsTax deadlines and failed card charges arrive together
Every monthReview the risk register and change what movedThe thing that ends the business is the one nobody wrote down
Every quarterCheck access: domains, payment accounts, backupsAn expired domain or a locked account takes the business offline
An operating rhythm for one person, not a measured schedule. The right column is why each row exists: every skipped duty has a delayed, specific cost.

Selling is the duty that is easiest to drop, because support and fixes feel urgent and selling does not. It is also the one whose absence shows up last, as a quiet month a quarter later. Protect a weekly slot for it, and move it by a day rather than skip it.

Keep a risk register, and read it every month

A risk register is a short list of the things that could end the business, each with what you are doing about it and a status. Ours is public on the page about how we might die. The statuses below are counted from that list, so this figure changes when the list does.

1
Open
7
Being tested
2
Closed
  • Founder tooling for founders is a crowded, low-trust marketBeing tested
  • No durable distribution channelOpen
  • Demand is not proven beyond our own painBeing tested
  • Subscription fights episodic usageBeing tested
  • Scope too broad for one personClosed
  • One wrong KILL destroys trustBeing tested
  • AI costs eat the margin at a low price pointClosed
  • Built from self-recognition, not external demandBeing tested
  • Payments friction (no direct Stripe; merchant-of-record instead)Being tested
  • A kill-biased product invites disputes and support loadBeing tested
WhittleOS's own register, read from the same list the public page renders. We wrote it by running our own idea through our own gate and keeping every reason it gave.

Three habits make a register useful rather than decorative. Each risk says what would settle it, so it can be closed, not just worried about. The list is read on a fixed date, not when something already feels wrong. And a risk that stays open for months is a decision waiting to be made; the guide to shutting down a project shows how to write that decision down in advance.

Three limits that end one-person businesses

These limits rarely end a business in a single day. They end it when the week stops fitting.

  • Support outgrows the week. Support rises with every customer. On the shared scale we use, below 2 hours a week is light and above 8 it has become the job. The support-burden tool shows where your idea lands.
  • Custom work creeps in. Each yes to one customer's special request makes the offer a little more bespoke, and bespoke work cannot be handed off or automated.
  • One platform holds the business. An app store, a marketplace or a single provider can change the rules. In our Discovery runs measured on 2026-09-06, platform dependence was named in 19.8% of the ideas the gate stopped. The platform-risk guide covers what to keep outside its control.

Plan for the week you cannot work

Alone, you are the single point of failure. A few preparations make an illness or a vacation an inconvenience instead of an outage:

  • An away message and a written promise. Tell customers how quickly they will hear back, and set the auto-reply before you need it.
  • A runbook for the common fixes. Write down the steps for the problems that recur, so someone you trust could follow them.
  • Access someone else can use. A trusted person should be able to reach the domain registrar, the payment account and the backups in an emergency.
  • Nothing that must be done by hand every day. If a daily manual step keeps the product alive, automate it or remove it before you take a week off. The guide on building a business that does not depend on you covers how to design those steps out.

Decide in advance what to drop when the week overflows

Some weeks will not fit. Decide the order in advance, so a busy week does not decide it for you:

  1. Anything that stops customers from using what they paid for. A broken product costs trust and refunds at once.
  2. Payments and billing. Failed charges and wrong invoices compound quietly.
  3. The weekly selling slot. It can move by a day, but a skipped month shows up later as a missing quarter.
  4. Improvements and new features. These are the first to wait, because nobody is harmed today when they do.

The order is easy to write and hard to keep, because the fourth item is usually the most enjoyable. Writing it down is what lets you defend it on a bad Thursday.

Which businesses stay runnable alone

The businesses one person can keep running share a shape: a fixed offer, setup customers can do themselves, support measured in minutes rather than hours, buyers who arrive without sales calls, and no single platform that can switch them off. Choosing that shape at the start is easier than retrofitting it. The guide on starting a business alone covers that choice, and the founder read suggests directions that fit the time and tolerance you actually have.

Common questions

Can you run a business completely alone?

Yes, if the offer is fixed, customers can set up without you, support stays in minutes per customer, buyers arrive without sales calls, and no single platform can switch the business off. Those five decide whether the week keeps fitting.

What is the hardest part of running a business alone?

Keeping the duties that are not urgent, above all selling and the monthly books, from being pushed out by support and fixes. A fixed weekly rhythm protects them.

What happens to a one-person business when the owner is sick?

It depends on preparation: an away message, a runbook for common fixes, emergency access for a trusted person and no daily manual step. With those, a week off is an inconvenience rather than an outage.