A one-page business plan, free — and why most tools stop there
A one-page plan is seven boxes, and it should take an afternoon at most. The generator below fills them in from four answers and does the arithmetic — free, no sign-up, nothing sent anywhere. Then the part nobody selling plan software wants to say out loud: the plan is the easy half, and everything that decides whether the business exists in six months happens in the week after you've written it.
Business plans have a reputation problem, and it's mostly deserved. The forty-page document with a five-year revenue projection in it exists to satisfy a bank, a landlord or a grant committee — and if one of those is asking, write it. But nobody ever started a business because the spreadsheet in appendix C said month thirty-one was profitable. The projection was invented. Everyone in the room knows it was invented.
What is genuinely useful is much shorter, and it's useful for a reason that has nothing to do with documents: writing it forces you to choose. One customer instead of everyone. One price instead of "depends". One channel instead of a list of nine you'll do a bit of each. Most first plans fall apart at exactly the point where a choice is required, and that's the whole diagnostic value of the exercise — the boxes you can't fill in are the ones your business hasn't decided yet.
So: seven boxes, an hour, and then go and do something. Here's what belongs in them.
What a one-page plan should contain
- 1 · The problem
- One sentence about a person, not a market. If it isn't true of somebody you could name and phone, the rest of the plan is fiction built on it.
- 2 · The customer
- Who it's for — and, in the same breath, who it isn't for. A plan that can't name its non-customer hasn't chosen anything yet, and choosing is the point.
- 3 · The offer
- One thing, described in a sentence a stranger would repeat correctly to someone else. If it takes a paragraph, it isn't an offer yet.
- 4 · Pricing
- An actual number, and how it moves. Raise on new customers first; leave existing ones alone for ninety days. That way a price rise costs nothing and tells you something fast.
- 5 · First channel
- One. Not five with a bit of each. Name the first twenty people or businesses you'll reach and what you'll say — that's a channel, a list of platforms isn't.
- 6 · The numbers
- Revenue at the rate you're actually running at, costs including your own pay, and break-even as a count of jobs or customers rather than a dollar figure.
- 7 · First milestone
- One number, ninety days, written somewhere you'll see it on a Monday. Not "grow" — "from twelve jobs a month to twenty-two".
Fill in the seven boxes
Four answers and three numbers. It composes the plan from a structured template and does the arithmetic — break-even, the monthly gap, what the milestone should be — in your browser. No sign-up, no email, nothing leaves the page, and no AI wrote a word of it, which is exactly why it's honest: a paragraph generated from four inputs about a business it has never seen would read like it knew something it doesn't. What it gives you is the right shape with your own numbers in it. Copy it, argue with it, change the sentences that are wrong.
- Short each month
- $2,200 $3,000 in, $5,200 out.
- Break-even
- 87 jobs a month to cover costs. You're at 50.
- The problemCommuters and weekend riders who can't get a bike to a shop currently put up with the inconvenient version of this — booking days ahead, taking time off work, or getting the thing to someone else. That is the whole opening — if it isn't true of a real person you could name, the rest of this plan is fiction.
- The customerCommuters and weekend riders who can't get a bike to a shop, in and around Denver. Not the customer: anyone shopping purely on price — they take the longest and leave the worst reviews. A plan that can't say who it is not for hasn't chosen anything yet.
- The offerMobile bike repair, done at the customer's door. Booked on the site, paid on the site, done at their address. One thing, described in a sentence a stranger repeats correctly to someone else.
- Pricing$60 a job. Raise it on new customers first and leave existing ones on the old price for ninety days — that way a price rise costs you nothing and tells you something within a fortnight.
- First channelLocal search, and the businesses that already have your customer. Two things, in this order. One page that answers the exact search someone in Denver makes when they need this, so you turn up when they look. Then a list of twenty local businesses that already serve the same person — the shops, gyms, offices and property managers nearby — and one honest email each, five a day, not twenty at once. Ask for a referral arrangement, not a favour.
- The numbers50 jobs a month at $60 a job is $3,000 a month coming in. $5,200 a month goes out. That's $2,200 a month short. Break-even is 87 jobs a month — the number to put on the wall.
- First milestoneGet from 50 to 87 jobs a month within ninety days — 37 more, which is about 3 extra a week for thirteen weeks. Write that number somewhere you'll see it on a Monday.
This is a structured template with your arithmetic in it — the seven boxes, filled in from what you typed, composed in your browser. Nothing is sent anywhere and no AI wrote a word of it. Velofound's own version starts from a conversation about your actual business, and then goes and builds the website the plan describes.
The plan is the easy half
You now have a page. Nothing has happened. There's no address a customer can visit, no way to take a payment, no LLC, no listing, no first email sent, and no number that tells you how long the money lasts. That gap — between a plan and a business — is where almost everything dies, and it isn't a motivation problem. It's twenty jobs, each of which is dull, unfamiliar and takes a Saturday: registering an entity you've never registered, choosing between six payment processors, working out what a website should even say, deciding whether $60 or $75 is the right price with nothing to base it on.
The tools that generate plans mostly stop precisely here, because this half is hard to build and impossible to fake. Anyone can produce a document. Producing a live page on a real address that charges a real card into a bank account in your name means integrating with Stripe, hosting the page, handling receipts and refunds and payouts, and being responsible when it breaks at 11pm on a Saturday. The gap between those two things is not a matter of degree.
The honest test to apply to any of these tools, including this one: at the end of the session, what exists that didn't exist before? A document is not nothing — but a document is not a business, and the difference between them is the entire job.
What the week after actually looks like
Northline Bike Repair is the business used as the running example across these pages: one founder, a van, Denver, mobile bike repair. Its one-page plan looked much like the one the generator above produces. Here is what each line of it turned into, and where it stands in week six — including the line that hasn't happened yet.
| What the plan said | What exists now |
|---|---|
| Offer — a 40-minute tune-up at the customer's door, $60. | A booking site live on its own address with the services priced, taking a $30 deposit at the point of booking. $1,286 collected this month across four paid bookings, into Northline's own Stripe account. |
| Customer — commuters and weekend riders who can't get a bike to a shop. | 214 people have visited from one Facebook ad and four of them paid. That is a small number and it's a real one; the rest of the plan's assumptions about who this is for are still assumptions. |
| First channel — local paid social. | Four ads on four angles, written and published to Facebook and Instagram paused for approval. $300 spent, $1,486 back — counted from Stripe charges rather than from Meta's reporting. The recommended budget is capped by runway. |
| Costs — van, insurance, tools, parts, and the founder's own pay. | $5,200 a month, read from a bank CSV. Burn and runway recomputed every night from real revenue against real costs: 8.4 months, up 0.3 on last week. |
| Legal — form the LLC, get an EIN. | Colorado's steps, fees and the official filing portal on a checklist, in order, with a guide for each. Velofound never files anything with a government, so the founder did that part. |
| Milestone — cover costs. | Not there yet. Break-even is in the morning brief every day and the gap is what the night shift works on: the leads it qualifies, the follow-ups it drafts, the ad angle it wants to move the budget onto. |
That last row is the point of showing the whole table. Six weeks in, Northline is not profitable, and no tool made it profitable. What changed is that the plan stopped being a document: there's an address, a charge, a real cost figure and a number of months. Every one of those is now a fact that can be argued with, which is the only useful state for a plan to be in.
When a longer plan is genuinely worth writing
Three cases, and they're all cases where somebody else is going to read it. A loan. An SBA lender or a bank will want a full plan with financial projections, and they will want them in a specific shape — usually three years, monthly for the first year, with the assumptions written down beside the numbers. Do it properly; the projection being invented is understood on both sides, but the assumptions being coherent is what's actually being assessed.
A grant. Grant applications ask their own questions in their own order and reward answering the question that was asked. A one-pager is the raw material for that, not a substitute.
A partner or a first employee. When someone else's livelihood attaches to the plan, the useful long document isn't a projection — it's the boring one: who owns what, who decides what, what happens if one of you leaves, and how profit gets split. That's an operating agreement, and it matters far more than the marketing section. The formation steps, state by state →
For everyone else: keep the page, put the milestone somewhere you'll see it, and go and get the first customer. If you want the runway maths done properly rather than estimated, that has its own calculator.
Common questions
Is this business plan generator really free?
Yes, and there's nothing to sign up for. It runs entirely in your browser: nothing you type is sent to a server, stored, or emailed to you in exchange for the plan. Copy the result and close the tab if you like.
Does it use AI to write the plan?
No, deliberately. It's a structured template plus arithmetic — the seven boxes, filled in from your answers, with break-even and the ninety-day gap calculated from your price and costs. A model given four inputs about a business it has never seen would produce fluent paragraphs that sound more certain than they are. Velofound's own version is different: it comes from a real conversation about your business and it ends with the website built, not with a document.
How long should a business plan be?
One page, unless somebody who is going to give you money has asked for more. A lender, a grant committee or a landlord will want the full version with financial projections; for everyone else the value is in being forced to choose a customer, a price, a channel and a number, and forty pages actively hides whether you've done that.
What's the difference between a business plan and a lean canvas?
Very little, at this size. A lean canvas is nine boxes on a landscape page and this is seven on a portrait one; both exist to make you commit to a customer, a problem, an offer and a way of reaching them in a form you can revise weekly. Use whichever you'll actually update.
Do I need a business plan to get a loan?
For an SBA loan or a bank loan, yes, and a proper one — three years of projections, monthly for the first year, with your assumptions written beside the numbers. A one-page plan is the raw material you write it from. It isn't a substitute, and no generator's output should be submitted as-is.
Should I write the plan before or after building anything?
Before, but only just. An hour on the page and then straight into the first real thing — usually a page a customer can pay on, because a charge answers questions a plan can only guess at. Writing for a week before building is the most common way to spend a month learning nothing.
What does Velofound do that a plan generator doesn't?
It builds what the plan describes and then runs it. The website goes live on its own address taking real card payments through Stripe into your own account; you get your state's legal checklist, pricing from real costs, social posts with photos on a calendar, a Money page with burn and runway from actual revenue, a business inbox with drafted replies, and an overnight shift that researches and drafts while you're asleep. Nothing gets sent, posted or charged without you approving it.
Can I edit the plan afterwards?
Copy it out and it's yours — it's plain text, so it goes into a document, an email or a note. Change the sentences that are wrong; the templated ones are a first draft written to be argued with, not a verdict on your business.
You have the plan. Now the other half.
Describe the business the way you just described it here. The website goes live in about a minute, ready to take payment, and the legal checklist, the pricing and the numbers follow. Free to start, no card.
Build it free →