How to write a business plan founders actually use
A practical guide to writing a business plan: the sections that matter, an evidence and assumption checklist, and how to keep the plan current as facts change.
Quick answer
A useful business plan states the problem, who has it, how you solve it, how you make money, how you reach customers, what it costs, and what you still need to prove. Write each section from evidence you already hold, label every assumption explicitly, and revisit the plan whenever a core fact changes.
Decide who the plan is for before you write a word
A plan written for yourself, a bank, a grant panel and a seed investor are four different documents drawn from the same underlying facts. Investors read for scale and evidence of demand. Lenders read for repayment capacity. Grant assessors read for eligibility and impact against published criteria. Internal plans exist to make trade-offs visible. Pick the primary reader first; write the rest of the plan to serve that reader without contradicting the others.
- Investor: market, wedge, traction, unit economics, team, use of funds
- Lender: cash flow, security, downside case, repayment schedule
- Grant panel: eligibility, defined outcomes, delivery capability
- Internal: priorities, resourcing, what you will not do this year
The sections that carry weight
Length is not credibility. A tight 10–15 page plan that answers the reader's real questions beats forty pages of filler. Work through these sections in order — each one should be readable on its own, because reviewers rarely read front to back.
- Executive summary — written last, and no longer than a page
- Problem and who has it, stated in the customer's own words where you have them
- Solution and what is actually built today versus planned
- Market: who you can realistically reach, not a global category number
- Business model: pricing, unit economics, and what drives each
- Go-to-market: the two or three channels you will test first
- Competition and your honest differentiator
- Team and the gaps you know you have
- Financials: assumptions first, numbers second
- Risks, and what would make you change course
Separate what you know from what you believe
The single biggest quality improvement in a founder plan is labelling. Every material statement is either evidenced (you can point to a source), inferred (a reasonable read of evidence you hold) or assumed (a placeholder you have not tested). Readers forgive assumptions. They do not forgive assumptions dressed as facts, and one over-claimed number can cost you the credibility of the whole document.
- Evidenced — a contract, an analytics export, a signed pilot, a published statistic you cite
- Inferred — derived from evidence you hold, with the derivation shown
- Assumed — a working number with no support yet; say so and say how you will test it
The evidence and assumption checklist
Before you send a plan out, run it against this list. If a line has no evidence and no explicit assumption label, it is a liability rather than a claim.
- Can you name the source behind every number in the plan?
- Does the revenue model in the plan match the model in your forecast?
- Does the customer described in the plan match the customer in your deck?
- Is any statistic older than two years, or from a source you cannot link?
- Have you stated what would falsify your core assumption?
- Does the team section describe roles you have, not roles you hope to hire?
- Is the use of funds tied to specific milestones rather than a runway figure?
Consistency is the quiet killer
Plans, decks and forecasts are usually written at different times by different means, and they drift. A pricing change made in a spreadsheet rarely reaches the plan's revenue narrative. An investor who spots two different customer-acquisition costs in the same pack stops reading for insight and starts reading for errors. Before any external send, cross-check the small set of facts that appear in more than one document.
- Pricing and packaging
- Target customer and segment definition
- Headline market sizing method
- Revenue and cost drivers
- Headcount plan and burn
- Funding amount and use of funds
Write from what you already have
Most founders have far more raw material than they think: product notes, customer emails, analytics exports, a landing page, competitor research, AI chat threads where the strategy was actually argued out, and the codebase itself. Starting from a blank template throws all of that away and invites invention. Start by gathering the material, then draft each section against it.
- Product notes, specs and roadmaps
- Customer conversations, support threads, survey responses
- Analytics and usage exports
- Pricing experiments and invoices
- Existing pitch materials, however rough
- AI chat exports where the thinking happened
Keep it living, not archived
A plan is only useful while it reflects reality. Set a light cadence: review monthly, and review immediately whenever a core fact moves — pricing, target customer, funding position, a key hire, or a strategy pivot. Record what changed and why. That trail is what lets you answer an investor's 'what have you learned since we last spoke?' with something specific.
Home · Features · How it works · Pricing · Compare · For founders · Try free