How to write a business plan that a lender or investor will actually read

Start with who is reading it

A business plan written for nobody is read by nobody, and the three audiences that matter want different things from the same document. Deciding which one you are writing for changes the order of the sections, the tone of the executive summary, and which risks you put in front rather than in an appendix.

If you are going to a lender — which is where most UK small-business plans end up — write the document they read. That means conservative revenue, an explicit sensitivity case, and a page that says what happens if the numbers miss. GigBlend’s business plan service is built lender-facing by default for that reason.

The structure, and why the order matters

The order below is the order a credit officer reads. The executive summary comes first physically but is written last, always.

1. Executive summary — one page, written when the rest is finished.
2. The business: what you sell, to whom, and how you get paid.
3. Market: who the customers are, how many there are, and how you know.
4. Sales and marketing: the specific route to the first £100k and the cost of it.
5. Operations: how the thing is actually delivered, by whom, and what could stop it.
6. The team: who is doing the work and what the gaps are.
7. Financial plan: three statements plus the assumptions behind them.
8. Risks and mitigation: the honest page.
9. Appendices: CVs, quotes, tenancy, letters of intent.

Executive summary: one page, four answers

A reader who gets to the bottom of page one should be able to answer, without turning over: what the business sells, who pays for it, what it earns today or in year one, and how much money you need and what it buys. If any of those four take them past page one, the summary is not finished.

Write this section last. It is the only part of a business plan that improves by being rewritten after the analysis exists, because the summary is a claim about the numbers, and you cannot make the claim before you have them.

The market section: sized, not asserted

The most common single error is the back-of-envelope global figure — “a £40bn market, we only need 0.1%”. A reader who does this for a living has seen that line a thousand times and it costs you the rest of the document. Build it up instead:

• Start from a number that exists: ONS business counts, Companies House filings for comparable companies, trade-association figures, or your own local count.
• Narrow to the people you can actually reach in the first two years — a geography, a sector, a size band.
• State a price and a plausible penetration rate, and show the multiplication.
• Say what would make your penetration assumption wrong.

Competitive analysis belongs here, and it belongs with no enemies. A plan whose competitor table gives every rival a weakness and the author nothing but strengths is read as marketing, not research. Put your own genuine weaknesses in the same table.

Financials: three statements and the assumptions page

You need a profit and loss forecast, a cash-flow forecast, and a balance sheet. The profit and loss tells the reader whether the business is viable; the cash flow tells them whether it survives; the balance sheet ties them together. The assumptions page — prices, volumes, growth rates, wage costs, the payment terms you have agreed and the ones you have assumed — is what makes the other three credible, and it is the page lenders check first for internal consistency.

• Monthly for year one, quarterly for years two and three. Annual-only forecasts read as a guess.
• Cash, not profit: a invoice raised on 31 March is not money in April if the terms are 60 days.
• VAT at the rate that applies to you, on the date it is actually paid to HMRC, not invoiced.
• Owner drawings and any director salary, in the costs. A plan that forgets the founder’s wage is a plan that fails.
• A downside case where revenue lands at a fraction of plan, shown against the same cost base.

The arithmetic is where amateur plans fail, and it is why ours is generated rather than written. In the fulfilment model behind this service, the financial model is computed by code — amortisation, VAT, cash timing, growth compounding — and the narrative is then written to match those numbers. Five-year totals reconcile to the monthly rows, and a reviewer can change one assumption and watch the statements move.

Risks: the page that gets you funded

Counter-intuitively, naming your real risks raises the chance of approval. A lender is not looking for a plan with no risks; they are looking for a founder who knows what kills the business and has thought about it. Three or four genuine risks, each with what you would actually do, in the body rather than an appendix.

Length, tone and what to cut

• Twelve to twenty pages of body for most applications. Longer than that and the reader skims, which is worse than being short.
• Present tense, plain verbs, no adjectives that could be argued with. “Category-leading” is a claim; “repeat rate of 41% across 24 months” is a fact.
• Every number in the narrative must appear in, or reconcile to, the financial model.
• Charts over tables where the shape matters; tables where the reader will check the arithmetic.

Doing it yourself, or paying for it

You do not need to buy anything to write a good plan, and for a very small loan a competent one-page plan is often enough. Where a paid plan earns its fee is when a third party will scrutinise it: a lender, an investor, a grant panel, a landlord, or a business sale. In those cases the value is not the writing, it is the model behind it being internally consistent and defensible under questions.

GigBlend sells that as a fixed-price service with the scope on the page: market research, written narrative, a three-year financial model computed in code, and revision rounds. The fulfilment route is disclosed before you pay — you can see whether a person, an AI agent, or both are doing the work.