There is a researched answer to the question of when a founder should write a business plan, and it is almost never “right now.” Francis Greene and Christian Hopp tracked 1,000 would-be American founders over six years using data from the Panel Study of Entrepreneurial Dynamics II, and the ones who fared best wrote their plan somewhere between six and twelve months after deciding to start. Sitting in that window raised their probability of reaching a viable venture by 8 percent. Writing it on day one did not.
That finding irritates people, because the advice handed to every new founder is to write the plan first and then go run the business. The data points the other way. Once you see why, most of the other decisions about a plan get simpler, so what follows is when to write it, how long to spend on it, and which of your three possible readers you are actually writing for.
Planning Works. That Was Never the Argument.
The same two researchers published an earlier study, written up in Harvard Business Review, showing that entrepreneurs who write formal plans are 16 percent more likely to reach viability than otherwise identical founders who skip it. So the question was settled a while ago. Plans help.
What the follow-up work added is that the help is conditional. A plan written before you have met a customer is a record of what you assumed in a room by yourself. Greene put it bluntly in a 2018 piece for the University of Edinburgh Business School: formal plans written at the very start are close to fiction, and the hours spent on them are hours not spent finding out whether the original customer was the right one.
There are exceptions worth naming. If a lender has given you a deadline, or an accelerator application closes in three weeks, you write the plan now and you write it to the deadline. Timing advice is for founders who have a choice.
Three Months Is a Ceiling, Not a Target
The Edinburgh research found the optimal amount of time to spend planning was about three months, which lifted the chance of a viable venture by 12 percent. Spending longer than that produced nothing, and the reason is worth understanding: the information you gathered in month one has gone stale by month seven. Competitor pricing moves. A supplier quote expires. The channel you were counting on changes its terms.
So the practical constraint is not “write more.” It is “finish before your inputs rot.”
It also argues for doing the cheap version first. A business model canvas fits on one page and takes an afternoon, and it will show you whether the revenue model has a hole in it well before a forty-page document would.
That constraint has an unglamorous implication. A meaningful share of the three months tends to disappear into deciding what the sections should be called and where the financial tables go. Starting from a five-year business plan template that already carries the executive summary, market analysis, operational plan, financial projection, funding requirement and risk assessment means the argument about section order never happens, and the three months go to the market instead of the margins.
Write It When You Start Talking to Customers
This is the sharpest finding in the research and the one almost nobody acts on.
Greene and Hopp looked at what else the founder was doing at the time the plan got written. Writing it alongside early activities like defining the market or collecting competitor information added nothing measurable. Writing it after the business had already hired people or raised outside money added nothing either. By then the decisions the plan was supposed to inform have been made.
The window that mattered was the one where a founder is talking to real customers and getting the product ready to sell. Planning at the same time as those two activities raised the chance of venture viability by 27 percent, which is more than three times the lift from timing alone.
The mechanism is not mysterious. During that stretch you are collecting the only inputs a plan genuinely needs: what people say when you quote a price, how long a decision takes, which objection keeps coming up, what a customer thought you sold before you corrected them. A plan written in that noise is a plan built on evidence. A plan written before it is a tidy guess.
This is also the stretch where founders find out the model needs to change rather than the wording. When that happens, reading how other companies put theirs together tends to be faster than reasoning your way to a new one from scratch.
Decide Who Is Reading It Before You Decide What Goes In It
A business plan has three possible readers, and they want three different documents.
A lender wants to know whether you can make the payment. An equity investor wants to know how big this gets if it works. You want to know which of your beliefs to check first. Most plans read badly because they try to serve all three at once, which produces a document that persuades nobody and guides nothing.
Pick the reader before you pick the sections. If you are writing for two of them, write two documents and let them share a spreadsheet.
There is a fourth reader worth separating out, because founders keep sending them the wrong document. A supplier deciding whether to extend terms, or a prospective client deciding whether you will still be around in a year, is not evaluating your growth thesis. They want to know who you are, how long you have been doing this, and what you sell. That is a company profile, a couple of pages long, and sending a business plan instead reads as though you did not understand the question.
What a Lender Actually Does With It
Founders tend to imagine an underwriter reading the plan the way they wrote it, front to back. That is not what happens. The narrative gets skimmed and the numbers get rebuilt.
The SBA spells out what its lenders have to produce, and reading that list is more instructive than most advice about writing plans. Under its 7(a) underwriting requirements, the lender has to document a justification for your projected revenue growth, naming things like new product lines or new sales channels, and compare your projections to current industry trends. The lender also has to build a pro-forma balance sheet and calculate a specific set of ratios: current ratio, debt to tangible net worth, and debt service coverage.
Debt service coverage is the one that decides most files. Under the current SOP, SBA lenders need to show a projected ratio of at least 1.15, meaning the business throws off $1.15 of cash for every dollar of loan payment. Plenty of lenders set their own internal floor at 1.25 or higher, and thresholds vary by lender and by deal, so treat 1.15 as the bottom of the range rather than the target.
What that list changes about your writing is specific. A revenue line that grows 30 percent a year because you typed 30 percent is not a justification, and an underwriter has to write a paragraph explaining it. Give them the paragraph: a signed contract, a second location opening in March, a distributor agreement with terms. If the growth has no name, take it out of the projection.
The same logic applies to the ratios. Since the current ratio and the debt service coverage are getting calculated whether you like it or not, you may as well see the numbers before the bank does. A financial projection spreadsheet that runs your sales forecast, payroll, operating costs and loan payments through an income statement, balance sheet and cash flow statement will produce those ratios off your own assumptions, and finding a coverage problem at your desk in July is a much better experience than hearing about it from an underwriter in September.
The Version You Keep for Yourself
The internal plan is a different animal, and it is shorter than founders expect.
Its job is to record what you believed and when, in a form you can check later. Most internal plans fail at this because they write down conclusions instead of assumptions. “We will reach $40,000 in monthly recurring revenue by Q3” is a conclusion. “About 3 percent of free trials convert, and we can put 400 trials a month through the funnel by June” is a pair of assumptions with dates attached, and in June you can look at both and see which one broke.
That difference is what separates a plan you use from a plan you file. When revenue misses, the founder with conclusions written down knows only that something went wrong. The founder with assumptions written down knows whether the belief was wrong or the execution was, and those two problems have nothing in common.
Keep it to a handful of numbers you would actually bet on. The forty pages of market sizing can stay in the lender’s copy.
Put the review date in while you are writing, not afterward. A plan you reassess on a schedule stays a working document. One you open only when something has already gone wrong is a postmortem with better formatting.
A Short Check Before You Start Writing
- Are you already talking to customers? If not, that comes first and the plan waits.
- Have you named the single reader, and can you say what decision they are making?
- Is there a hard deadline that overrides the timing advice? Write to the deadline.
- Does every growth number in the forecast have a reason with a name attached?
- Have you run the coverage and liquidity ratios yourself before anyone else does?
- Is the internal version short enough that you will reread it in four months?
The Bottom Line
A business plan is a dated document. It is worth what its inputs were worth on the day you wrote it, which is why the research keeps landing on timing rather than length or format.
Write it once you have customer conversations to write about, give it about three months, aim it at one reader, and make the numbers checkable by the person who has to check them. The founders in that six-year study who did roughly this reached viability more often than the ones who wrote a beautiful document in week one and never opened it again.
