By Amit Jain · curated with Vinod Kumar Jain · All Frontier Global · 2026-07-05
A business plan is a document written to change what a specific reader believes and does next — approve a loan, write a cheque, sign off a grant, or simply stop arguing about the same unresolved question at every board meeting. Most of the plans that fail do not fail because the business behind them is weak; they fail because the document was written for no reader in particular, or for the wrong one. This page is about what a business plan actually contains, who reads it, how the financial model inside it is built, and why so many plans are assembled backwards — numbers first, story second, reader never.
| Document | What it owns | What it deliberately leaves out |
|---|---|---|
| Business plan | The whole enterprise: what it does, how it makes money, what it needs, what could go wrong, what the numbers look like if it works | Channel-by-channel execution detail, campaign calendars, individual seller quotas |
| Go-to-market strategy | The route to market — which segments, which channels, in what sequence, with what resourcing | Full company financials, cap table, legal structure, risk register |
| Marketing plan | Demand — positioning, messaging, channel mix, content, brand, the demand-generation budget | Sales process, quota-setting, the enterprise-level P&L |
| Sales plan | Conversion — the selling motion, pipeline stages, quota, compensation, the sales team itself | Product roadmap, funding requirement, market sizing methodology |
This site carries three companion pages that sit close enough to a business plan to be confused with it, and the confusion is worth clearing up before anything else, because writing the wrong one wastes the same weeks whether you are a founder or an owner-manager. A go-to-market strategy is about sequencing and channel choice for taking a specific offering to a specific market — which segment first, through which route, with what resourcing behind it. A marketing plan owns demand: how prospects come to know the offering exists and want it. A sales plan owns conversion: what happens once a prospect is in front of someone whose job is to close them, including quota, compensation and pipeline management. Each of those three is bounded to a function.
The business plan is not bounded to a function. It is the document that describes the whole enterprise — what the business does, how it makes money, what it needs to operate and grow, what could plausibly go wrong, and what the numbers look like if the plan succeeds and if it does not. Where the plan touches go-to-market, it should say so in summary and point at the GTM document for depth rather than duplicating it. The same is true of marketing and sales detail. A business plan that tries to also be a full GTM strategy, a full marketing plan and a full sales plan becomes unreadable and, worse, becomes stale in four different ways at four different speeds. Keep the boundary. The business plan asks whether the enterprise as a whole holds together; the three siblings ask whether one function inside it is executed well.
This distinction also settles a common argument about sequencing. Some founders want to write the business plan first and derive the GTM, marketing and sales plans from it; others build GTM and marketing plans first and let the business plan summarise them once they exist. Both orders work. What does not work is writing all four as one undifferentiated document, because a reader evaluating the whole enterprise does not want the sales team's compensation structure, and a reader evaluating the sales motion does not want the balance sheet.
A business plan is written for someone. That someone determines almost everything about the document — what goes in, what is left out, how long it runs, and what tone it takes. Writers who skip this question produce a document that tries to serve everyone and persuades no one.
An equity investor is reading to answer one question: can this return a multiple of the money on a timescale that matters to their fund, and what has to be true for that to happen. They are comfortable with an unproven market and an unfinished product if the team, the mechanism and the size of the opportunity justify the risk. They read the team section and the market section hardest, and they read the financial model as a test of the founders' judgement more than as a forecast they believe line by line. A plan for this reader can be shorter than founders expect, because a great deal of the real work happens in conversation and diligence rather than in the document itself; the document's job is to earn that conversation.
A bank or lender is reading a completely different question: can this business service debt on the agreed schedule regardless of whether it grows spectacularly. Lenders are not buying upside; they are pricing downside. They read the cash flow statement and the security available against the loan harder than they read the market opportunity, and they want to see conservative assumptions, not ambitious ones. A plan written for a lender that reads like an investor pitch — heavy on vision, light on how repayments get made in a slow month — will be read as evasive, because it is answering a question the lender did not ask.
A grant body or a visa or licensing authority is reading against a published set of criteria, not against a general sense of merit. The plan for this reader has to map explicitly onto whatever the scheme rewards — job creation, export potential, innovation criteria, minimum investment thresholds, sector eligibility — and it should say so in the plan's own language rather than making the assessor hunt for the match. This is the one audience where matching a checklist matters more than narrative persuasiveness, and founders who write a compelling story instead of an explicit criteria match often lose to a duller plan that ticks every box.
A board is reading a plan they already partly own. Board members were, in most cases, present for the strategic choices the plan documents, so the plan's job for this reader is less to persuade and more to record: what was decided, what is being measured, what the resourcing implications are, and what the board is being asked to approve or note. A board plan can be blunter about weaknesses than an investor plan, because candour with a board that already has skin in the game builds trust, whereas the same candour with a first-time investor can read as a red flag before the relationship exists to interpret it correctly.
The plan written to make the founders think is the one nobody else will read, and it is arguably the most useful of the five. Its purpose is to force the founders to write down, in full sentences rather than a slide's worth of bullet points, what they actually believe about the business — the mechanism by which it makes money, the customer they are serving, the assumptions underneath the numbers. Writing forces a kind of precision that talking does not; ideas that sound settled in conversation frequently fall apart the moment they have to survive a paragraph. This version of the plan can be as unpolished and as honest as the founders like, because its only reader is them.
The failure mode is not usually that the plan is badly written. It is that the writer answered a question the reader did not ask. A founder who sends an investor-style plan — heavy on market size and vision, thin on downside protection — to a lender will find the lender unmoved, because the lender's question was never "how big could this get" but "what happens to my money if it does not". A founder who sends a bank-style plan — conservative, cash-focused, low on ambition — to an investor will find the investor uninterested, because caution is not what an equity return is bought with. The document can be well written and still fail completely, simply because it was aimed at the wrong target.
The practical fix is to write the reader down before writing anything else: name them, name the decision they are making, and name what they need to believe to make that decision in the founder's favour. Everything else in the plan should be traceable back to that one sentence. A plan with two very different readers — a bank and an investor, say, in the same funding round — often needs two documents sharing a common core, not one document trying to do both jobs.
These four are frequently used interchangeably and they should not be. A pitch deck is a presentation aid, built to be talked over, not read cold; it compresses the plan's argument into a sequence of visual claims and relies on the presenter to supply the connective reasoning live. A one-pager is a filtering document — its job is to get a meeting, not to close one, and it should contain only enough to make the reader want the next conversation. A financial model is a working spreadsheet, built to be interrogated, stress-tested and rebuilt as assumptions change; it is an instrument, not a narrative. The business plan is the only one of the four that stands alone as a complete, self-contained written argument — a reader should be able to pick it up with no other context and understand the business, the opportunity and the ask.
The common failure is producing only one of these four and trying to make it serve all the purposes. A deck with no underlying plan collapses under any serious follow-up question, because the reasoning that should sit behind each slide was never written down. A plan with no deck is hard to present live because prose does not compress well into a room's attention span. A model with no plan around it is numbers without a story, and readers who are asked to fund a spreadsheet rather than a business tend to decline. The four objects are complementary, built from a shared understanding of the business, each doing a job the others cannot.
There are situations where writing a full business plan is a waste of the time it takes. A very early idea that has not yet had a single real conversation with a prospective customer usually does not benefit from a fifteen-thousand-word document — it benefits from those conversations, after which the plan can be written with something to say. A business that is simply the founder replicating a known, well-understood local service — a single-site trade business with an established local market and no external financing need — may need a one-page operating plan and a cash flow forecast far more than it needs a document written for an investor who will never read it. A plan written purely because "you're supposed to have one," with no reader and no decision attached, tends to be shelved unread the day after it is finished, and the hours spent on it would often have been better spent talking to customers or building the thing. The test is always the same: name the reader and the decision. If neither exists yet, it is too early to write the plan, though it may already be time to start keeping the notes that will become one.
Before a plan can address market, strategy or numbers, it has to describe the business plainly: what problem it solves, how it makes money, what has to happen daily to deliver it, and the structural facts — legal form, ownership, intellectual property, regulation, supply chain, location — that shape everything downstream.
Every business plan opens, in substance if not always in its literal first paragraph, with a claim about a problem worth solving and an opportunity to solve it profitably. The discipline here is separating the two: the problem is what a customer currently experiences as friction, cost, risk or unmet need, described in the customer's terms rather than the founder's. The opportunity is the founder's specific answer to why this problem is solvable now, by this business, at a profit. Plans weaken themselves when they blur the two — describing the problem so generically that any number of businesses could claim to solve it, or describing the opportunity so specifically to the founder's product that the underlying customer need never gets independently established.
A useful test is whether the problem statement survives being read by someone who has never heard of the company. If it does not — if it only makes sense once you already know the product — the plan has skipped a step. The strongest problem statements are ones a sceptical reader could, in principle, go and verify by talking to five customers.
The offering section describes what is actually sold, in plain terms: the product or service, its form, what is included and excluded, how it is packaged and priced, and what distinguishes it from doing nothing or from an obvious alternative. This section should resist the temptation to sell — its job is to describe accurately enough that a reader unfamiliar with the space understands exactly what changes hands for money. Ambiguity here (vague packaging, unclear pricing logic, an offering that seems to change shape depending on which paragraph you read) is one of the more common reasons diligence conversations stall, because the reader cannot get a straight answer to "what exactly am I being asked to believe people will pay for."
The business model answers three linked questions and the plan should answer them in that order. Value creation: what is actually produced or assembled that did not exist before, and why does it matter to the customer. Value delivery: how that value physically or digitally reaches the customer — the mechanism of supply, whether direct, through partners, through a platform, or some combination. Value capture: how the business converts delivered value into revenue — subscription, transaction fee, licence, one-off sale, usage-based charge, or a blend — and why that particular capture mechanism suits this particular value.
Founders sometimes describe their business model as a single word — "SaaS," "marketplace," "subscription" — as though the label does the explanatory work. It does not. Two subscription businesses can have entirely different economics depending on acquisition cost, churn, and what triggers cancellation; two marketplaces can have opposite unit economics depending on who pays the take rate and how liquidity is built on each side. The plan needs to walk through creation, delivery and capture as three distinct, connected steps rather than reaching for a category label and assuming the reader fills in the rest.
Separate from the business model, which is the economic logic, is the operating model: what actually has to happen, day in and day out, for one customer to be served correctly. This is where plans often go thin, because it is less glamorous to write than strategy or market size, and yet it is frequently where the real risk sits. If the offering depends on a small skilled team that cannot be hired at the pace the revenue plan assumes, that is an operating-model risk that undermines the whole forecast regardless of how sound the market analysis is. If delivery depends on a manual, founder-dependent process that has never been tested at twice today's volume, the plan should say so rather than assuming scale is a numbers exercise.
A useful discipline is to trace a single customer's journey from the moment they decide to buy to the moment they have fully received what they paid for, listing every step that has to happen and who or what performs it. Steps with a single point of failure — one person, one supplier, one piece of undocumented knowledge — are operating-model risks worth naming explicitly, both because a sophisticated reader will look for them and because naming them is the first step to fixing them.
The legal and ownership facts belong in the plan even though they rarely change a reader's mind on their own, because a reader who cannot find them will assume the worst or simply stop and ask, which slows everything down. State the legal structure and jurisdiction of incorporation, the ownership split among founders and any existing investors, and any options or convertible instruments outstanding that would dilute a new investor. State plainly what intellectual property the business holds or has applied for, what is registered versus what is merely claimed, and — this matters more than founders often realise — whether any of it was created while the founders were employed elsewhere, which is a question investors and acquirers ask early and one a founder should have a clean answer to before it is asked. Where IP is central to the defensibility claim made elsewhere in the plan, this section and the competitive-defensibility section later in the plan should agree with each other; a plan that oversells IP strength in the market section and underplays it here, or vice versa, reads as either careless or evasive.
Where the business operates in a regulated space — financial services, healthcare, food, alcohol, education, anything requiring a licence — the plan should state the regulatory position honestly: what licences or approvals are held, which are pending, and what the business does if an approval is delayed or refused. Readers in regulated sectors will discount a plan heavily if the regulatory section reads as an afterthought, because regulatory risk is frequently the single largest risk to timeline in these businesses.
Supply chain and dependency should be described concretely: who supplies the inputs the business cannot do without, how concentrated that dependency is, and what happens if a key supplier fails or changes terms. A business with a single supplier for a critical input carries a specific, nameable risk that belongs in this section and is cross-referenced in the risk register later. Location and premises matter for different reasons in different businesses — a business with physical footfall depends on a fact about a specific address; a fully remote business may need only a sentence on this. Write proportionately: enough to answer the question, not padding for its own sake.
Market sections are where plans most often lose credibility, usually through a sizing method that does not survive scrutiny. This part covers how to size a market honestly, define the target customer, assess competition including the option of doing nothing, and treat claims of defensibility with appropriate scepticism.
Market sizing is usually presented one of two ways, and a sophisticated reader treats them very differently. A top-down number starts from a large published figure for an industry or category and applies a percentage to arrive at an addressable slice — "the total market is a large figure, and if we capture even a small percentage of it, that is still a substantial business." This method is popular because it is quick and because the resulting number is always impressively large, but it explains nothing about why the business would capture that percentage rather than any other, and readers who have seen enough of these plans discount them heavily on sight.
A bottom-up number is built the other way: start from the actual mechanism by which the business gets paid — customers times price, or units times price, or accounts times average contract value — and build up from a defensible estimate of how many of those the business can realistically reach in a given period through its actual channels. A bottom-up estimate is harder to produce and is usually a smaller, less exciting number than the top-down one, but it is the number a careful reader will trust, because it can be interrogated line by line: where does the customer count come from, why that price, why that channel capacity. The strongest market sections show both — the top-down figure to establish that the category is not trivially small, and the bottom-up figure to show the founders understand exactly how they get paid — and are explicit about which one is being used to justify the forecast in the financial model. Using the top-down figure to imply forecast credibility while the actual model is bottom-up, or vice versa without saying so, is the kind of inconsistency that experienced readers catch immediately.
Segmentation should say who the business sells to now, in enough specific detail that a reader could recognise one of these customers if they met one, and should distinguish that from who the business might sell to later. A common weakness is a target customer description broad enough to include almost anyone — "businesses that want to grow" is not a segment. A usable segment description states shared, observable characteristics: size, sector, behaviour, existing tooling, a triggering event that creates the need. The tighter and more specific the segment description, the more credible the sales and marketing assumptions built on top of it later, because a specific segment can be reached through specific, nameable channels, whereas a vague one cannot.
A competitive analysis that lists only direct competitors — other companies selling something similar — is incomplete. The honest version also names substitutes: adjacent products or workarounds that solve the same underlying problem in a different form, which are often the business's real competition even when they do not look like it on a feature comparison. And it names the option customers are actually most likely to choose against any new offering, which is very often simply doing nothing — continuing with the manual process, the spreadsheet, the existing supplier, the status quo, because switching has a cost even when the alternative is objectively better. A plan that never mentions "doing nothing" as a competitor has usually not spoken to enough real prospective customers, because inertia is the most common reason a genuinely good product fails to sell.
The comparison itself should be evaluative, not merely descriptive. Naming three competitors and their features tells a reader nothing about why customers will choose this business over them; the plan needs to state, plainly, what this business does that the alternatives do not and why that difference matters enough to change a buying decision.
Founders are structurally inclined to overstate defensibility, because believing the business is hard to copy is part of what makes founding it feel rational. A plan that survives scrutiny treats defensibility claims with the same scepticism a sophisticated reader will apply. Network effects, switching costs, proprietary data, regulatory licences, brand and exclusive relationships are all genuine sources of defensibility in the right circumstances, but each is frequently claimed where it does not actually apply — a small user base described as having "network effects" before any evidence of the effect existing, or "proprietary data" that is simply data any competitor could also collect. The useful version of this section names the specific mechanism, states the evidence for it existing today rather than hypothetically, and is honest about the fact that, for most early-stage businesses, execution speed and quality of relationships are the real, if less durable, forms of defensibility — which is a legitimate answer, provided it is stated as such rather than dressed up as something more permanent.
A market section should address timing directly: why is this the right moment for this business, rather than five years ago or five years from now. Genuine answers usually point to a specific shift — a technology becoming viable at a workable cost, a regulatory change opening or closing a door, a change in customer behaviour or expectation, an incumbent's weakness or exit creating room. Weak answers gesture vaguely at "the market is growing" without identifying a mechanism that makes now specifically different from before. If there is no good answer to why now, that is worth sitting with honestly before writing the rest of the plan, because a business with a compelling offering but poor timing is a much harder proposition than the offering alone suggests.
This part covers the strategic choices the business is making, how those choices translate into a route to market and a roadmap, and the team and operational capacity that will actually carry them out.
A strategy is, at its core, a decision about what not to do as much as what to do, because resources are finite and an attempt to serve every segment through every channel with every feature usually serves none of them well. This section should state the deliberate trade-offs the business is making: which segment gets priority and which is explicitly deferred, which capability gets built now and which is deliberately left thin, which market gets entered first and why. A strategy section that reads as a list of ambitions rather than a list of choices has not yet done the work strategy requires. Readers — particularly investors and board members — are testing whether the founders can say no to attractive-looking options, because the ability to prioritise ruthlessly is one of the better predictors of execution quality.
The business plan should describe the route to market briefly — which channels, roughly what sequencing, what the resourcing implication is for the numbers in the financial model — without attempting to reproduce the full go-to-market strategy document. A paragraph or two, plus a pointer to the dedicated GTM page for the detailed channel-by-channel plan, is normally sufficient here. The discipline is restraint: this section exists to let a reader confirm the financial model's revenue assumptions are connected to a real plan for reaching customers, not to relitigate the whole GTM strategy inside the business plan.
Milestones translate strategy into a sequence a reader can hold the business to. Useful milestones are specific and dated — a product capability shipped, a number of paying customers reached, a licence obtained, a hire made — rather than vague statements of intent. The roadmap section is also where a plan can quietly demonstrate realism: a roadmap with sensible spacing between milestones, acknowledging dependencies and the time real work takes, reads very differently from a roadmap that compresses eighteen months of work into a first quarter because the model needed the revenue to start early. Readers who have seen many roadmaps can usually tell the difference at a glance, and an implausibly compressed roadmap undermines trust in every other date in the document.
For most external readers, especially investors, the team section is read as closely as the numbers, on the reasoning that a strong team can recover from a flawed initial plan but a weak team will struggle to execute even a good one. The section should state, plainly, what relevant experience each founder or key team member brings to this specific business — not a generic career history, but the parts of it that bear directly on this problem, this market or this operating model. It is equally important to state the gaps honestly: the capability the team does not yet have, the hire that has to be made before a particular milestone can be hit, the area where the founders know they are weakest. Readers who spot an unacknowledged gap read the whole plan more sceptically than readers who see the same gap named and addressed with a hiring plan.
The hiring plan should connect directly to the milestones and the financial model — each significant hire dated to the point in the roadmap where it becomes necessary, with a note on what happens if that hire is delayed. Partners and advisers worth naming are ones who materially change the business's capability or credibility, not a long list of loosely affiliated names assembled to pad the section. Operations and delivery capacity should state plainly whether current capacity can absorb the growth the revenue forecast assumes, and if not, what has to be added and by when — this is the same discipline as the operating-model section in Part Two, applied forward against the growth plan rather than against today's volume. Technology and systems should describe what exists, what is being built, and what is bought in or outsourced, with enough specificity that a technically literate reader can judge whether the plan is realistic about build time and cost.
This is the part most readers turn to first regardless of what order the plan presents them in, and it is also the part most often built backwards — a target number chosen first, with assumptions reverse-engineered to justify it. This part covers what the three core statements actually tell a reader, how to build a bottom-up revenue model, cost structure and unit economics, break-even, working capital, the funding requirement, and why a single forecast is not a credible one.
The profit and loss statement answers whether the business is, or is forecast to become, profitable — revenue less costs over a period, arriving at a profit or loss figure. It is the statement most founders think about first and it is genuinely important, but it answers a narrower question than most people assume: it says nothing directly about whether the business has cash in the bank on any given day, because it recognises revenue and costs on an accounting basis that frequently does not match when cash actually moves.
The balance sheet answers a different question: what does the business own and owe at a single point in time — its assets, its liabilities, and the resulting equity. It is a snapshot, not a flow, and it is where a reader checks the underlying financial structure of the business: how much is funded by debt versus equity, what assets exist to support that debt, whether liabilities are building up faster than assets.
The cash flow statement answers the question the other two cannot: does the business have enough cash, at each point in time, to meet its obligations as they fall due. This is the statement lenders read hardest, and it is the statement most likely to reveal a problem the profit and loss statement hides entirely.
A business can report a profit for a period and still run out of cash within it, and this is one of the more counterintuitive and more common causes of business failure. It happens through timing mismatches: revenue is recognised when a sale is made or a service delivered, but the cash for that sale may not arrive for weeks or months if customers pay on credit terms, while costs — wages, rent, supplier payments — are frequently due on a much shorter cycle. A business growing quickly can make this worse, not better, because rapid growth often means paying for the inputs to serve next month's larger volume of customers before this month's revenue from a smaller volume has been collected. The plan's cash flow statement, built monthly rather than only annually in the early, more fragile years of a business, is the tool that catches this before it becomes a crisis rather than after.
The weakest, most common way to forecast revenue is to pick a plausible-sounding growth percentage and apply it to a current or estimated starting figure, year over year. This method produces a smooth, reader-friendly chart and tells the reader almost nothing about whether the number is achievable, because a growth percentage is not a mechanism — it does not say where the customers come from, at what cost, or why that particular rate rather than any other.
A bottom-up model instead forecasts each real driver of revenue separately and multiplies them together: the number of new customers a given level of sales and marketing activity can plausibly generate in a period, the price or average order value, the retention or renewal rate for existing customers, and any expansion revenue from existing accounts buying more over time. Each of these drivers can be interrogated on its own terms — is this many new customers realistic given the channel capacity described in the route-to-market section, is this retention rate consistent with what similar businesses experience, is this price consistent with the competitive positioning described in Part Three. When the drivers disagree with what the rest of the plan says elsewhere, that is a sign the model and the narrative have been built separately and need to be reconciled before either is trusted.
Costs divide usefully into fixed costs, which do not change materially with volume in the short run — rent, core salaries, insurance, software licences — and variable costs, which scale with volume — materials, transaction fees, delivery costs, sales commission tied to units sold. This split matters because it determines how the business behaves as volume changes: a business with high fixed costs and low variable costs needs volume to cover the fixed base but becomes highly profitable once it does, while a business with low fixed costs and high variable costs scales more predictably but with a lower ceiling on margin at any given volume.
Gross margin — revenue less the direct, variable cost of delivering what was sold, expressed as a share of revenue — tells a reader how much of each pound of revenue is available to cover fixed costs and, eventually, profit. Contribution margin is the closely related figure used per unit or per customer: revenue per unit less variable cost per unit, which is the figure the break-even calculation is built from.
Unit economics ask whether a single representative customer or unit is, on its own, a profitable proposition once the full cost of acquiring and serving them is accounted for — independent of overall company scale. For a subscription business, this typically means comparing the cost of acquiring a customer against the contribution that customer generates over the period they are expected to remain a customer; for a transactional business, it means comparing the contribution per transaction against the cost of generating that transaction. A business can have poor unit economics and still grow revenue quickly, which is one of the more dangerous patterns a forecast can hide, because rapid growth on a loss-making unit simply grows the loss. Readers who understand this will ask about unit economics specifically, separate from the top-line growth numbers, and a plan that has not done this analysis internally will struggle to answer well.
Break-even is the volume at which contribution exactly covers fixed costs, leaving neither profit nor loss. The calculation is straightforward once fixed costs and contribution per unit are known: break-even volume equals total fixed costs divided by contribution per unit. The table below works this through with clearly hypothetical, stated figures, purely to show the mechanism.
| Line | Value | Note |
|---|---|---|
| Selling price per unit | £40 | Assumed, illustrative |
| Variable cost per unit | £15 | Materials, fulfilment, payment processing, assumed |
| Contribution per unit | £25 | £40 − £15 |
| Total fixed costs (monthly) | £10,000 | Rent, core salaries, software, assumed |
| Break-even volume | 400 units/month | £10,000 ÷ £25 |
| Break-even revenue | £16,000/month | 400 units × £40 |
Working capital is the cash tied up in running the business day to day — principally stock sitting unsold, and money owed by customers who have been invoiced but have not yet paid. Debtor days measure how long, on average, it takes customers to pay after being invoiced; stock days measure how long inventory sits before it is sold. Both numbers directly affect how much cash the business needs to hold, because a business with long debtor days or high stock days has, in effect, lent money to its own operating cycle before it collects any cash back. A growing business with lengthening debtor days or rising stock levels can look healthy on the profit and loss statement while quietly consuming more and more cash to fund that growth, which is another route by which a profitable-looking business can run out of money — closely related to, but distinct from, the timing mismatch discussed earlier.
The funding requirement should be stated as a specific figure, tied directly to the cash flow forecast — the amount needed to reach a stated milestone or to cover the cash trough the forecast shows, plus a stated margin for contingency, rather than a round number chosen because it sounds appropriate for the stage. The use of funds should break that figure down by category — hiring, product development, marketing and sales investment, working capital, contingency — in enough detail that a reader can judge whether the request is proportionate to what it is meant to achieve. A funding ask with no clear use-of-funds breakdown, or one where the breakdown does not obviously connect to the milestones in Part Four, invites the reasonable question of what, specifically, the money buys.
A single forecast, presented as though it were a known future, is not a credible one — every forecast rests on assumptions that could turn out differently, and presenting only the case that happens to work is a signal to an experienced reader that the exercise has not been stress-tested. The base case represents the founders' honest central estimate. The downside case models what happens if key assumptions underperform — slower customer acquisition, lower retention, a delayed product milestone, a cost overrun — and, critically, shows what the business does in response: what costs are cut, what the runway looks like, at what point the business would need to raise again or change course. The upside case shows what happens if the business performs better than the central estimate, which matters to investors assessing return potential more than it matters to lenders, who are typically more interested in the downside case than the upside one. Presenting all three, with the assumptions that separate them stated explicitly, demonstrates that the founders understand the range of outcomes rather than having anchored on one convenient number.
Valuation is a negotiated figure, arrived at through methods that each have real limitations, not a fact the financial model can calculate to a precise pound. Comparable-company and comparable-transaction methods anchor to what similar businesses have recently been valued at, but "similar" is doing a great deal of work in that sentence and genuinely comparable transactions are frequently scarce, especially for an early-stage or unusual business. Discounted cash flow methods derive a value from projected future cash flows, but for an early-stage business those cash flows are themselves highly uncertain, which means the valuation inherits that uncertainty rather than resolving it. The business plan's job is not to assert a valuation as though it were settled, but to give a reader the inputs — the financial model, the market context, comparable transactions where genuinely available — needed to form their own view, and to be candid that the final figure in an equity raise is a negotiated outcome between what a founder believes the business is worth and what an investor is willing to pay, informed by but not dictated by any single method. Where valuation is central to a transaction, an independent, appropriately qualified adviser is the right source for a defensible figure — this is not something a plan on its own should try to settle.
A plan that has been through this much rigour still needs to state, honestly, what could go wrong, how the business is governed, and how the document itself gets used and revised after it is written — because a plan filed away and never revisited is a wasted effort regardless of how good it was on the day it was finished.
A risk register lists, plainly, the things that could materially harm the business, alongside a stated likelihood, a stated impact, and a mitigation for each — what the business is doing, or would do, to reduce the chance or the severity of that risk. Useful categories to check systematically include market risk (the target market does not materialise as sized), execution risk (the team cannot deliver the roadmap on the stated timeline), key-person risk (the business depends unhealthily on one individual), financial risk (the funding runway is shorter than the plan assumes, or a key customer or supplier concentration creates fragility), regulatory risk, and competitive risk (an incumbent or new entrant responds more effectively than assumed).
Certain risks are systematically understated by founders, for reasons that are understandable but worth naming so a plan can correct for them deliberately. Timeline risk is understated because founders are close to the work and tend to forget how much of any roadmap depends on things outside their direct control — hiring, third-party integrations, regulatory approval, customer decision cycles. Key-person risk is understated because founders rarely want to contemplate their own unavailability, yet a business overwhelmingly dependent on one person's knowledge or relationships is a real risk a sophisticated reader will spot immediately even if the plan does not name it. Competitive response risk is understated because it is more comfortable to assume competitors are slow or unaware than to plan for a capable, well-funded response. A risk register that reads as though nothing could plausibly go wrong is not a reassuring document; it is one that has not been honestly stress-tested, and experienced readers treat it as a warning sign rather than a strength.
Every number in the financial model rests on an assumption, and the discipline that separates a trustworthy plan from an untrustworthy one is whether those assumptions are written down somewhere a reader can find and check them, rather than buried invisibly inside spreadsheet cells. A dedicated assumptions log — even a simple table listing each key assumption, its value, and the reasoning or evidence behind it — lets a reader interrogate the model without needing to reverse-engineer it from the output. It also gives the founders themselves a single place to update when reality diverges from the plan, which happens to every plan eventually and is addressed directly later in this part.
Sensitivity analysis tests how much the outcome — profit, cash position, break-even timing — moves when one assumption is varied while the others are held constant. It answers a different, more precise question than the base/downside/upside scenarios in Part Five, which vary several assumptions together to tell a coherent story: sensitivity analysis isolates one driver at a time to show a reader which assumptions the business's fate actually hinges on. A business whose outcome is highly sensitive to, say, customer acquisition cost, but only mildly sensitive to price, is telling the founders and any reader something specific and actionable about where to focus attention and where the real fragility in the plan lies.
Governance describes how decisions actually get made and reviewed as the business runs — the board's composition and its powers, what decisions require board approval versus what the executive team can decide alone, and what reporting the board or investors receive and how often. For a plan written for an investor or board reader, this section states what oversight they are being offered in exchange for their capital or their seat. For an internal-use plan, it can be brief, since the founders and the reader are the same people, but it is still worth stating so that reporting rhythm does not drift into ad hoc as the business grows.
No plan survives contact with reality unchanged, and treating this as a failure of the plan rather than an expected feature of planning is a mistake that costs businesses real time. The right response when actual results diverge from the forecast is not to discard the plan, and it is not to quietly keep reporting against a forecast everyone privately knows is stale — both responses waste the document's usefulness. The right response is to trace the divergence back to the specific assumption that turned out to be wrong, using the assumptions log described earlier, update that assumption with what has actually been learned, and re-run the model forward from today rather than trying to force the rest of the year to still hit the original annual figure. A plan revised honestly in this way becomes more valuable over time, not less, because each revision is grounded in real, learned information rather than in the guesses that were available on day one.
A plan that is written once and reviewed only when the next funding round or annual filing forces it open has mostly stopped being useful as a management tool, whatever value it retained as a historical record. A working review rhythm typically checks the cash position and the key operating metrics monthly, checks progress against the milestones and roadmap quarterly, and revisits the full plan — strategy, market assumptions, competitive position, the funding requirement — at least annually or whenever a material assumption is shown to have changed. The frequency should be proportionate to how fast-moving and how fragile the business currently is; a business burning through a short cash runway needs a tighter rhythm than a stable, profitable one with a long history.
The appendix exists to keep the body of the plan readable while still giving a diligent reader everything they might want to check. Detailed monthly financial model outputs, full CVs, letters of intent or reference, detailed technical specifications, full legal documents or extracts, and the complete assumptions log all belong in the appendix rather than the body, where they would interrupt the argument the body is making. The test for what stays in the body versus what moves to the appendix is simple: does a first-time reader need this sentence or table to follow the argument on first read, or does a diligent reader need it available to check the argument once they have already been persuaded to look closely. The first stays in the body; the second moves to the appendix.
| Reader | What they are really testing | Reads hardest |
|---|---|---|
| Equity investor | Can this return a multiple on their capital, and what has to be true for that | Team, market opportunity, founders' judgement in the model |
| Bank or lender | Can the business service debt on schedule regardless of upside | Cash flow statement, security, downside case |
| Grant or visa body | Does the business match the scheme's published criteria | Explicit criteria mapping, not narrative persuasiveness |
| Board | What was decided, what is being measured, what is being asked of them | Milestones, resourcing, honest progress reporting |
| The founders themselves | Does the business actually make sense when written in full sentences | Every section, especially the ones that were vague in conversation |
| Statement | Question it answers | What it cannot tell you |
|---|---|---|
| Profit and loss | Is the business profitable over this period, on an accounting basis | Whether there is cash in the bank on any given day |
| Balance sheet | What does the business own and owe at this single point in time | How performance is trending period to period |
| Cash flow | Does the business have enough cash to meet obligations as they fall due | Whether the business is fundamentally profitable in the accounting sense |
A business plan is not a document written once and filed. It is one turn of a cycle that keeps running for as long as the business does, each turn grounded in more real information than the one before it.
What follows is the skeleton a written business plan typically takes, presented as a checklist. Not every plan needs every section in full — a plan for internal use can compress the market and legal sections that a lender or investor plan would expand, and the notes below flag where the emphasis usually shifts by reader.
A grant or visa application plan follows a variant of this skeleton dictated by the scheme's own published criteria, and should be checked against that criteria list directly rather than assumed to fit this generic order. In every case, the version of the plan that succeeds is the one built around a named reader and a named decision, with the numbers built bottom-up from real drivers rather than reverse-engineered from a target, and with the document treated as a living record that gets revised as the business learns, not as a one-time exercise filed away once the immediate ask has been answered.
Developed by Amit Jain at allfrontierglobal.com
© 2026 All Frontier Global · Panchkula, Haryana, India
Developed by Amit Jain at allfrontierglobal.com · purposed.in · purposed · purposed2 · merchcomp.com · uuka.org
Hand-authored essays — perspectives and figures reflect their writing date; verify current rules with official sources.
A question, a correction, or something you'd like covered. It goes straight to his inbox — no list, no newsletter.