Skip to content
Velloreth
Entrepreneurship 9 min read

Writing a Business Plan That's Actually Useful

How to write a lean business plan that helps you make decisions — not a 40-page document nobody reads. Focus on assumptions, unit economics, and milestones.

By

Nobody writes a business plan because they love writing business plans. You write one because you need to think clearly about whether this idea can survive contact with reality — or because a bank, grant committee, or investor demanded a document. The problem is that the traditional format optimizes for the wrong audience: a hypothetical reader who wants narrative and polish, not the only reader who matters — you, six months from now, making decisions under pressure.

A useful business plan is not a document. It is a set of explicit assumptions you are willing to be wrong about in public, with the math that connects them laid bare. One to three pages is often enough. The value is in the clarity, not the page count.

Why most plans fail their only reader

Classic business plans run 20–40 pages: executive summary, company overview, market analysis, products, marketing strategy, operations plan, financial projections, appendix. The format comes from loan applications and MBA assignments. It has two structural problems.

First, it front-loads narrative over evidence. Pages of market description and mission statements, with the actual unit economics — does a customer cost more to acquire than they are worth? — buried or absent. That is backwards. The numbers either work or they do not; everything else is context.

Second, it is written once and abandoned. Plans get polished before a pitch, then never revisited. When reality disagrees with the plan, there is no process for noticing or responding. The document becomes a artifact of optimism rather than a tool for learning.

CB Insights’ analysis of hundreds of failed startups found the leading root causes were poor product-market fit (43%), bad timing (29%), and unsustainable unit economics (19%). Running out of capital tops the list at 70%, but that is the final symptom — money dried up because one of the earlier three was broken. A plan that pressure-tests those three things — fit, timing, unit economics — earns its keep. A plan that does not is decoration.

The lean plan: five sections that matter

Write these five sections. Keep the whole thing under three pages. Update it when reality disagrees with you.

1. The problem and who has it (one paragraph)

Not “the market is fragmented and inefficient.” That is true of every market ever. Write something falsifiable:

Small accounting firms (5–30 people) spend 6–10 hours per month manually reconciling client bank feeds across tools. They know better software exists but lack the time to evaluate and switch. The person who feels this pain is the firm’s operations lead or senior manager, not the founding partner.

Every clause should be verifiable by talking to five people in that segment. If you cannot name the buyer, the pain, and the current workaround (including “spreadsheet” and “nothing”), you do not have a problem definition yet — you have a wish.

2. The solution and why now (one paragraph)

What you are building, in the same plain language. Then: why this is viable now when it was not two years ago. Timing failures kill companies that were otherwise right — 29% of failures in CB Insights’ recent data cited bad timing or macro conditions. Possible “why now” triggers:

  • A platform change (APIs, regulations, AI capabilities, payment infrastructure).
  • A behavior shift (remote work, new tool adoption patterns).
  • A cost collapse (what used to cost $100K to build now costs $5K).

If your “why now” is “we finally got around to it,” that is not timing — that is a hobby schedule. Find the structural reason.

3. How you will reach customers (one page)

Name the one primary channel and the fallback. Not “content, social, partnerships, paid, email, events.” One channel you will dominate first, described concretely:

Primary: founder-led outbound to operations leads at accounting firms with 5–30 staff, using LinkedIn Sales Navigator lists plus warm introductions from the two accounting associations we belong to. Target: 40 personalized emails per week, 3–5 calls booked per month.

Fallback if reply rates stay under 2% after 30 days: pivot to association webinars and co-marketing with complementary tools (practice management software).

Include the rough conversion math. If you need 10 customers this quarter and expect 3% of contacted leads to convert, you need roughly 333 contacts — about 28 per week over 12 weeks. If that volume is impossible for you to execute, the channel is wrong. Do this math before you fall in love with a strategy.

4. Unit economics (half a page, the most important half)

This is where most plans are silent and most businesses die. Four numbers:

MetricDefinitionWhy it matters
CACTotal sales and marketing cost ÷ new customers acquiredWhat one customer costs you to win
LTVAverage revenue per customer × gross margin × average lifetimeWhat one customer is worth
LTV:CACThe ratioBelow ~3:1, growth usually destroys value
Payback periodMonths to recover CAC from gross marginLonger than 12–18 months strains cash

Example for a $49/month SaaS product with 80% gross margin and 24-month average lifetime:

  • LTV: $49 × 0.8 × 24 ≈ $940
  • Target LTV:CAC of 3:1 → maximum sustainable CAC ≈ $313
  • If actual CAC (including your time at a loaded rate) is $600, the model is broken at scale — more growth just burns cash faster

You do not need precision. You need honesty. If you do not know your CAC yet, put your best estimate and the plan for measuring it in the next 90 days. Ignoring the question is how founders discover unit economics at $50K MRR when it is far too late to change course cheaply.

5. Milestones and kill criteria (half a page)

A plan without kill criteria is a plan you will never execute honestly. Write the falsifiable version:

90-day milestones:

  • 50 discovery calls completed
  • 10 paying customers
  • Gross churn under 10% monthly among those 10
  • At least 3 customers acquired through a repeatable outbound sequence (not pure favors)

If not hit: reassess problem definition and ICP before spending more on acquisition. If call volume is fine but conversion is near zero after 100 calls, the product-pain fit is wrong — pause development and re-interview.

Kill criteria: If after 6 months we have fewer than 5 paying customers and no clear evidence that the pain is urgent, we shut down or pivot. Continuing “because we are close” is not evidence.

Writing kill criteria in advance is the single highest-integrity move in planning. It converts a vague sunk-cost spiral into a decision you made while thinking clearly.

Financial projections that do not insult reality

Do not project exact revenue five years out. Model relationships instead.

Months 1–12, monthly:

  • Customers added per channel
  • Revenue (new + existing)
  • Key costs: tooling, marketing spend, your living costs
  • Cash balance

Months 13–36, quarterly:

  • Same lines, coarser granularity
  • Include a downside case (50% of planned customers, CAC 50% higher) and ask: how many months of runway does that leave?

The downside case is the useful one. The base case is what you hope; the downside case is what you survive. Founders who only model the happy path discover their fragility at the worst possible moment.

A simple rule of thumb: your plan should tell you, at any point, how many months of cash remain under the current burn. If the answer is “we will figure it out when we get there,” you do not have a financial plan — you have a prayer.

What to cut

Everything else. Specifically:

  • The 10-year vision slide. Nice for a wall. Useless for next quarter’s decisions.
  • TAM calculations with five decimal places. “Large enough to support a $50M business if we get 1% of the mid-market” is more honest than a bottom-up model built on assumptions you made up to hit a number.
  • The org chart of a company you do not have. Hiring plans belong in the plan only when they connect to specific revenue milestones (“first sales hire at $25K MRR”).
  • Feature roadmaps beyond the next validation milestone. You will be wrong. Plan the learning loop, not the product fantasy.

How to actually use the plan

A plan used once is a pitch. A plan used repeatedly is a management system. The loop:

  1. Write it. One to three pages. Explicit assumptions.
  2. Share it with five people who will be blunt. Not cheerleaders. Customers, skeptical founders, someone who has failed at a similar idea.
  3. Set a monthly review. Compare actuals to plan on the four or five numbers that matter (calls, customers, CAC, churn, cash).
  4. Update when reality disagrees. If a number is off for two consecutive months, the assumption behind it is wrong. Change the assumption, do not rationalize the gap.
  5. Rewrite fully every 6–12 months, or after any major pivot.

The document should visibly age. Edits, strikethroughs, and “we believed X, turns out Y” notes are features, not embarrassments. They are the trail of a business learning.

For investors and banks, separately

When someone else needs a formal plan — a loan officer, a grant committee, a corporate partner — write it for them, but do not abandon the lean version for your own use.

The formal version can include the classic sections, but front-load what they actually care about:

  • Banks: cash flow, collateral, repayment ability, conservative projections. They care about downside, not disruption.
  • Grant committees: mission alignment, measurable outcomes, sustainability beyond the grant period.
  • Strategic investors: market position, defensibility, synergy with their existing business.

Keep the internal lean plan as the source of truth. The external plan is a translation.

The plan is the thinking, not the document

The useful output of planning is not a PDF. It is having explicitly written down: who has this problem, why you can reach them, what a customer is worth versus what they cost, and what evidence would tell you to stop. Most founders who fail did not fail for lack of a document. They failed because those four questions were never forced into the open, one by one, while there was still time and money to adjust.

Write the plan for the person who has to make the hard call at month eight with half the runway gone and a vague feeling that things are not working. That person does not need narrative. They need numbers, assumptions, and the permission — written in advance, in their own handwriting — to change course.

Frequently asked questions

Do I still need a business plan to get funding?
For most early-stage rounds, a crisp deck plus a working product beats a 40-page plan. Some investors, banks, and grants still ask for a formal plan — write it when required, but build the lean version first because it is what you will actually use.
How detailed should financial projections be?
Detail the next 12 months monthly and the following 24 months quarterly. Beyond that, ranges are fine. Precision five years out is fiction; the value is in modeling the relationships between customers, costs, and cash, not in predicting exact revenue.
How often should I update the plan?
Review monthly against actuals, rewrite fully every six to twelve months, or whenever a major assumption fails. A plan you have not touched in a year is a historical document, not a management tool.
What if my idea changes while planning?
That is a success signal, not a failure. Pivots born from real customer evidence are the plan working as intended. Update the assumptions, keep the unit economics honest, and do not fall in love with the original version.