No shortcuts here: plan fundamentals turn on the basics

Most plans are not rejected for a single fatal flaw. They are discounted gradually, as a reader accumulates small reasons to stop believing the document. The faults below are the ones that do that damage, and all nine are fixable in an afternoon.

They are ordered by how early in the reading they cost you.

What to take away

1. Written for nobody

The draft addresses a general audience, so it explains things the actual reader knows and omits the things they need.

A loan officer wants repayment capacity, collateral, and who guarantees it. An investor wants the size of the outcome and what has to be true to reach it. A landlord wants to know you can pay rent for the term. A partner wants to know what they are agreeing to. These are four different documents built from the same research.

The fix. Write the reader's name at the top of your working copy. Ask them what they require, because most of them have a list and will send it. Then cut whatever does not serve that reader.

2. The summary written first and never revised

The executive summary is written on day one, when the business is still an idea, and then never updated as the model changes. By submission it describes a slightly different business from the rest of the document, and it is the first thing anyone reads.

The fix. Write it last, from the finished plan, against what the summary guide says has to be in it. Then check every number in it against the model. If the summary contains a figure that appears nowhere else, one of the two is wrong.

3. The market sized downward from a large number

An industry total, a chosen percentage, and a revenue forecast. The percentage was picked because it produced an acceptable answer, and no reader has ever believed one.

The fix. Build upward from something countable: households in your radius, seats times turns, qualifying firms you could contact, quotes you can produce in a week. That is the whole job of the market section. The resulting number will be smaller and it will survive questioning, which is the only property that matters.

4. Projections that do not match the rest of the document

The operations section describes one van. The forecast delivers volume that needs two. The team section has three staff; the payroll line pays for two. The marketing plan has no budget; the profit forecast assumes customers arrive anyway.

Reviewers find these quickly because checking internal consistency is the fastest way to assess a plan they cannot otherwise verify.

The fix. Before sending, reconcile explicitly: head count against payroll, capacity against volume, premises against rent, marketing activity against marketing spend, equipment against the cash outflow. Ten minutes, and it removes the most common reason plans are doubted.

5. A profit forecast with no cash flow

The plan shows a business that is profitable from month two and never shows the bank balance. Profit and cash are different, and the difference is exactly where small businesses fail.

The fix. Build a monthly cash forecast on the pattern in the projections guide, with real timing: when customers pay rather than when they are invoiced, when suppliers must be paid, when annual costs fall, and loan principal, which never appears on the profit line but leaves the account every month. Then report the lowest cash point of the year. That single figure tells a lender more than the profit total.

6. The owner works for free

Nowhere in the model is the founder paid. The business looks viable because one person's labor is priced at zero, and every reader notices, because they are being asked to fund a business that cannot afford its own operator.

The fix. Put a real salary in the model, at a rate you could actually hire that work at. Published wage data by occupation and area is a defensible starting point when you have no quote. If you intend to defer it, show the deferral with an end date and show the business affording it from that date. A plan that shows a modest but paid founder is more fundable than one that shows a large paper profit resting on unpaid work.

7. No competitors, or a table with all the ticks in your column

"There is no direct competition" tells a reader either that the research was not done or that there is no market. The feature grid with your column full of ticks tells them the section is a sales sheet.

The fix. Name the three or four businesses that would take your customers. Describe how they operate: pricing, waiting times, what they refuse to do, how they get customers, what their reviews complain about. Include the substitute nobody counts, which is the customer doing nothing or doing it themselves. Then say what happens when a competitor responds to you.

8. A team section with no weaknesses

Four biographies of uniformly impressive people, no gaps, no vacancies, and no mention of who is part-time. Readers know teams have holes, so the absence of any reads as concealment rather than strength.

The fix. Name what the team cannot do, as the team section requires, say who covers it now, say when it gets hired properly, and put that hire in the payroll schedule with a date. State the hours each founder actually commits, particularly anyone holding another job. The admitted gap makes the rest of the section believable.

9. A funding request with no sources and uses

The plan asks for an amount, and the amount appears nowhere else. There is no itemized list of what it buys, no statement of what the owner is contributing, and no demonstration that the cash flow services the repayment.

The fix. Two tables that total the same figure. Sources: owner's cash, the amount requested, anything already spent, and any other financing, described by type rather than by brand. The federal summary of funding routes is a fair checklist of what exists. Uses: equipment, fit-out, deposits, opening stock, working capital, and a named contingency. Then a short paragraph showing the repayment covered by the cash forecast after the owner is paid, using the term and rate you were actually quoted rather than assumed.

The pass that catches most of it

Read the finished draft once with a single question: which sentence would I be unable to answer a follow-up question about?

Every sentence that fails is either research you still owe or a sentence to delete. Most drafts lose a third of their length to this pass and get stronger, because what remains is the part you can defend.

Two habits prevent the whole list from recurring. Keep every number in one model with an assumptions page, and never type a figure into prose that is not taken from it. And date the document, so that when it stops matching reality you know how much of it to revisit.

Work through the fundamentals if you are starting from scratch, and use this list on the draft before anyone else reads it.

Common questions

Which of these matters most for a bank?

The cash flow, the owner's compensation, and the sources and uses. Those three carry the credit decision. The narrative sections support them.

My plan is for internal use. Do these still apply?

Most of them, and the internal version can drop the introductory framing entirely. The consistency checks and the cash forecast matter more internally, not less, because you are the one who has to live with the result.

How long should I spend fixing a draft?

Longer than you spent formatting it. The reconciliation pass and the follow-up-question pass together take an afternoon and do more than a week of rewriting.

Should I get someone to review it before submitting?

Yes, and pick someone who reads plans professionally rather than someone who will be encouraging. Ask them for the first question they would ask, then go and answer it.